Here is the daily technology #threadcast for 12/4/24. The goal is to make this a technology "reddit".
Drop all question, comments, and articles relating to #technology and the future. The goal is make it a technology center.
Here is the daily technology #threadcast for 12/4/24. The goal is to make this a technology "reddit".
Drop all question, comments, and articles relating to #technology and the future. The goal is make it a technology center.
But for every dystopian AI ad that stirs social media users into a frenzy, there are thousands of AI advertisements that fly under the radar. Why? Because most ads for AI mean nothing at all. Lots of companies have resorted to painting AI as this amorphous, magical children’s book character with no specific use case, and yet, implying that it can do almost anything.
Here’s some examples of odd AI billboards seen around San Francisco:
“Intelligence so big, you’d swear it was from Texas,” said one.
“Adapt your workforce at the speed of AI,” said another.
“AI that talks to cars and talks to wildlife,” said a third.
Uncomfortably honest
While these AI ads feel dystopian, there’s something about them that also feels honest. These ads represent the ways people are actually using AI for today, even though it’s unsettling when it’s demonstrated on your screen.
Some of the most common use cases of AI today are AI-generated art and AI companions. The former is usually a pretty low-stakes, creative task such as creating a picture or a short song. The latter can be surprisingly valuable: People are using chatbots to learn about things or talk through personal problems, much like they would with an intelligent or sympathetic friend.
Art and companionship both feel very central to the human experience, and the fact that AI is being used for both of those things today is a reality some find uncomfortable to acknowledge.
The US also added 24 additional types of chipmaking equipment to the export-control list, along with three types of software tools used for developing or producing semiconductors.
It was expected that Beijing would respond to the new US sanctions with a ban on the materials.
The concern now is that China could impose export bans on other critical items, including nickel or cobalt. The only US nickel mine will be depleted by 2028.
The system works like a giant radiator, with cold water in the loop being heated underground and then coming back to the surface where it will be used to generate electricity and piped into nearby homes through a district heating system. Eavor expectes to begin generating energy at the site in the first half of 2025, says John Redfern, Eavor's CEO and president.
However, things become truly daunting on the right half of the game's spec sheet, which lists the demands for path tracing, a more comprehensive and taxing form of ray tracing. All path-tracing presets assume players will use upscaling and frame generation to reach 60 frames per second, and AMD GPUs disappear at this point.
An industry insider, speaking to the Financial Times, revealed that Japan and several European nations have secured exemptions from the FDPR by committing to implement their own export restrictions. South Korea is reportedly pursuing a similar agreement, given the significant investments by Samsung and SK Hynix in the production of high-bandwidth memory (HBM) products.
They believe that ongoing progress in quantum computing and artificial intelligence makes the creation of advanced quantum robots very likely.
These robots could be better than regular robots in processing power, adaptability, and security, making them essential tools for many uses in research, industry, and daily life.
“To conclude, we note that while the technologies to realize advanced qubots are still dispersed and the knowledge for their integration into sophisticated units is challenging, our review has highlighted the recent developments in different fields that would together coalesce into the state-of-the-art qubots we envision,” asserted the researchers.
“The gradual maturity and renewed interest in quantum information and quantum control theories suggest that the realization of qubots is a matter of when, not if.”
To be extra clear: This operation consisted of federal bureaucrats taking it upon themselves to shut legal businesses out of the banking system.
Thanks largely to the efforts of Rep. Blaine Luetkemeyer (R-Mo.), who helped lead a multi-year effort to shut the program down, documents were unsealed in 2020 that revealed some of the highest ranking officials at the Federal Deposit Insurance Corporation helped orchestrate the program. Those documents also revealed how easy it was—and still is—for federal banking regulators to pull off this kind of debanking effort.
The romance fantasy would be fine with many men (not all men) is that men have figured out that it is a one-sided fantasy. Meaning only the man falls in love, while the woman is not in love, she has conquered the man. Similar to the Beauty and the Beast story. So the woman wants a man to fall in love with her, but she still reserves the right to upgrade to the next alpha male she comes across, which is why women are not loyal. Men realize that women are not really romantic, they use romance as a tool to get what they want.
International concerns, risk, and more factors
The report also cites critics who state that the hypersonic weapon’s high speed and unpredictable flight path could generate uncertainty about its intended target. This could heighten the risk of miscalculation or unintended escalation in the event of a conflict.
Moreover, the ambiguity over whether a hypersonic weapon is carrying a nuclear or conventional warhead could also cause unintended escalation.
The other side’s arguments that Russia and China possess these type of weapons which can be used to target the U.S. and wreak havoc on its defense system in case of escalation are also taken into account.
The report also quotes experts who state that the U.S. should take some measures like expanding New START, negotiating new multilateral arms control agreements, and undertaking transparency and confidence-building measures with other countries to mitigate risks.
"Our solid Q3 results were underpinned by continued strong profitability and cash flow," Okta CEO Todd McKinnon said in a statement. "The focused investments we've made in our partner ecosystem, the public sector vertical, and large customers are materializing in our business with each of these areas contributing meaningfully to top-line growth."
For the fourth quarter, Okta said it expects to report revenue between $667 million and $669 million, topping the $651 million average estimate, according to LSEG. The company expects to report earnings of 73 cents to 74 cents per share for the period, which also exceeded estimates.
Salesforce also raised the low end of its revenue guidance, expecting a range of $37.8 billion to $38 billion for its fiscal 2025. That's up slightly from $37.7 billion to $38 billion previously. The new range puts the midpoint for Salesforce's fiscal 2025 revenue guidance at $37.9 billion, ahead of analysts' expectations of $37.86 billion.
"We delivered another quarter of exceptional financial performance across revenue, margin, cash flow, and cRPO," Salesforce CEO Marc Benioff said in a statement. "Agentforce, our complete AI system for enterprises built into the Salesforce Platform, is at the heart of a groundbreaking transformation."
In a call with analysts, Benioff boasted about Salesforce's latest artificial intelligence push, including the company's AI-powered chatbots dubbed Agentforce, which investors are closely monitoring for growth. Salesforce's Agentforce product is an example of so-called AI agent technology. Several companies have said they believe that these advanced chatbots represent the next logical step from ChatGPT and other related tools powered by large language models.
"We're delivering these incredible Agentforce capabilities as well," Benioff said. "This is a bold leap in the future of work, where AI agents let humans unite to transform all of our customer interactions."
Benioff also revealed that he ruptured his achilles tendon on a recent birthday scuba-diving trip to Fakarava, an atoll in French Polynesia. Benioff expressed disappointment that the hospital that treated him couldn't schedule his follow-up appointments using AI agents.
"That is the message to our customers, which is how are you going to give some of your people a break, let them get back to their strategic work, let them focus on what really matters," Benioff said.
The company in August announced that Amy Weaver would step down from her role as chief financial officer but remain in the position until the company appoints a successor, after which she will become an advisor. That same month, activist investor Starboard Value revealed that it boosted its position in Salesforce by roughly 40% in the second quarter following the firm issuing a letter earlier in the year saying that Salesforce was continuing to move "in the right direction" in regard to improving its profit margin.
Starboard Value released a presentation in October in which it noted that Salesforce "can continue to become more efficient and more profitable."
The shortlist of candidates could not be fully learned, but Bloomberg citing people familiar with the matter reported that Marvell chief Matt Murphy was among those being considered.
Representatives for Intel did not return multiple requests for comment. Tan, whose involvement in the replacement process was first reported by Reuters, could not be reached for comment.
Interim executive chair Frank Yeary and a board-level committee are overseeing the search for a new CEO, Intel disclosed in a regulatory filing, and the process remains in its early stages.
Intel's board may opt for an internal promotion, and internal candidates could include CFO David Zinsner or Intel products chief MJ Holthaus, the interim co-CEOs.
Intel abruptly ousted CEO Pat Gelsinger over the weekend, with people familiar with the matter previously saying that the board had lost faith in his ability to execute on a turnaround at the chipmaker, which has been lagging behind Nvidia for more than a year.
Gelsinger will receive roughly $10 million in severance, Intel disclosed in a regulatory filing.
The email feels more like something you’d write to a distant work colleague, rather than the way you’d speak with a loved one you see every day. While it’s not an inappropriate message, it’s cold and could’ve been sent to anyone.
This example from the Browser Company was the latest AI ad that told a different story about the technology than it intended — but perhaps a truer one. It strikes the same sensitive nerve that so many other AI advertisements have in the last year. In trying to promote AI, tech companies can’t help but show how it removes us from the very activities that make us human.
Of course, Miller could (and probably should) have customized the prompt to be warmer and address his wife as such, but that’s missing the larger point.
Miller didn’t really talk to his wife in this case. The AI browser took a genuine act of human kindness and turned the exchange into something that feels impersonal — largely because it is. AI is further abstracting what it means to connect. At one point, connection meant talking in person; then, around the turn of the century, it migrated to texts sent over the internet. Now humans are starting to experiment with using AI to talk with each other, and in some cases, just talking to AI — removing the need to connect with a human altogether.
You could say I’m cherry-picking this ad, but it’s a story that tech companies keep accidentally telling over and over again. This part of the ad was likely intended to show how Dia could retrieve links from multiple web pages and understand their context — an impressive feat for an AI system these days.
But this was yet another example of how generative AI can reduce our humanity.
Consider Google’s ad earlier this year, where a father and daughter used Gemini to create an AI-generated fan letter to their favorite Olympian. The company later pulled the ad after facing backlash for taking a sweet father-daughter exchange and automating it away.
Or maybe you remember how Apple unveiled its AI features at WWDC this year: showing how you can go up to a stranger’s dog, point your iPhone at it, and have Apple Intelligence tell you what breed it is. Many people pointed out that you could have just asked the stranger what type of dog they have, and maybe you would have found a friend alongside the dog’s breed.
Apple's Visual Intelligence enables you to use the camera to look things up. "It will change the way you interact with iPhone"
Maybe, it's just "cynical me", but this example of the looking up the cute dog was a perfect opportunity for two humans to connect in a human way by… pic.twitter.com/CZpPb0ufCU
— Willem (@wlmiddelkoop) September 10, 2024
Months earlier, Apple apologized for an ad it ran where the company quite literally crushed objects representing human creativity, in favor of an iPad. It wasn’t an ad for AI, but it had the same effect: technology that reduces our humanity.
The most extreme example of these AI ads came from an AI startup called Friend. The startup released a promotional video showing how lonely young people could have a virtual companion in the startup’s AI device that they wear around their neck, instead of talking to others.
“Geminiiiiiiiiiice,” said yet another.
See what I mean? I have no idea what these things do, and yet, it all feels inoffensive, vaguely describes AI in a magical way, and gets the product in front of my face.
Maybe that’s the point.
This banal tapestry of AI advertisements depicts the industry more accurately than any one company can. Most companies don’t really know what AI is good for, and the ways people use AI today are somewhat discomfiting, automating many of the very tasks that make us human.
‘Big picture problem’
Founded out of New York in 2022 by Misha Herscu (CEO) and Skyler Thomas (CTO) — pictured above — Cake launched last year and is already working with customers like AI bioscience startup Altis Labs and data intelligence insurtech Ping. However, the company hasn’t been making much noise in public until now.
On top of its formal unveiling today, Cake said it has raised $13 million since its inception. This includes $3 million in pre-seed funding through its formative couple of years, and a recent $10 million seed round led by Google’s Gradient Ventures.
“We haven’t been super secretive; we’ve just been building, and working with customers,” Herscu explained to TechCrunch in an interview last week.
Previously, Herscu founded an AI company called McCoy Medical Technologies that was focused on machine learning infrastructure for radiology, and sold it in 2017 to IT vendor TeraRecon. He later joined New York VC firm Primary Venture Partners as “operator in residence,” where he pursued his next venture by chatting with hundreds of data science and AI executives.
“I did over 200 customer discovery calls, asking what their biggest pain points and bottlenecks are,” Herscu said. “The biggest problem wasn’t a single part of the stack, such as setting up a vector database or data pipeline. It was that there are a ton of different components across a very rich ecosystem. How do you go about integrating everything reliably, and making it production ready?”
This is what Herscu refers to as the “big picture problem,” and is where his new business enters the fray.
Cake is all about making sense of the myriad open-source components that constitute the modern AI stack, and providing bundled, managed, open-source AI infrastructure for small teams. This isn’t about building a business around a single open-source project as countless companies have done; instead, it’s about assembling and serving a curated selection of open-source projects across an entire stack and making it run smoothly.
Let’s say a large financial services company has millions of documents containing complex financial data, and it wants to do RAG (retrieval augmented generation) against these files to improve the quality of the responses to natural-language queries. If an off-the-shelf product isn’t up to the task, or is unsuitable for compliance reasons, the company would have to build its own system by installing and stitching multiple different components. That’s a time-consuming endeavor that Cake can take care of.
Elsewhere, a hospital might need to construct a secure system for analyzing images from CT scans, or an e-commerce company might want to upgrade its recommendation engine. These are all potential use-cases for Cake.
Parallel development
Cake’s CTO Thomas previously worked at IBM as a chief architect, and more recently he was a distinguished engineer and director of strategy at Hewlett Packard Enterprise, which acquired a previous company he worked at called MapR.
Thomas says he has worked across hundreds of projects through the years, with large and smaller customers, and he noticed a trend permeating pretty much all of them — every one was using open-source tools in some way, much of it fresh out of research labs. Still, using them in the enterprise wasn’t easy.
“It takes a huge amount of time for even the largest enterprises to take what’s coming out of the labs and integrate it into what they do,” Thomas told TechCrunch. “A lot of that is because most of it isn’t ready for the enterprise — it might not have authentication and authorization, and enterprises have to do that themselves.”
There are parallels to what Cake is striving for here. In Europe, we have the likes of Finnish Aiven, a $2 billion unicorn, which is doing something similar but with a focus on data infrastructure. Perhaps the most obvious comparison would be Red Hat, which IBM acquired for $34 billion and is best known for its enterprise-grade Linux operating system (RHEL).
“In the early days of Linux, there were thousands of open-source packages that everyone wanted to use, but weren’t integrated and weren’t secure,” Thomas said. “There just wasn’t a support model for it, and so the Red Hats of the world made Linux safe for the enterprise. We want to do a similar thing for AI today.”
While there are plans to eventually introduce a hosted version of Cake, for now companies have to run it in their own environments. For many, this won’t be an issue because data privacy stipulations mean they can’t send data outside their own systems anyway. But a hosted version might be appealing to organizations with lower compliance obligations.
“It is actually easier for us if we can control the cloud,” Herscu added.
You might wonder why companies aren’t making the obvious AI ads: AI does your boring job so you can spend more time at the beach, with your friends and family, or pursuing your passions. That’s what Zoom’s CEO laid out as his vision for AI, and it’s probably the most optimistic outcome we’ve seen someone describe.
Perhaps the reason we’re not seeing more tech companies promise that future is because AI is not ready to do your job. There’s also a conflicting vision there: If AI can do some of your job, couldn’t it also replace you altogether? While it may be a while before AI can actually do your job, it seems most companies are steering clear of that message altogether.
I can’t say what the “right way” to be promoting AI is right now, but I do think the status quo for AI ads is objectively strange. Whereas previous generations of technology promised to liberate us, connect us, and make us smarter, the overarching promise of AI is still unclear. If companies are looking for another uplifting message to sell their software with, automating core aspects of the human experience ain’t it.
Aside from lead investor Gradient, Cake’s seed round saw participation from its pre-seed investor Primary Venture Partners, as well as Alumni Ventures, Friends & Family Capital, Correlation Ventures, and Firestreak Ventures.
The hitherto unannounced $10 million seed round, which closed back in April, is indicative not only of the founders’ backgrounds but also the company’s traction. Herscu said that the company is already looking toward its next financing round, with tentative plans to raise again around the middle of 2025.
“From a traction standpoint, we look more like a Series A company already. We were able to get there pretty quickly,” Herscu said. “When we go to the Series A, it’ll probably look more like a Series B.”
On the hardware front, Growl is divided into two main parts. There’s a frame that’s designed to be screwed to your wall. You then attach a boxing bag component onto the frame. Most of the sensing intelligence is located in the frame. The boxing bag itself is mostly made of foam and artificial leather, with five or six accelerometers inside to detect punching power etc.
The frame includes a key component of the device: a 4K projector. It can project a human-sized coach onto the boxing bag and various metrics on the wall. Growl has also been using the projector for a gaming experience (think Tap Tap Revenge on a boxing bag).
Using a projector is a smart approach, as you aren’t punching a display, which means that you can’t break it.
“That’s the beauty of our technology. We mostly use light, and you’ll never break light,” Desrumaux said.
On each side of the frame, there’s one speaker and two cameras with infrared sensors. There’s also an additional camera near the top of the frame, along with some time-of-flight sensors.
The cameras and sensors are used to calculate the impact zone of your punch and your posture. The accelerometers behind the boxing bag are used to calculate the intensity of the punch.
“The core feature of our product is that we reproduce the physical presence of a one-on-one coach, as if they were physically there with you,” Desrumaux said.
Growl plans to price its device in the same range as Tonal devices. The full price is slated at around $4,500, but most customers will likely opt for a financing option with monthly payments. There will be a subscription plan to access new content as well. The idea is that it shouldn’t cost more than a premium gym membership in total, the startup said.
The Growl team is well aware that the connected fitness industry is crowded these days. But most of these companies have focused on cycling, strength and rowing. And those companies have faced some growing pains.
“The worst thing that ever happened to Peloton was the Covid pandemic,” Desrumaux suggested. “When you look at Peloton, at the time of their IPO, that is a few months before Covid, in September 2019, they have 800,000 households in the U.S., they’re growing 100% a year, they’re making a 50% margin on hardware, a 60% margin on software, and they have 0% in EBITDA. So they’re practically break-even.”
During the lockdown periods, Peloton launched several new products and spent a lot of money. “And then, in the end, Covid was just a brief interlude. The market returned to its original growth trajectory,” he added.
Now, Peloton seems to be out of the woods after a difficult reality check. That’s why Desrumaux believes connected fitness is still an interesting industry as long as you maintain financial discipline — which is his aim with Growl.
Based in Austin, Texas and Paris, France, the company counts Sam Bowen, the former VP of hardware engineering at Amazon, Peloton and Tonal, as an advisor. The startup has raised $4.75 million in seed funding from Skip Capital, Kima Ventures, Teampact Ventures and various business angels, such as former UFC Heavyweight Champion Ciryl Gane.
Veeam’s CEO, Anand Eswaran (pictured above), told TechCrunch that while the company didn’t need to raise capital, the timing was right to give their early investors and employees some liquidity. It also wanted to bring on strategic investors on the journey to an eventual IPO, Eswaran said, although he added the company doesn’t have a timeline for an exit yet.
“Welcoming select blue-chip, top-tier investors was critical,” Eswaran said. “If you look at the next phase of the journey, several milestones like an IPO, having a diversified, top-tier set of investors for such milestones is critical.”
Veeam seems to have accomplished part of that goal, given that the investors who have now come in all invest in both private and public assets — they would be less likely to sell their stake in Veeam at an IPO, unlike many private equity or venture capital funds.
The fresh $2 billion will be used to bulk up the company’s research and development team, and for acquisitions as the right opportunities arise, Eswaran said.
While the company doesn’t have specific M&A goals, it’s keeping an eye out for potential acquisition targets. Eswaran noted a few areas, like companies with AI technology that could help Veeam scale, data resilience businesses focused on workloads Veeam isn’t currently active in, and companies with adjacent or complementary tech.
When Veeam was founded 18 years ago, there wasn’t much competition in the data resilience sector, but the company now operates in a relatively crowded market, especially with the rise of both cloud and AI.
There are numerous other companies looking to help enterprises manage their data and keep it secure: Rubrik, for one, hit the public market with a $5.6 billion valuation earlier this year, and we also have late-stage startups like Cohesity (raised nearly $1 billion in venture capital) and Druva (raised more than $475 million).
“Competition is actually a good thing,” Eswaran said. “It pushes the industry and all of us to innovate faster.”
Seattle, Washington-based Veeam was founded in 2006 and has always been focused on data resilience and recovery. Its tech covers 150 different workloads, spanning SaaS to AI, and its customers include large corporations like Shell and Deloitte as well as federal and local governments like the City of New Orleans.
“We feel we are the best platform right now as shown by our No. 1 market share,” Eswaran said. “If anything, this oversubscribed round just validated our direction.”
The CEO has previously stated that Chinese AI is quickly catching up to Western AI thanks to its embrace of the open source movement.
Delangue warned on the podcast that the strong concentration of top open source models coming from China is a “fairly new development and I’m a little worried about it to be honest,” he said. “It’s important that AI is distributed between all countries — that there isn’t one or two countries which are much stronger than the others.”
Hugging Face is the world’s biggest platform for AI models and a popular place for Chinese AI companies to showcase their latest LLMs. In fact, Hugging Face’s CTO announced this week that the default model on HuggingChat is Qwen2.5-72B-Instruct, which was developed by Chinese e-commerce giant Alibaba.
This particular model does not appear to censor questions about the Tiananmen Square massacre or other issues typically censored by the Chinese government.
A different model from Alibaba’s Qwen family available on HuggingChat, QwQ-32B, however, clearly does when TC asked:
Alibaba’s QwQ-32B model won’t answer a question about the Tiananmen Square protests. Image Credits: Hugging Face (screenshot)
DeepSeek, another Chinese model that went viral in the AI community for its reasoning capabilities, also extensively censors topics deemed sensitive by the Chinese government, TechCrunch previously reported.
Chinese AI companies are in a tough spot, as the Chinese government forces their models to “embody core socialist values” and comply with its already-extensive censorship system.
A Hugging Face spokesperson declined further comment but pointed out that Delangue recently predicted China would start to lead the global AI race in 2025.
Unwinding expensive private capital investments is not what most companies say they will do with their IPO funds. They tend to dedicate the money to running their businesses, or for possible acquisitions. In this case, ServiceTitan says it will use whatever is left over as working capital for the company or other corporate uses.
This latest disclosure follows news that ServiceTitan sold its soul, so to speak, in 2022 when it raised a Series H round by agreeing to grant the investors in that round a “compounding IPO ratchet structure.”
This price range means that the company will almost certainly have to grant those Series H shareholders a bunch more shares as part of the IPO, too. If its IPO share price was less than what the Series H investors paid, ServiceTitan agreed to cover the losses, and every quarter it delayed an IPO after May 22, 2024, the company agreed to owe those investors even more. They paid $84.57 a share, it disclosed.
VC Alex Clayton, general partner at late-stage firm Meritech Capital and known for his IPO analysis, was the first to point out that painful ratchet structure in a blog post that went viral. He tells TechCrunch that spending a large chunk of its IPO cash for ServiceTitan to get out of the preferred stock deal “makes sense.”
“They clearly want to have a cleaner cap table so are using the proceeds to buy these back. They could buy this back anytime and now have the cash to do so,” he said.
Still, the company also appears to need the cash for the business. While losses are narrowing, at the end of its fiscal 2024, it lost $183 million from operations and logged a net loss of $195 million when factoring in interest and other costs.
Clayton, however, also believes the bankers are playing their typical IPO pricing games with that $52 to $57 range, which was lower than he expected. That means that the company might actually price above the range — which helps create positive headlines and excitement for the offering. If so, ServiceTitan can clean its cap table and walk away with more cash.
“This is just the initial range; it’s likely to be priced and trade higher. Remember, bankers want an ‘IPO pop’ and it will not excite companies to work with them if they price the IPO too high and it trades below issue price. I suspect the company will trade in the high $60s or low $70s,” he said.
In that vein, ServiceTitan also better clarified who will be eligible to buy stock in its direct share program. ServiceTitan is setting aside 5% of its shares to sell to friends and family of the founders and, it clarified, to certain C-suite decision-makers of its customers.
While there could be some conflict-of-interest issues there — a customer who is also a shareholder selecting a vendor — such private stock sales have become more common. Reddit, which went public earlier this year, for instance, did so for its moderators.
The China Association of Automobile Manufacturers accused Washington of having "arbitrarily amended the control rules, seriously affecting the stable supply of US chip products".
"The Chinese auto industry's trust and confidence in the procurement of US chip products is being shaken, and US auto chip products are no longer reliable and safe," the association said.
The newly announced limits also cover super-hard materials, such as diamonds and other synthetic materials that are not compressible and extremely dense, reports AP. These are used in cutting tools, disc brakes and protective coatings.
"To safeguard national security interests and fulfil international obligations such as non-proliferation, China has decided to strengthen export controls on relevant dual-use items to the United States," China's Commerce Ministry said.
China accounts for 94% of the world's production of gallium, and 83% of germanium. Exports of the two materials from China to the US had fallen to zero following the introduction of restrictions last year.
China accounted for 48% of globally mined antimony last year. The country's overall shipments of the material in October dropped by 97% from September following the introduction of the export controls.
China's move comes a day after the Biden administration announced a new set of sanctions designed to restrict China's access to critical Western technologies that could have AI and military applications. At least 140 Chinese organizations were added to the Entity List, joining the likes of Huawei, meaning US companies must apply for and acquire a license, which is rarely granted, if they wish to do business with them.
From Wires and Gels to a Simple Spray
For anyone who’s endured an EEG, the process feels beyond cumbersome. A technician marks a dozen or more points on your scalp, applies sticky gel, and painstakingly glues electrodes to your head. These electrodes connect to a machine via a tangle of wires. The setup is clunky, uncomfortable, and prone to failure as the gel dries out.
Lu and her collaborators wanted something better. They’ve spent years designing tiny sensors, known as electronic tattoos or e-tattoos, which stick to the skin to measure everything from heartbeats to muscle fatigue. But applying these tattoos to the scalp posed a challenge.
“Designing materials that are compatible with hairy skin has been a persistent challenge in e-tattoo technology,” Lu explains.
Then, their breakthrough came in the form of a new kind of liquid ink made from conductive polymers. The ink flows effortlessly through hair to reach the scalp, where it dries into a thin, flexible film. Once in place, the tattoo acts as a sensor, picking up brain signals with precision.
The process is remarkably simple. Researchers map out the electrode locations using a computer and then use a specialized inkjet printer to spray the liquid ink onto the scalp. Within minutes, the ink dries, forming a lightweight, unobtrusive tattoo. No discomfort; no glue; no wires.
A More Reliable Brain Listener
To test their invention, the researchers applied these temporary tattoos to the scalps of five volunteers with short hair. For comparison, they also attached traditional EEG electrodes next to the tattoos. The results were telling.
After six hours, the gel on the conventional electrodes began to dry out. A third of them stopped working entirely, and the remaining electrodes delivered weaker, noisier signals. The e-tattoos, in contrast, performed flawlessly for at least 24 hours.
The researchers didn’t stop there. They redesigned the tattoos to replace the wires typically used in EEG setups. By tweaking the ink’s formula, they printed thin lines connecting the sensors to a small data collection device. These printed “wires” conducted signals reliably without picking up interference.
Brain-Machine Interfaces, Redefined
Beyond making EEG tests more comfortable and accurate, the researchers see their technology playing a role in the rapidly evolving field of brain-computer interfaces. These devices, which translate brain signals into commands for external systems, hold immense promise for people with disabilities, offering the ability to control prosthetic limbs or communicate without speaking.
Right now, brain-computer interfaces are bulky and unwieldy. Users typically wear large headsets bristling with electrodes. José Millán, another co-author from the University of Texas at Austin, believes e-tattoos could change that.
A Glimpse of What’s Next
The idea of wearing electronics on your skin might seem futuristic, but it’s becoming a reality. In recent years, e-tattoos have made their way onto athletes’ chests, tracking heart rhythms, and onto people’s arms, monitoring muscle fatigue. Now, they’re climbing the ladder—straight to the brain.
For now, the liquid ink is confined to the lab, but its creators see a world of possibilities ahead.
“E-tattoos represent a new frontier in wearable technology,” says Lu. “This is just the beginning of what we can achieve.”
The data also includes the average selling price (ASP) for the boards. Intel's products were more expensive with an ASP of $183, while AMD's ASP was $170.
It's not just in Germany where AMD is dominating. Eight of the top ten best-selling motherboards on Amazon are AMD mobos, including the top two best sellers.
Intel's Arrow Lake desktop CPUs have proved very disappointing, especially when it comes to gaming. AMD's Zen 5 chips proved quite underwhelming too, admittedly, but then the AMD Ryzen 7 9800X3D arrived and was crowned the new gaming CPU king. The processor was so popular that it sold out everywhere, leading to scalpers selling it on eBay for up to $1,500.
Intel is going through the roughest patch the company has endured in decades. The situation has become so dire that the board forced out CEO Pat Gelsinger this week.
In the long run, the photonic processor could lead to faster and more energy-efficient deep learning for computationally demanding applications like lidar, scientific research in astronomy and particle physics, or high-speed telecommunications.
“There are a lot of cases where how well the model performs isn’t the only thing that matters, but also how fast you can get an answer. Now that we have an end-to-end system that can run a neural network in optics, at a nanosecond time scale, we can start thinking at a higher level about applications and algorithms,” says Saumil Bandyopadhyay ’17, MEng ’18, PhD ’23, a visiting scientist in the Quantum Photonics and AI Group within the Research Laboratory of Electronics (RLE) and a postdoc at NTT Research, Inc., who is the lead author of a paper on the new chip.
Bandyopadhyay is joined on the paper by Alexander Sludds ’18, MEng ’19, PhD ’23; Nicholas Harris PhD ’17; Darius Bunandar PhD ’19; Stefan Krastanov, a former RLE research scientist who is now an assistant professor at the University of Massachusetts at Amherst; Ryan Hamerly, a visiting scientist at RLE and senior scientist at NTT Research; Matthew Streshinsky, a former silicon photonics lead at Nokia who is now co-founder and CEO of Enosemi; Michael Hochberg, president of Periplous, LLC; and Dirk Englund, a professor in the Department of Electrical Engineering and Computer Science, principal investigator of the Quantum Photonics and Artificial Intelligence Group and of RLE, and senior author of the paper. The research appears today in Nature Photonics.
Machine learning with light
Deep neural networks are composed of many interconnected layers of nodes, or neurons, that operate on input data to produce an output. One key operation in a deep neural network involves the use of linear algebra to perform matrix multiplication, which transforms data as it is passed from layer to layer.
But in addition to these linear operations, deep neural networks perform nonlinear operations that help the model learn more intricate patterns. Nonlinear operations, like activation functions, give deep neural networks the power to solve complex problems.
In 2017, Englund’s group, along with researchers in the lab of Marin Soljačić, the Cecil and Ida Green Professor of Physics, demonstrated an optical neural network on a single photonic chip that could perform matrix multiplication with light.
But at the time, the device couldn’t perform nonlinear operations on the chip. Optical data had to be converted into electrical signals and sent to a digital processor to perform nonlinear operations.
“Nonlinearity in optics is quite challenging because photons don’t interact with each other very easily. That makes it very power consuming to trigger optical nonlinearities, so it becomes challenging to build a system that can do it in a scalable way,” Bandyopadhyay explains.
They overcame that challenge by designing devices called nonlinear optical function units (NOFUs), which combine electronics and optics to implement nonlinear operations on the chip.
The researchers built an optical deep neural network on a photonic chip using three layers of devices that perform linear and nonlinear operations.
A fully-integrated network
At the outset, their system encodes the parameters of a deep neural network into light. Then, an array of programmable beamsplitters, which was demonstrated in the 2017 paper, performs matrix multiplication on those inputs.
The data then pass to programmable NOFUs, which implement nonlinear functions by siphoning off a small amount of light to photodiodes that convert optical signals to electric current. This process, which eliminates the need for an external amplifier, consumes very little energy.
“We stay in the optical domain the whole time, until the end when we want to read out the answer. This enables us to achieve ultra-low latency,” Bandyopadhyay says.
Achieving such low latency enabled them to efficiently train a deep neural network on the chip, a process known as in situ training that typically consumes a huge amount of energy in digital hardware.
“This is especially useful for systems where you are doing in-domain processing of optical signals, like navigation or telecommunications, but also in systems that you want to learn in real time,” he says.
The photonic system achieved more than 96 percent accuracy during training tests and more than 92 percent accuracy during inference, which is comparable to traditional hardware. In addition, the chip performs key computations in less than half a nanosecond.
“This work demonstrates that computing — at its essence, the mapping of inputs to outputs — can be compiled onto new architectures of linear and nonlinear physics that enable a fundamentally different scaling law of computation versus effort needed,” says Englund.
The entire circuit was fabricated using the same infrastructure and foundry processes that produce CMOS computer chips. This could enable the chip to be manufactured at scale, using tried-and-true techniques that introduce very little error into the fabrication process.
Scaling up their device and integrating it with real-world electronics like cameras or telecommunications systems will be a major focus of future work, Bandyopadhyay says. In addition, the researchers want to explore algorithms that can leverage the advantages of optics to train systems faster and with better energy efficiency.
This research was funded, in part, by the U.S. National Science Foundation, the U.S. Air Force Office of Scientific Research, and NTT Research.
Tokenization will occur using Backed’s Tokenization-as-a-Service (TaaS) platform. Those tokens will be backed 1:1 by units, ensuring their value corresponds to the real-time Net Asset Value (NAV) of the fund—essentially the fund’s worth per share.
“Tokenizing the Fortlake Sigma Fund is a major step forward in the blockchain landscape,” said Adam Levi, co-founder of Backed. “It showcases how institutions can use blockchain technology to simplify back-office operations and reach new capital.”
Currently only 32 countries in the world have geothermal power plants in operation. There are fewer than 700 power plants around the world, generating around 97 Terawatt hours (TWh) in 2023 between them. That is less than half the amount of electricity generated by solar in the US alone and far short of estimates for the potential contribution that geothermal could make to the global energy mix. Some estimate that geothermal could contribute around 800-1400TWh of electricity annually by the middle of the century with a further 3,300-3800TWh per year of heat.
"The Earth itself has the potential to address a variety of hurdles in the transition to a clean energy future," argued Amanda Kolker, geothermal programme manager at the National Renewable Energy Laboratory (NREL) in the US, when releasing a report on the potential of geothermal energy in 2023.
One reason geothermal is not more widespread is the high upfront investment needed to extract that energy. But physically reaching it has also been beyond us so far.
For other parts of the world to enjoy a part of this geothermal bonanza of clean energy, we need to drill deeper to reach the temperatures needed to generate electricity or provide large-scale heating for nearby neighbourhoods.
Across much of the planet, temperatures increase by 25-30C (77-86F) on average every kilometre you go down through the Earth's crust. In the UK, for example, the subsurface temperature at around 5km (3 miles) down is about 140C (284F), according to the British Geological Survey.
Drill down far enough, though and it is possible to reach a point where water temperatures surpass 374C (705F) at pressures above 220 bars (one bar being average pressure on the Earth's surface at sea level). This is where water enters an energy-intense state known as supercritical, where it exists in a form that is neither liquid or gas. The hotter and more pressurised it is, the greater energy it contains.
In fact, a single superhot geothermal well could produce five to 10 times the energy that commercial geothermal wells produce today, according to the NREL.
One major hurdle, however, is that conventional rotary drills – even those tipped with diamond – are ill-equipped to excavate to the kind of depths needed to access these kinds of temperature. In the mysterious deep underworld of uncertain geology, extreme temperatures and huge pressures, drill components can fail frequently, while keeping holes from becoming blocked is a constant battle.
In 2009, for example, a team working on the Iceland Deep Drilling Project inadvertently tapped supercritical conditions when it drilled into a magma chamber at the Krafla volcano, about 1.2 miles (2km) below the surface. The superheated steam emitted from this well was highly acidic, making it difficult to use. The high pressures and temperatures involved also made it difficult to control, and it had to be intermittently discharged for around two years before a valve failure forced the hole to be sealed.
The deepest hole ever dug by humans dates back to the Cold War, however, when there was a race between the superpowers to drill as far into the Earth's crust as possible. The Soviets managed to plough their way through 7.6 miles (12.2km) of rock – creating the Kola Superdeep Borehole, on the Kola Peninsula, high in the Arctic Circle. It took them almost 20 years to reach that depth and it remains the deepest humans have managed to delve into the Earth. (Read more about the Kola Superdeep Borehole in this article by Mark Piesing.)
The NREL estimates that the cost of drilling a 1km deep well is around $2m (£1.57m) while drilling four times that depth can cost between $6m-$10m (£4.7m to £7.87m) with current technology.
These depths will allow nearly universal access to geothermal power – Igor Kocis
Yet deep geothermal energy could provide some considerable cost savings when compared to conventional geothermal, due to the higher temperatures and pressures that can be accessed further into the Earth's crust. Some studies have suggested deep geothermal energy could supply heating for communities at costs similar to other forms of heating, such as using gas, but with fewer greenhouse gas emissions.
With this in mind, some pioneering researchers and companies are turning to new types of drills and drilling techniques to bore some of the deepest holes ever created in the hope of bringing geothermal energy to parts of the world that never thought it was possible.
Quaise Energy, a spin-off from the Massachusetts Institute of Technology (MIT), for example, are aiming to drill holes as deep as 12 miles (20km) to access temperatures of 500C (932F) or more. To do so, they are turning to a tool that draws on years of research into nuclear fusion power. "While others are putting shovels in the ground, we're putting microwaves in the ground for the first time," says the company's co-founder Matt Houde.
He and his colleagues are experimenting with millimetre-wave directed energy beams that vaporise even the hardest rock. It focuses high-powered beam of radiation similar to microwaves but at a higher frequency onto a segment of rock, heating it up to 3,000C (5,432F) so that it melts and vaporises. By directing the beam so it bores through the rock, holes can be created without the debris and friction created by traditional drilling techniques.
"Millimetre-wave drilling is a process that can operate largely independent of depth," says Houde. "And millimetre-wave energy can also transmit through dirty, dusty environments."
The technology has grown out of nuclear fusion plasma experiments conducted by Paul Woskov, an engineer at MIT's Plasma Science and Fusion Centre. Millimetre-wave directed energy has been explored as a way of heating up plasma in nuclear fusion reactors since the 1970s, but a few years ago Woskov hit upon another use for the technology. He started using millimetre-wave beams generated by a device known as a gyrotron to melt through rock.
But so far the technology has only been tested in the laboratory, drilling shallow holes in relatively small samples of rock, but the company claims it can drill through rock at around 3.5m (11.5ft) per hour. While this is slow compared to traditional drilling techniques, there are other benefits as the "drill bit" isn't physically grinding through the rock, it should not wear out or need replacing. Quaise Energy are now at the final stage of laboratory testing of millimetre-wave technology prior to beginning field trials in early 2025.
But transferring the millimetre-wave drilling technology from the laboratory to a full-scale drilling operation will still be a challenge.
"They have never been used before in the deep high-pressure subsurface environment," says Woskow. "Changes due to intense energy-matter interaction applied to drilling require a new learning curve."
Slovakia-based GA Drilling, meanwhile, is exploring a different high-energy drill technology to bore into the Earth's crust. It is using a pulse plasma drill, based on very short high energy electric discharges that disintegrate rock without causing it to melt. This avoids creating any viscous molten rock, which can be difficult to remove and can stop drill bits penetrating further. "Since the process is very swift with short shocks crumbling the rock, there isn't time for melt to form – so the need to pull up and replace the bit is greatly reduced," says Igor Kocis, chief executive and chairman of GA Drilling. "Five to eight kilometres (3-5 miles) is a target for our current development programme – and later 10km-plus," he adds. "These depths will allow nearly universal access to geothermal power."
Research into pulse plasma drills – using very short energy pulses that disintegrate rock using ionised gas as hot as 6,000C (10,832F) – is another avenue being explored by a European consortium led by the Geothermal Energy and Geofluids (GEG) group, with partners in Germany and Switzerland.
GA Drilling has also been collaborated with Konstantina Vogiatzaki, associate professor of engineering science at the University of Oxford to adapt advanced mathematics looking at how supercritical fluids can be controlled when tapping deep earth energy sources accessed via plasma drilling. "We worked on defining the optimum combustion system for a full-scale drilling tool, opening new horizons in controlling ultra-high pressure combustion through plasma drilling," says Vogiatzaki.
Others are looking beyond our own planet for ways to help us drill down into it. Technology developed for planetary exploration missions on the scorching surface of Venus, where temperatures can reach 475C (887F), are being adopted by geothermal drilling companies. Ozark Integrated Circuits – an electronics manufacturer based in Fayetteville, Arkansas – has been adapting circuits capable of withstanding extreme temperatures that can be used on deep Earth geothermal drilling rigs.
For its own part, the NREL has turned to AI to analyse complex subterranean environments to try to find the best places to drill for supercritical water, as well as helping to predict and detect faults with drills before they cause major issues.
And some companies are already making inroads into the deep Earth. Geothermal company Eavor told the BBC that in 2024 it reached a depth of three miles (5km) with two vertical wells at a site in Gerestried, Bavaria, Germany. It has been using two of the largest land-based drilling rigs in Europe in an effort to create a commercial-scale plant in Geretsried that aims to bring geothermal heat to the surface by circulating water inside a closed loop design it calls the Eavor Loop.
"Our technology is looking to drill up to 11km (6.8 miles) in the future," says geologist and Eavor co-founder Jeanine Vany. "I believe we can make meaningful progress towards unlocking superhot rock in the next three to five years."
Their closed-loop approach also helps to avoid some of the contamination problems that can occur when superheated water is extracted from deep geothermal wells – as the Iceland Deep Drilling Project discovered in 2009. It can also help to reduce the emissions of hazardous gases, such as hydrogen sulphide, that open-loop geothermal systems can emit.
Vany also points out that deep geothermal energy doesn't need a lot of space on the surface, which means it could be slotted into urban locations in the future.
The laptop also integrates the Thomson Computing Wallet alongside an underlying token ecosystem, which will incentivize users through activities such as resource sharing and engaging with decentralized applications.
“The Web3 laptop offers a gateway to a decentralized future, empowering users to truly own their digital experiences. More than just a device, it’s designed to help individuals connect, create, and reclaim control of their digital lives.”
Stephan Français, founder & CEO of METAVISIO
Thomson Computing, the computing branch of Metavasio, has a strong global presence, with its devices sold in over 52 countries and through more than 16,000 physical points of sale.
The launch follows the completion of a eur 50 million (roughly $52.53 million) SAFE financing round in October, secured from American AI and Hardware Investments LTD, to support Metavasio’s expansion into blockchain technology and hardware innovation.
Recent trends in the tech industry have focused on integrating Web3 capabilities into consumer devices. Web3 smartphones, for instance, have previously emerged as an early attempt to integrate blockchain technology into everyday life.
Solana Mobiles, a subsidiary of Solana Labs, developed Saga and Seeker, which are Android smartphones that feature Web3 functionalities and are integrated with the Solana blockchain.
Similarly, HTC introduced the Exodus 1, a blockchain-focused smartphone featuring a built-in cryptocurrency wallet and a secure enclave to protect users’ digital assets.
Indiana Jones demands more than those games and Cyberpunk 2077's path-tracing mode. Unsurprisingly, Machine Games lists the RTX 2060, RX 6600, and Intel Arc A580 as the minimum required GPUs for 60fps gameplay at native 1080p on low settings with path tracing disabled, as the game still uses more modest ray tracing at its lowest settings. However, our recent look back at the 2060 shows that Nvidia's oldest mainstream RTX card can't handle RT in modern titles, even at 1080p.
The game's system memory specs are also alarming. Prior high-end titles have recommended 32GB of RAM for 4K native gameplay, but Indiana Jones also requires it at 1440p.
For example, Machine Games recommends an RTX 4070 to hit 60fps at 1080p using DLSS3 quality mode on low settings with path tracing. This implies a 720p 30fps performance profile with AI making up the difference.
Meanwhile, the spec sheet's ultra path tracing column contains a rare appearance from the RTX 4090. Recent high-end games typically stop at the 4080 for 4K native gameplay, but Indiana Jones requires Nvidia's flagship to reach 4K 60fps, upscaled from 1080p.
An SSD with 120GB of free space is also needed. Interestingly, Machine Games omits the requirements for playing the game with upscaling enabled and path tracing disabled, which could be substantially less demanding.
In late November 2024, xAI’s valuation nearly doubled in six months after closing a $5 billion fundraising round, valuing the startup at around $50 billion.
Recently, Musk hinted at a stand-alone xAI app featuring its large language model (LLM) Grok, which is coming soon. The LLM is currently available on X. A dedicated app for Grok would expand its reach beyond X users.
Musk did not detail the specific legal hurdles that the vehicle may face in the country, though it should be noted that pickup trucks in China face significant regulatory restrictions. This was likely one of the reasons why Tesla called the Cybertruck a “crossover touring vehicle” when it brought the vehicle on a tour around China.
As noted in a report from CNEV Post, the speed limit for pickup trucks on Chinese highways is capped at 100 kilometers per hour, while passenger cars can legally travel up to 120 kilometers per hour. Additionally, pickup trucks have a mandatory scrapping requirement after 15 years of use. Pickups face urban access restrictions in many cities as well, although these have been eased somewhat in recent years.
These restrictions have resulted in China’s pickup truck market being quite small. From January to October this year, China’s pickup truck sales were at 427,000 units, as per data from the China Passenger Car Association (CPCA). In comparison, passenger car sales, which include sedans, SUVs, and MPVs, have totaled 17.84 million units in the same period.
Tesla has been quite successful in China. The Model Y crossover has become particularly popular in the country, with 373,852 units delivered in China from January to October, making up 74.67% of Tesla’s total deliveries. Overall, Tesla delivered 500,685 vehicles in China in the same timeframe, an 8.29% increase year-on-year, as per data compiled by CNEV Post.
A lot has changed since June. First, Elon Musk’s open support of President-elect Donald Trump has boosted his companies, including SpaceX. For example, Tesla shares are up about 40% since November 5.
SpaceX was operating well before Trump’s win, but its most challenging and lucrative projects are still ahead. It is currently developing its Starship rocket to land on the moon and Mars. SpaceX also has its Starlink satellite internet service entering another phase. Starlink’s internet services have proven to be a game-changer.
Now, Starlink will prove its mettle as a cellular provider. Last month, Starlink Direct-to-Cell received a US commercial license from the Federal Communication Commission.
For China, the sanctions pose a significant hurdle in its drive for technological self-sufficiency amid what experts describe as a "Chip Cold War." US Commerce Secretary Gina Raimondo emphasized the gravity of the measures, calling them "the strongest controls ever enacted by the US to degrade the People's Republic of China's ability to make the most advanced chips that they're using in their military modernization."
The latest set of US export controls will add at least 140 Chinese organizations to the "Entity List," which now includes companies like Wingtech Technology and chip investment firm Wise Road Capital. Any US-based companies or international firms seeking to do business with entities on this blacklist must apply for export licenses, which are notoriously difficult to obtain.
For the Biden administration, this marks the third major effort to curb Beijing's AI ambitions, coming just weeks before President-elect Donald Trump is set to be sworn into office. During his previous term, Trump also advocated for stricter export rules targeting companies like Huawei and SMIC while pushing for increased government funding to revitalize US semiconductor manufacturing.
The new regulations expand the scope of export controls to include 24 additional types of chipmaking equipment that were not previously restricted. Moreover, the US will invoke the Foreign Direct Product Rule (FDPR), which affects nearly every non-US company utilizing US-made or US-designed components in their products.
This indicates that the Biden administration may have previously underestimated China's ability to bypass US export controls and is now taking stronger measures to close those loopholes. However, the new rules could also pose challenges for US companies like Applied Materials and Lam Research, making it harder for them to grow their international manufacturing of chipmaking tools.
That said, experts like Greg Allen from the Center for Strategic and International Studies and Dylan Patel of SemiAnalysis remain skeptical that the revised export controls will achieve their intended objectives. They argue that the Biden administration has designed an overly complex set of rules, riddled with potential loopholes and overly dependent on international cooperation to implement multilateral export controls effectively.
Allen and Patel point out that the poor timing of each regulatory iteration has allowed Chinese companies to adapt and stay one step ahead, significantly diminishing the impact of these measures. For instance, a company like Huawei can easily create a network of subsidiaries or source materials and tools from entities not yet listed on the Entity List. By the time sanctions catch up, Huawei and similar companies have often already established fully operational facilities equipped with technology they technically shouldn't have been able to acquire.
In the journal Optical Materials Express, researchers led by Nicolas Laurand at the University of Strathclyde describe their new recycling method and show that quantum dots from supraparticle lasers can be recycled and reused to create new lasers with similar performance.
“Nanoparticle aggregates and supraparticle lasers are expected to play an increasingly prominent role in everything from wearable medical devices to ultrabright LEDs,” said Downie.
“We envision this method being used to extend the life cycle of supraparticles, which could be repurposed for various applications such as medical biosensors, representing a significant advance toward sustainable nanoengineering.”
Recovering costly quantum dots
To create supraparticle lasers, colloidal quantum dots are suspended in an oil-in-water emulsion stabilized with a surfactant. This forms microbubbles in which the colloidal quantum dots naturally aggregate. However, not all batches successfully create supraparticle lasers, and even those that do will degrade with time.
The idea to develop the new recycling method arose during a meeting when members of the research team expressed frustration over the costly loss of nanoparticle quantum dots in faulty supraparticle batches. In an off-the-cuff remark, Downie proposed a potential recycling technique. With Laurand’s support, they tried the method on a defective sample, aiming to recover and reuse the expensive quantum dots.
“Our eureka moment came when we could very clearly see new, albeit crude, supraparticles under the microscope,” said Downie. “Encouraged by this success, we began refining the recovery method for producing and validating the quality of our recycled nanoparticles.”
To recycle the supraparticle lasers, the researchers first had to encourage them to disassemble. They did this by suspending them back into an oil phase, applying moderate heat and subjecting them to mechanical stress in the form of ultrasonic sound waves. Then, they combined the mixture with water to separate the oil with the quantum dots from the water with the impurities.
The quantum dots in the oil phase were then filtered, treated with additional surface coating and tested to see if they could efficiently fluoresce before being reassembled into aggregates for use in a laser. The liquid separation coupled with filtration was carried out in an enclosed separating funnel system, maximizing the sample purity while minimizing solvent use and nanoparticle loss.
Turning recycled nanoparticles into new lasers
Using this recycling process, the researchers demonstrated recovery of 85% of the quantum dots from initial supraparticles. The recycled quantum dots retained a photoluminescence quantum yield of 83 ± 16%, compared to 86 ± 9% for the initial batch. They successfully used the recycled nanoparticles to create lasers that performed similarly to their precursors.
“The discoveries about quantum dot behavior over the past 40 years, culminating in last year’s Nobel Prize in Chemistry, allowed us to understand each stage of the recycling process at a fundamental level,” said Downie. “This enabled us to develop a simple method that is practical even for labs that lack specialized equipment like centrifuges, scrubbers or magnetic field generators.”
The researchers plan to perform additional testing to better understand how nanoparticle performance changes each time they are recycled. The supraparticles tested in this study were unmodified, but the researchers are exploring techniques for recycling functionalized or embedded supraparticles.
About Optica Publishing Group
Optica Publishing Group is a division of the society, Optica, Advancing Optics and Photonics Worldwide. It publishes the largest collection of peer-reviewed and most-cited content in optics and photonics, including 18 prestigious journals, the society’s flagship member magazine, and papers and videos from more than 835 conferences. With over 400,000 journal articles, conference papers and videos to search, discover and access, our publications portfolio represents the full range of research in the field from around the globe.
About Optical Materials Express
Optical Materials Express is an open-access journal focusing on the synthesis, processing and characterization of materials for applications in optics and photonics. It is published by Optica Publishing Group and emphasizes advances in novel optical materials, their properties, modeling, synthesis and fabrication techniques; how such materials contribute to novel optical behavior; and how they enable new or improved optical devices. The Editor-in-Chief is Andrea Alù from City University of New York, USA. For more information, visit Optical Materials Express.
By mid-2025, Nova will introduce its most ambitious capability yet: an “any-to-any” model that can transform text, speech, images, and video seamlessly. This vision, as Jassy described it, is “the future of how frontier models are going to be built and consumed.”
Musixmatch is already integrating Nova Reel to help emerging artists create music videos, while Dentsu Digital is using it to accelerate campaign production. Nova’s speech-to-speech model, expected in early 2025, could further transform industries like customer service and media.
Internal Potential and Strategic Branding
While Nova is marketed as an enterprise tool via Amazon Bedrock, its internal applications could prove equally transformative. With Amazon’s ownership of MGM Studios, Nova Reel could streamline video production, from trailers to visual effects, while Amazon’s gaming division might use multimodal models to enhance storytelling and design in its titles.
The branding as Amazon Nova rather than AWS Nova signals something larger. It suggests the company sees Nova not just as a cloud service but as a central piece of Amazon’s strategy, integrating AI into its many verticals, from retail to logistics.
An Emphasis on Accessibility and Responsibility
Jassy also highlighted Nova’s affordability, claiming the models are up to 75% cheaper than competitors. This accessibility, combined with support for over 200 languages, positions Nova as a tool that could drive AI adoption globally, including in markets like South Africa, where cost barriers have slowed adoption.
However, Jassy didn’t shy away from addressing AI’s risks. Nova includes built-in content moderation and watermarking, features aimed at ensuring responsible use. “Many human beings don’t go to one human being for expertise in every single area,” Jassy remarked, underscoring the need for flexibility and the human-like adaptability of AI.
Amazon’s Strategic Bet on the AI Frontier
Amazon Nova is more than just a response to a shifting market, it’s a calculated manoeuvre in the evolving cloud market. AWS, once the undisputed leader in cloud computing, now faces stiff competition in an era increasingly defined by AI-driven innovation. Companies like Microsoft, with its OpenAI partnership, and Google, with its Gemini models, have been gaining ground rapidly, reshaping what it means to lead in this space.
By branding Nova as Frontier Models, rather than the more common Foundation Models, Andy Jassy signalled Amazon’s intention to position itself at the forefront of this next wave of AI. It’s a barely veiled claim that Nova represents not just the present but the future of AI development, a future where cloud dominance will be defined by intelligence, not just infrastructure.
As Jassy concluded his keynote with his signature “Giddy up,” he left little doubt about Amazon’s ambitions. With Nova’s integration across Amazon Bedrock, the company is making its case to restake its claim as the world’s leading cloud provider. These models aren’t just for tech demonstrations — they are designed to address the demands of an undefined but fast-emerging technological frontier. From enterprise workflows to creative industries, Nova isn’t merely catching up — it’s Amazon’s bid to lead the next chapter of cloud and AI innovation.
It is not yet clear who has been eliminated from the Soto sweepstakes.
The Post’s Joel Sherman reported Tuesday that the expectation is Soto’s contract will exceed $600 million.
Shohei Ohtani’s monster 10-year, $700 million contract he signed with the Dodgers last offseason is the current record, with $680 million being deferred.
Soto turned down a 15-year, $440 million deal from the Nationals in 2022 and was later traded to the Padres. He was acquired by the Yankees last December in a blockbuster trade with San Diego.
Managing the end-to-end software life cycle
Sivasubramanian told me that he believes what differentiates Q Developer from competing platforms is its focus on the entire software development life cycle. So far that meant helping developers troubleshoot issues and perform multistep tasks to fix them (or build entirely new apps), as well as scan the code for security vulnerabilities.
At re:Invent, the company is taking this a step further. Q can now, for example, automatically generate unit tests. But what’s maybe even more important is that it can now do the one thing that many developers hate the most: write and maintain the documentation for that code. To complete this cycle, Q can now generate a first code review when developers check in their code.
“In Amazon, we have this rule that no code ever gets checked in without a code review,” Sivasubramanian said. “So if you don’t do a code review, then you cannot check in code. But not many enterprises actually have either enough senior engineers to review or the senior engineer says: ‘I can’t deal with so many reviews. Can somebody first review it before we do so?’ Q will streamline the code review process by being the first line of reviewer and takes care of automatically checking code quality, security vulnerabilities, and so forth.”
Amazon
Then, once the code is in production, a new operations agent for Q can now automatically pull in data from AWS CloudWatch, the company’s monitoring service, and immediately start investigating when an alarm goes off. “It utilizes the [knowledge it has about an] organization’s AWS resources and then it sifts through hundreds of data points across various resources sitting in CloudWatch. Then, after analyzing it, Q comes up with potential hypothesis for the root cause and then it guides the users through how to fix it,” Sivasubramanian explained.
All you wanted for Christmas was help with your Cobol and .NET migrations, right?
For those enterprises with older codebases, transitioning to the cloud often involves rewriting a lot of their existing code. One of the earliest differentiating features of Amazon Q Developer was its agent for code transformation. At the time, the focus of this agent was on modernizing older Java apps. Today, the team is expanding this by also helping developers update their older .NET-based applications from Windows to Linux.
And while this may at first seem like a curiosity, AWS is also launching an agent for modernizing COBOL mainframe applications. A lot of large enterprises still rely on this old code, after all, which few developers know to work with today. These are very complex migrations, Sivasubramanian stressed, and so the goal here is not to simply translate the existing code 1:1.
“Our goal is not to actually just like fully COBOL project in, code out,” he said. “The reality is, these projects are inherently extremely complex. You need to have a human in the loop to leverage it, but I’ve heard customers say, ‘Hey, this takes multiple years and customers have explicitly told us this is a game changer and would significantly drop that timeline.”
Sivasubramanian noted that while there is less COBOL code to train models to automate the code migration, the team was able to leverage AWS’ overall experience in modernizing mainframe applications, as well as more traditional methods for code translation.
“Taking code from one language to another one arguably is the easy part,” he said. “But the harder part is: how do you know you got it right? And how do you even know what the code does? And then the challenge in these [codebases] is they are usually poorly documented and dependencies are not well understood. So what we have built is really extremely innovative, and [the system] also understands, at a project level, what are the objectives of each of the module, and then plans out and creates a migration planning timeline to actually generate the code, and then generate the test — and bringing humans in the loop to see how you validate it.”
However, several factors warrant consideration when interpreting these metrics:
YouTube views alone do not guarantee sustained retail participation or predict market direction.
Previous cycles have shown that social media engagement can be a lagging indicator of market interest.
The current view count of 4.72 million represents about 50% of the all-time high, suggesting retail participation may still be in the early stages compared to previous cycles.
This trend emerges as part of a broader shift in market dynamics. Traditional finance continues to establish stronger footholds in the crypto space through ETFs and other investment vehicles, attracting a more diverse group of retail users to the crypto space.
We all suffer from stress sometimes; it’s just human nature and affects us all differently. You could be worried about work, money, or just everyday stress. One key area where it can affect you is in the bedroom, so looking at ways to manage stress is important to relationships.
What Is Stress?
Before we discuss how it impacts sexual health, let’s take a look at what it is and the different types that are out there. Sometimes, some stress is good for us as we all need it occasionally, but too much of it is bad.
What Is Stress?
Before we discuss how it impacts sexual health, let’s take a look at what it is and the different types that are out there. Sometimes, some stress is good for us as we all need it occasionally, but too much of it is bad.
Physiological Effects
When you are stressed, you release hormones such as cortisol and adrenaline. These can be good in short bursts, but chronic stress can keep these high, which isn’t good for the body; these can lead to the following symptoms.
Reduced Libido – If you have high levels of cortisol, this can overpower the body’s natural sex hormones and reduce your sex drive.
Erectile Dysfunction – Stress does not help with performance anxiety, which might mean you struggle to keep an erection.
Menstrual Cycle – For ladies, stress might mess up their menstrual cycle, which makes them less up for sex.
Psychological Effects
Stress can massively psychologically impact you as well. This can lead to some of the following issues.
Body image problems – you might see yourself as not sexually attractive, and therefore, you won’t be as up for sex as if you thought you were attractive.
Anxiety and depression – Chronic stress doesn’t help with anxiety and depression, which in turn will reduce your sexual desire.
Ideas For Managing Stress To Improve Your Sex Life
There are plenty of ways you can look at to help lower your stress levels. Not all of these will work for you, so it’s important to investigate them to see which works for you.
Exercise
We hear a lot in the press about how exercise can help lower stress. This is because it releases endorphins, which our body naturally creates. This lowers stress levels, and a workout can only improve your health, which leads to our body functioning better sexually. You could even look at working out together as a couple. This would allow you to spend time together, which will also improve your relationship.
Talk To your Partner
Talking is essential. They might not realize you are quite as stressed as you are, and opening up to them might help them. They say a problem shared is a problem halved. Partners also love people being honest, and it shows that you can trust each other.
Think about when the right time to bring it up is, and make sure you are 100% honest and open. Let them know about your concerns, fears, and desires. It’s also important you let them speak so you can hear their side of the story, as it should be a two-way thing.
It could also be that you need counseling not just on your own but as a couple. This could be hard to bring up, but it is about being open and honest.
Diet
We all go through bad times when it comes to diet. The important thing is not to let this become too much of a regular thing. Eating takeaways and sugary food every day is not good for you. Evaluate your diet and make sure you are taking in all the key nutrients that your body needs. Diet can massively help with stress levels.
Joint Relaxation Techniques
In addition to exercising together, joint relaxation techniques could be a good idea. This could include things like meditation or yoga. Doing these together can help increase your bond.
Make Sex Fun
This might sound obvious, but maybe making sex fun again might be a good idea to try. If you are stressed and have been having sex on the same day, in the same positions, and just going through the motions, then it could be time to mix it up.
Why not do some sex shopping together and have a look at what you could buy? Toys like vibrators and dildos could add some extra stimulation. Or you could look at role play or some sexy outfits that you could both try on. Watching porn together can also be a good way to make sex more fun.
What Else Could Be Causing It?
It could be worth booking an appointment with your general doctor to discuss your issues. You might be aware of other underlying issues with your body or mind, such as diabetes or fibromyalgia, which they can help diagnose.
Final Thoughts
You are not alone. Not everyone goes through life having the perfect sex life. Everyone has their ups and downs. What is important is that if you have more downs than ups, you get some help, which can’t be done by shutting yourself off.
Using SageMaker Unified Studio, customers can publish and share data, models, apps, and other artifacts with members of their team or broader org. The service exposes data security controls and adjustable permissions, as well as integrations with AWS’ Bedrock model development platform.
AI is built into SageMaker Unified Studio — to be specific, Q Developer, Amazon’s coding chatbot. In SageMaker Unified Studio, Q Developer can answer questions like “What data should I use to get a better idea of product sales?” or “Generate SQL to calculate total revenue by product category.”
Explained AWS in a blog post: “Q Developer [can] support development tasks such as data discovery, coding, SQL generation, and data integration” in SageMaker Unified Studio.
Beyond SageMaker Unified Studio, AWS launched two small additions to its SageMaker product family: SageMaker Catalog and SageMaker Lakehouse.
SageMaker Catalog lets admins define and implement access policies for AI apps, models, tools, and data in SageMaker using a single permission model with granular controls. Meanwhile, SageMaker Lakehouse provides connections from SageMaker and other tools to data stored in AWS data lakes, data warehouses, and enterprise apps.
AWS says that SageMaker Lakehouse works with any tools compatible with Apache Iceberg standards — Apache Iceberg being the open source format for large analytic tables. Admins can apply access controls across data in all the analytics and AI tools SageMaker Lakehouse touches, if they wish.
In a somewhat related development, SageMaker should now work better with software-as-a-service applications, thanks to new integrations. SageMaker customers can access data from apps like Zendesk and SAP without having to extract, transform, and load that data first.
“Customers may have data spread across multiple data lakes, as well as a data warehouse, and would benefit from a simple way to unify all of this data,” AWS wrote. “Now, customers can use their preferred analytics and machine learning tools on their data, no matter how and where it is physically stored, to support use cases including SQL analytics, ad-hoc querying, data science, machine learning, and generative AI.”
I’m always looking for new ways to help adult content creators expand their brands and reach new audiences so, ultimately, they can make more money. Recently I found something new to help you do just that.
As an adult content creator, navigating the often-restrictive landscape of social media can be a challenge. Shadowbans, account suspensions, and content restrictions can hinder your ability to reach your audience and build your brand.
Quantum computing overcomes classical computing’s limitations
Traditional robots rely on classical computing, which utilizes bits to represent information as 0s or 1s. This binary system, while effective for many tasks, struggles to handle the increasing complexity and data volume characteristic of modern robotic applications.
Quantum computing, in contrast, employs qubits, which can exist in a superposition of states, representing 0, 1, or both simultaneously. This property allows quantum computers to process and analyze information at speeds and scales that are theoretically unattainable for classical computers.
The study outlines two main research areas within quantum robotics. The first focuses on making current robot tasks better. For example, quantum algorithms have shown promise in improving how robots navigate, allowing them to move through complex environments with greater efficiency and accuracy.
These algorithms can also improve decision-making, letting robots adapt to unexpected situations and make the best choices in real-time. Furthermore, quantum algorithms can help multiple robots work together better, leading to more efficient teamwork in difficult tasks.
The second research area looks at how to include quantum mechanics in the physical design and function of robotic systems. This includes research into robots that communicate through quantum entanglement, a phenomenon that allows instant information transfer between linked particles no matter how far apart they are.
Such communication could lead to very secure and efficient robot networks that can coordinate their actions with incredible speed and precision. Additionally, researchers are looking at using quantum-controlled designs to improve how well robots perform
A typical quantum robot is imagined to have a mix of quantum computing modules for complex information processing, regular computer systems for handling everyday tasks and connecting with existing technologies, and ways to communicate with outside systems and other robots. These parts work together to improve the robot’s thinking, sensing, and acting abilities, allowing it to do things that are currently impossible for regular robots.
Challenges and future of quantum robotics
However, the researchers have also pointed out certain challenges that can hinder the development of quantum robotics. These challenges include hardware limitations and the early development stage of quantum computing software.
While building fully working and adaptable quantum robots faces challenges, including creating stable and reliable quantum processors and dealing with environmental effects on quantum states, researchers are hopeful about the future of this field.
Amount of energy produced depends on levels of shade
The company maintains that the amount of energy produced depends on levels of shade, intensity of the sun, and geographical location.
Mercedes-Benz drivers in Stuttgart, Germany, drive an average of 32 miles per day. Around 62 percent of this distance would be covered using solar energy. In Los Angeles, there is even a surplus of solar energy. It could be used for 100 percent of their driving, on average, and the surplus of energy could be fed directly into the home network via bidirectional charging, according to Mercedes.
This highly efficient solar paint contains no rare earth elements or silicon, relying only on non-toxic, readily available raw materials. It is easy to recycle and considerably cheaper to produce than conventional solar modules. The Mercedes-Benz research department is currently working to enable the use of the new solar paint on all exterior vehicle surfaces – regardless of shape or angle, as mentioned in a press release.
“Innovative strength was and is among the most important driving forces of Mercedes-Benz. Our brand has been shaping automotive progress for almost 140 years with pioneering technologies,” said Markus Schäfer, Member of the Board of Management of Mercedes-Benz Group AG.
“For us, innovation only makes sense if it offers our customers real added value. Their perfect composition is what makes a Mercedes a Mercedes. That’s why we believe progress means developing innovative technologies in dialogue with society. It’s how we maintain our innovative power and create a new kind of automotive experience.”
Over time, CUE’s abilities expanded significantly. It progressed from making simple free throws to receiving the ball, positioning itself, and sinking shots from the three-point line and beyond.
Later iterations allowed CUE to independently pick up the ball, shoot from varying positions, move on wheels, and even dribble. Each upgrade showcased Toyota’s commitment to pushing robotics’ limits, making CUE a groundbreaking blend of technology, precision, and adaptability.
CUE’s shot evolution
For the record-breaking attempt, CUE6 underwent significant enhancements, such as hand modifications and the addition of foot cameras to detect ball movement better.
By trying a record-breaking shot from an unprecedented distance, the team hoped to push limits. GWR claims that the robot made use of sophisticated AI; CUE6 examined its structure and refined its throwing technique to produce the best form for long-range shots.
After several tries and a great deal of practice, the longest basketball shot ever made by a person is 34.6 meters (113 feet 6 inches). The creators of CUE6 were optimistic that a flawless first try would allow it to surpass expectations. The second effort was successful, landing the historic shot from 24.55 meters (80 feet 6 inches), while the first shot missed by a close margin and bounced off the rim.
Uses multiple nozzles
Traditional 3D printing builds tissues layer by layer on a bio-compatible substrate. In bioprinting, the “ink” is a living material composed of cells and often cells suspended in a hydrogel.
But bioprinting faces challenges. One major hurdle is achieving the high cell density found in natural tissues. This is crucial for the tissue to function properly.
Spheroids — with their human cell-like density — offer a promising alternative for tissue bioprinting.
While 3D printing spheroids is a promising approach to creating tissue with human-like cell density, current techniques have limitations. Existing methods can damage cells or are too slow and cumbersome to be practical.
Researchers have developed this HITS-Bio to address these issues. HITS-Bio utilizes a “digitally controlled nozzle array” that consists of multiple nozzles capable of three-dimensional movement.
Enables faster recovery in rats
To test HITS-Bio, the team fabricated a one-cubic centimeter cartilage tissue using 600 spheroids in less than 40 minutes.
This quick fabrication outperforms existing bioprinting techniques.
The team went on to further test this novel bioprinting technique in rat models. The technique was used to repair bone tissue in a rat model.
The researchers demonstrated the clinical potential of HITS-Bio by directly printing spheroids into a skull wound in a rat during surgery.
For this surgery, these spheroids were specifically programmed with microRNA to differentiate into bone cells. This accelerated the healing process.
“Since we delivered the cells in high dosages with this technique, it actually sped up the bone repair,” Ozbolat said in the press release.
The wound exhibited a 91 percent healing rate after three weeks and a 96 percent healing rate after six weeks.
This is not the first time we’ve been warned that humans cannot find AI work on their own. Last year, a study from the University of South Florida concluded that linguists could not tell the difference between text created by AI and text written by humans.
A different study last year – this one from American universities in Vietnam – found that AI detectors were far better at picking out AI text than human teachers were. The team in Vietnam wrote, “Although [AI detection system] Turnitin correctly identified 91% of the papers as containing AI-generated content, faculty members formally reported only 54.5% of the papers as potential cases of academic misconduct.”
The paper, by Peter Scarfe and others at the University of Reading in the U.K., examined what would happen when researchers created fake student profiles and submitted the most basic AI-generated work for those fake students without teachers knowing. The research team found that, “Overall, AI submissions verged on being undetectable, with 94% not being detected. If we adopt a stricter criterion for “detection” with a need for the flag to mention AI specifically, 97% of AI submissions were undetected.”
You read that right – 97% of AI work in university courses was not flagged as possible AI by teachers. But it’s actually worse than that, as the report also says, “Overall, our 6% detection rate likely overestimates our ability to detect real-world use of AI to cheat in exams.”
In that study, the teachers were told in advance that papers using AI would be submitted in their courses and they only found little more than half of them. That study also used “prompt engineering” to make the papers more difficult for AI detectors to spot. Still, the machines caught 91%. The humans, 55%. And schools are actually turning these machines off.
Again, the humans were told to look for them and still whiffed. When humans are not put on alert and AI detectors are not used, as was the case in the more recent U.K. study, AI work is missed nearly every time.
Worse, the U.K. study also found that, on average, the work created by AI was scored better than actual human work. “We found that in 83.4% of instances the grades achieved by AI submissions were higher than a random selection of the same number of student submissions,” the repot said.
In other words, a student using the most basic AI prompt with no editing or revision at all, was 83% likely to outscore a student peer who actually did the work – all while having a generous 6% chance of being flagged if the teachers did not use any AI detection software. Keep in mind that, in real classrooms, a flag of suspected AI use doesn’t mean much since professors are very reluctant to pursue cases of academic integrity, and even when they do, the schools are often even more reluctant to issue any sanctions.
Recently, the BBC covered the case of a university student who was caught using AI on an academic essay, caught by an AI detector by the way. The student admitted to using AI in violation of class and school rules. But, the BBC reported, “She was cleared as a panel ruled there wasn’t enough evidence against her, despite her having admitted using AI.”
Right now, if a school or teacher is not getting assistance from AI detection technology, using AI to cheat is very likely to boost your grade while carrying nearly zero chance anything bad will happen. In other words, if your school is not using AI detection systems, you’d have to be an idiot to not use AI on your coursework and exams, especially if your classmates are. Spoiler alert – they are.
Not to be overlooked in the results of the new U.K. study is that the fake students were in online classes and their fake coursework was submitted online – where teachers could not possibly know anything about their students, including whether they were even actually human. It highlights the reality that online courses are more vulnerable to cheating, and AI cheating in particular, because teachers do not know their students and cannot observe their work.
Naturally, schools could use technology to solve this problem too, either by proctoring assignments and exams, or by using writing environments that track changes and revisions. But just like using AI detectors in the first place, many schools don’t want to do those things either because they take work and cost money.
That said, there’s been chatter in recent weeks that the Red Sox are making a serious push. Meanwhile, SNY’s Andy Martino and Jim Duquette of MLB Network Radio on SiriusXM (X link) each suggested last night that the Blue Jays might come in with the highest bid at the end of the day. Reporting has generally cast the Dodgers as something of a long shot, but it’s impossible to count them out on any top-tier free agent.
In any case, Soto seems mere days away from establishing a record for the top net present value in MLB history. That’s currently held by Shohei Ohtani, whose deal is valued by MLB just shy of $461MM after accounting for deferrals. There was never much doubt that Soto would beat that, though it was fair to wonder whether the bidding would stop somewhere between $500MM and $600MM.
"While the Biden-Harris administration may not want to inform the American people of just how closely they are collaborating with foreign governments, at the very least, the American people have a right to know what foreign actors are trying to impose these radical regulations on American companies," Cruz wrote.
The European Union has led the global charge for regulations tailored to AI technology, having this year passed the AI Act, the world's first sweeping set of laws governing the technology.
Cruz cited a memorandum of understanding the U.S. issued alongside Britain in April, as well as a treaty signed by the Biden administration and several European governments, as examples of foreign influence over domestic AI policy.
Cruz also said the Centre for the Governance of Artificial Intelligence (GovAI), a UK-based non-profit organization which recently co-hosted a summit for AI policy experts in San Francisco, had not followed the law by failing to register as a foreign agent despite engaging in political activities.
"GovAI is aware of the request made by Sen. Cruz to the Department of Justice, and, if required, will cooperate fully with the relevant authorities," a GovAI spokesperson said.
"GovAI is a strictly independent organization which is committed to operating responsibly and transparently."
Elon Musk and Vivek Ramaswamy are apparently floating the idea of doing away with changing the clocks twice a year and staying on daylight savings time permanently, The Washington Post reported Tuesday.
In a follow-up post, Musk told Sen. Mike Lee, R-Utah, he did indeed want to end the time changes.
Advocates for year-round daylight saving time say permanently advancing the clocks one hour would permit more people to enjoy sunshine and avoid the frustrations involved with resetting clocks.
But politicians in the center of the country have pointed out that such a policy would mean winter sunrises that could happen past 9 a.m. in cities such as Indianapolis and Detroit.
Complaints about clock changes are not new, and lawmakers in the early 1970s even moved to permanently adopt daylight saving time.
However, that decision backfired almost immediately, as there was nationwide criticism of the move, because children were waiting in the dark for school buses to arrive.
Congress ended up rolling back the change after 10 months.
Brian Freeman, a Newsmax writer based in Israel, has more than three decades writing and editing about culture and politics for newspapers, online and television.
As with many subjects, Adams' relationship to crypto is not without controversy. A year after taking a chunk of his salary home in BTC and ETH, the mayor landed in hot water for failing to disclose his crypto holdings in a mandatory report filed with the city’s Conflict of Interest Board.
Since then, Adams’ legal troubles have increased dramatically. In September, the politician became the first mayor in New York City’s history to be indicted on criminal charges while in office.
The U.S. Department of Justice (DOJ) has accused him of receiving luxury travel accommodations and other perks from Turkish government officials and businessmen in exchange for favorable treatment of the country’s interests in New York.
The NAV per share chart for GSOL mirrors Solana’s price trajectory, highlighting growing investor confidence in the asset. The upward trend in NAV signifies a maturing market with rising demand for professionally managed investment products tied to Solana.
Industry-wide context: Solana ETF momentum
Grayscale’s filing follows a slew of applications for spot Solana ETFs in 2024.
VanEck and 21Shares submitted proposals last summer, targeting the burgeoning Solana ecosystem. Canary Capital joined the race with its application in October, while Bitwise Asset Management recently entered the fray.
What Is Quantum Robotics?
Quantum robotics uses quantum computing principles — such as superposition, entanglement, and quantum algorithms — to tackle challenges that traditional robots face. These include processing vast sensory data, meeting real-time response needs and enabling cognitive and emotional functions that mimic human intelligence. Unlike classical systems, quantum robots — or “qubots” — leverage the probabilistic nature of quantum mechanics to expand the scope and capability of robotics.
Traditional robots depend on binary computing frameworks, which struggle to handle the increasing complexity of modern demands. Quantum computing offers an alternative, with the ability to process and analyze data at speeds and scales — theoretically at least now — that are unattainable by classical systems.
Methods and Approaches
The study identifies two key areas of research in quantum robotics. The first focuses on enhancing existing robotic tasks. Quantum algorithms have been applied to improve navigation, decision-making and multi-robot coordination. For instance, quantum reinforcement learning allows mobile robots to detect faint signals and execute strategies with greater efficiency, according to the researchers.
The second area involves integrating quantum mechanics into robotic systems. Early models, such as Paul Benioff’s concept of quantum robots cited by the team, envision mobile units equipped with quantum processors capable of interacting with their environments. More recent efforts explore robots communicating via quantum entanglement or using quantum-controlled designs to improve performance and adaptability.
pen Questions And Current Challenges
Despite its promise, quantum robotics is still in its early stages and faces significant technical and practical challenges, according to the researchers. Hardware limitations remain a significant obstacle, as fully functional and scalable quantum processors are still under development. Quantum states are also highly sensitive to environmental disturbances, creating difficulties in maintaining stability during operations.
Bridging quantum and classical systems poses another challenge. Efficient converters to translate information between these two computing paradigms are essential but remain underdeveloped. The researchers add that the field suffers from a lack of robust quantum-specific software, which slows the deployment of advanced quantum robotics.
One of the worst kept secrest in the influencer world is that it’s a pay to play game. For a few thousand dollars you too can be on the cover of Playboy magazine. Okay sure it’s Playboy Africa or something like that, but still, you technically are on the cover of a real Playboy magazine. I think Maxim and FHM have pay-to-play deals with photographers as well. These are, however, real magazines with REAL followers and really established brands.
Some people have used this information to scam models, influencers, and adult performers into paying to be on the cover of a fake magazine. This particular group of con men is out of India, and in May of this year, they created a website called Billboard Lifestyle, which is meant to mimic the website owned by Billboard magazine, which does, in fact, have a lifestyle section.
The real Billboard magazine website is very well known. It’s like Playboy but for Musicians. Being featured in Billboard magazine is a really big deal for those in the music industry.
Their website (Billboard.com) gets over 20 million visitors a month. The company that owns them is called Pekse Media Corp, and they own a ton of other really big brands like Variety, Deadline, Rolling Stone, and so many more. Let’s just say these guys are a big company.
In May of this year, a group of conmen out of India set up their own website called Billboard Lifestyle to mimic the original website. They made a ton of posts and then went out to con girls to pay them to get a cover of Billboard Lifestyle magazine.
I’m told they are charging anywhere from $5,000 to $9,000, and so far, they have tricked quite a few girls into doing this. These rates are on par with what it takes to get a cover for other magazines like International Playboy’s so I get why some might not question it.
But they should have. Because, in the end, they are paying for nothing. They are not getting exposure for being in a real magazine with any real subscribers or readers, nor are they featured on a website with any traffic.
Here are some of the fake Billboard Lifestyle covers.
et’s look at one of the girls who fell for it. She’s an up-and-coming iG influencer named Ariana Amadeus, who, as you can see by her post, actually tags Billboard (the real company), not the con men. Then, it provides the link to a fake website. It couldn’t be more obvious that this girl thought she was being featured in a story by the real Billboard.
Their website is nothing more than an attempt to sell the scam. Doing a deep dive, we see that their website has only 2300 (estimated) visitors per month, which includes the guys running the site and making fake posts. They get less than 100 visitors a day.
And as you can see, most of that is from India. So it’s all fake “bot” traffic.
The Billboard Lifestyle website created its first post on May 4, 2024. By June, it had already recruited several influencers and aspiring models.
This is a copy of their first post.
This was created just a few months ago. Now, they have posted thousands of seemingly real posts, but they are all just a way to mask the fact that their website is a scam.
There is no such thing as Billboard Lifestyle magazine.
As Andreessen discussed, one of the reasons federal officials use to justify this pressure is that crypto companies do not comply with federal anti-money laundering laws, the ones imposed by the Bank Secrecy Act of 1970. The BSA requires, among other things, banks to collect a great deal of personal information on their customers. (There are various layers involved, and they’re often grouped under the heading of “Know Your Customer” rules).
And if a bank regulator thinks a bank is failing to comply with those laws, even by association with certain customers who are guilty of no crime, that’s a problem the bank can’t have.
Bigger Than the Bank Secrecy Act
But the source of this debanking problem goes well beyond the BSA/KYC/AML rules. In fact, federal regulators justified the original Operation Choke Point based mainly on protecting banks’ safety and soundness and reputation risk.
For those who don’t recall, Operation Choke Point was launched during the Obama administration in 2013. The name comes from federal regulators’ efforts to pressure banks to “choke off” access to banking services by merchants and industries the administration didn’t like. Specifically, regulators pressured banks to stop doing business with several types of companies, including payday lenders, small dollar loan companies, and ammunition and gun sellers.
Operation Choke Point Predates 2013
The federal regulatory framework gives regulators enormous discretion. In fact, they have so much discretion that they have the authority to warn banks about dealing with certain types of customers for almost any reason they choose to justify. And any kind of warning from the FDIC, for example, is enough to get someone debanked.
Operation Choke Point may have been formalized in 2013, but regulators have been pressuring banks for decades. Both the FDIC and the Office of the Comptroller had forced banks to close customer accounts based on concerns over reputational risks dating to at least 2003, long before Operation Choke Point.
Reputational risk is vague enough, but there’s more. The FDIC can terminate a bank’s deposit insurance if it finds that the bank is engaging in “unsafe or unsound practices.” And if the FDIC hints to a banker that engaging with certain customers amounts to unsafe or unsound practices, guess what’s going to happen?
The 2010 Dodd-Frank Act also gave the Fed the authority to impose special regulations “to prevent or mitigate risks to the financial stability of the United States that could arise from the material financial distress or failure, or ongoing activities, of large, interconnected financial institutions.” The law does not, however, define financial stability or distress. That’s up to the regulators. And these are just three examples—there are many more.
This broad discretion has stark implications for future debanking efforts, and some activists are already pressuring the Fed and the FDIC to use their authority to freeze oil companies out of the banking sector in the name of stopping climate change. The only surprising fact is that we’ve only reached two versions of Operation Choke Point.
Bipartisan Solutions are Needed
Many people are hopeful that the Trump administration will end Operation Choke Point 2.0 and regulators will end their outward hostility to crypto firms. But history has already proven that the problem is much bigger than any one industry and hoping for the “right” administration to be in power is far from a solution.
Rep. Ritchie Torres (D-NY) is right—there is essentially “no real limitation on the ability of banking regulators to de-bank law-abiding citizens and businesses without due process of law.” Regulators’ broad discretion is an “insidious threat to civil liberties,” and it is one that “transcends partisanship.”
The only way to stop this kind of debanking in the future is for Congress to pare back the discretion it has given to the regulatory agencies. And neither party can do that on its own.
China’s LiDAR sensors could have hidden vulnerabilities
FDD claims that these custom chips can be specifically engineered to include hidden vulnerabilities, known as “hardware trojans,” which could provide a LiDAR company or other hostile actor with unauthorized access or control over LiDAR devices.
The report also maintained that LiDAR systems frequently receive software updates over the internet, often affecting network settings, cybersecurity protocols, and core sensing algorithms. These over-the-internet updates, especially those originating in foreign countries of concern, could introduce additional cybersecurity vulnerabilities, making Chinese-made LiDAR systems highly vulnerable to manipulation.
LiDAR, which stands for Light Detection and Ranging, is a remote sensing method that uses light in the form of a pulsed laser to measure ranges (variable distances) to the Earth. These light pulses—combined with other data recorded by the airborne system — generate precise, three-dimensional information about the shape of the Earth and its surface characteristics.
The think tank maintains that the threat of attacks that cause LiDAR sensors to transmit false information is equally severe. Compromised sensor activity could erode policymaker confidence in the reliability of LiDAR-dependent systems, including advanced weapons platforms. This potential loss of faith could undermine strategic decision-making and policymaker resolve to deploy these systems during future conflicts, according to FDD.
“During wartime, LiDAR-enabled weapons could fail when needed most. Similarly, sensor malfunctions could sow doubt about the integrity of LiDAR-dependent critical infrastructure systems,” said the report.
“This extends to systems needed to support future force mobilization, such as domestic rail networks, potentially jeopardizing U.S. military readiness and responsiveness.”
Budget constraints
An earlier Congressional Budget Office assessment had stated that hypersonic weapons might cost one-third more than ballistic missiles of the same range with maneuverable warheads.
It had also pointed out that they probably would not be more survivable than ballistic missiles – unless the latter encounters a highly-effective long-range air defense system.
Moreover, in FY 2025 the DOD had requested $182.3 million for the hypersonic defense program element and $6.9 billion for offensive hypersonic weapons program.
The report states that due to the lack of defined mission requirements for weapon development, it may be challenging for Congress to evaluate the balance of funding for hypersonic weapons programs, enabling technologies, supporting test infrastructure, and hypersonic missile defense.
US and other countries with hypersonic weapon capability
Several countries are developing or have already developed weapons, which fly at speeds of at least Mach 5 (five times the speed of sound).
There are two primary categories of these weapons:
• Hypersonic glide vehicles are launched from a rocket before gliding to a target.7
• Hypersonic cruise missiles are powered by high-speed, air-breathing engines, or
“scramjets,” after acquiring their target.
The DOD is currently developing hypersonic weapons under the Navy’s CPS, and HALO program, as well as US Army’s LRHW, and HACM missions. These are in the process alongside several other programs being run by DARPA.
Its backers state that the weapons would act as deterrence and allow the US to win the wars of the future.
The report states that Russia is pursuing two hypersonic weapons programs—the Avangard and the 3M22 Tsirkon (or Zircon)—and has reportedly fielded the Kinzhal (“Dagger”), a maneuvering air-launched ballistic missile.
The Avangard, as per Russian news sources, entered into combat duty in December 2019, and it reportedly carries a nuclear warhead.
The Zircon is a ship-launched hypersonic cruise missile capable of hitting both land and sea-based targets. It had been first launched into Ukraine in February 2024.
Similarly, the air-launched Kinzhal missile has also been used by Russia for attacking Ukraine.
China, on the other hand, has fielded the DF-ZF hypersonic glide vehicle (previously referred to as the WU-14) in 2020, alongside the DF-17 missile. It has also successfully tested Starry Sky-2 (or Xing Kong-2), a nuclear-capable hypersonic vehicle prototype, in August 2018.
Some reports mention that the Starry Sky-2 could be operational by 2025. Moreover, China is devoted to weapon development in this category and is carrying out several tests at wind tunnels and on ground to develop other systems.
Other countries like India, South Korea, Australia, France, are also involved in experimenting with hypersonic weapon development programs and may soon be able to field them.
Prosegur’s decision to set up its third cryptocurrency vault in Argentina aligns with the country’s leading position in Latin America in terms of crypto transaction volume, as reported by Chainalysis from June 2023 to June 2024. Continuing with Argentine reports in CNF, Argentina ranked first in the region, with 61.8% of transactions involving stablecoins. Internationally, the country is 15th in cryptocurrency adoption.
The introduction of a military-grade technology facility for cryptocurrency custody in Argentina underscores the increasing importance of advanced security solutions in the management of digital assets, reflecting a significant step toward aligning with global standards in the digital financial sector.
Isai Alexei
As a content creator, Isai Alexei holds a degree in Marketing, providing a solid foundation for the exploration of technology and finance. Isai's journey into the crypto space began during academic years, where the transformative potential of blockchain technology was initially grasped. Intrigued, Isai delved deeper, ultimately making the inaugural cryptocurrency investment in Bitcoin. Witnessing the evolution of the crypto landscape has been both exciting and educational. Ethereum, with its smart contract capabilities, stands out as Isai's favorite, reflecting a genuine enthusiasm for cutting-edge web3 technologies. Business Email: info@crypto-news-flash.com Phone: +49 160 92211628
68% of women never EVER share their core level fantasies with their significant other out of a legitimate fear of rejection. Conversely, 32% will share it if she trusts her man and feels secure that he won't judge her and leave her. I was with one woman for five years and her fantasy was that she wanted to be dominated and punished / spanked. Another I was with for three years was a romance version with the mountaintop Bed and Breakfast chalet with a huge bathtub filled with rose petals and 30 lit candles...she actually did this with me. These things are real, and finding out which version your woman desires is key.
That romance novels explain the root female fantasy... I had a real lightbulb moment there.
It seems to me that the main difference between male and female fantasies is that women want a long and intricate story around the relationship/bonding, but men are more into immediate and physical/sensual connection. Hence men dismiss romance novels as crap.
Also, men get tired of having to live up to the image of being strong all the time. It's really exhausting. So a common male fantasy is of a woman who is not only physically sexy, but also gives a man the space to be weak (or to relax) when he needs to be... something like a mother figure.
The greatest sexual fantasy I can even imagine would involve a woman that I was in love with. That deep pair bonding connection is rare and intoxicating. Sleeping with most women doesn't have that "I can't believe this is happening" quality. When I sleep with a woman I am in love with, it changes my world. It changes my whole experience of life. It is a fantasy too valuable to dwell on. It is forbidden fruit. But when it actually happens...it is fire.
The way I heard it described once was that women want the most powerful and wild man that they can control. That matches this description of romance with the "alpha" being that primordial force of a man that the woman is able to bond to her. Side note, its also why women love horses. The symbology of having control of a powerful beast between your legs doesn't get more blatant. Not that women see horses as romantic partners of course, its not about the animal its the dynamic playing out that is
OnlyXHub welcomes visitors to a unique platform where exclusivity and fun merge into an exciting combination. The platform serves as a vibrant mixtape of news from the adult entertainment scene and offers a fantastic opportunity to get to know their unique OnlyFans models.
OnlyXHub welcomes visitors to a unique platform where exclusivity and fun merge into an exciting combination. The platform serves as a vibrant mixtape of news from the adult entertainment scene and offers a fantastic opportunity to get to know their unique OnlyFans models.
OnlyXHub is a kick-ass platform with user-friendly features. With just a click, visitors will land on the page that interests them the most. The site is operational 24/7, giving users all the time in the world to fully explore its offerings. And here’s the kicker – OnlyXHub boasts a special feature not found everywhere else: the Top Models of the Week. This means that every Monday, based on the highest ratings, three models who were the hottest of the week are selected. This feature keeps users in the loop, ensuring they don’t miss out on checking out some sexy, hot babes.
Get ready for a full range of categories where all kinds of models come together, united by one thing – to bring pleasure and make you happier. On OnlyXHub, along with the models, you can live out any fantasy.
For example, MILFs will share their sexual experience with you, Latinas will show off their voluptuous curves, BDSM babes will make you submit and reveal their most intimate fetishes
And what about cosplayers? They’ll transform into your favorite character and play by their own rules with you. Asians will show their cutest side, while ebony models will captivate you with their ethnicity. We also have a gem – an incredibly attractive trans girl who will introduce you to the unusual side of sex. Blondes, brunettes, and redheads will show their true colors and immerse you in a world of pleasure and passion. Their models love sex, and you definitely won’t get bored with them.
Now, what about the content, you ask? You’ll find high-quality photos of sexy and diverse girls of all types. Models put a lot of effort into their content, so you’ll definitely be satisfied.
OnlyXHub is cranking out jaw-dropping content daily that users will love. The articles on the site fall into two distinct and exciting categories. First, there are reviews of top porn stars, providing exclusive glimpses into their hottest scenes. It’s akin to a buffet of porn, making it easy for users to find the perfect flick when they’re feeling frisky.
Thanks to OnlyXHub, users are well-informed about these actresses, which makes their solo sessions far more satisfying than watching random videos with fake names. Additionally, the platform offers reviews of different porn platforms, helping users know where to get their fix and which sites to avoid.
The second category is a whole different beast. Initially, the focus was going to be on reviewing adult toys for men, but the scope has broadened to help readers become sexperts. Users can look forward to articles about BDSM, Shibari, and guidance on selecting the right Tenga masturbator or a tight flashlight like Riley Reid’s. OnlyXHub isn’t just about text; it also provides visuals that immerse users in the experience, making them feel as if they’re part of the action.
Though still in its early stages, OnlyXHub is driven by a dedicated team committed to improving and expanding the platform. The team’s goal is to streamline the user experience, making it easier for visitors to transition from exploring models on the site to engaging with them on OnlyFans. This approach aims to eliminate the inefficiencies of navigating OnlyFans without prior knowledge, ensuring users can connect with models that truly appeal to them. With a growing lineup of models and diverse content, OnlyXHub invites users to join in and explore the dynamic world of adult entertainment it offers.
$BTC ETFs are growing fast. On Dec. 3 Which was yesterday, they gained $675 million. So they're closing in on Satoshi Nakamoto’s Bitcoin stash.
As of now they need about 13,000 BTC more to surpass him. But that's a lot of money too.
Part 1/8:
In the realm of technology and cybersecurity, staying updated is crucial, particularly when it comes to security patches. However, recent discussions have shed light on the drawbacks of automatic updates—specifically the risks of feature changes that don’t always align with user preferences or security needs.
Part 2/8:
The conversation began with a common sentiment that underscores the importance of automatic updates: they are essential for getting timely security updates, especially for vulnerabilities that may be actively exploited. Yet, it was also pointed out that updates can sometimes come with unintended consequences, such as breaking existing functionality or altering privacy settings.
There’s a notable divide between security updates and feature updates. Users can often manually choose what type of updates to receive for operating systems, but applications frequently lack this flexibility. A recurring theme is the importance of balancing the need for the latest security patches against the potential instabilities that new features might introduce.
Part 3/8:
For many, especially in production environments, the decision becomes a trade-off between maintaining stability and ensuring adequate security measures. The consensus leans towards automatic updates being beneficial for most users, as they protect against current exploits—though there will always be exceptions where updates inadvertently introduce vulnerabilities.
Regular review of one’s privacy and security settings becomes essential, especially with the risk of updates reverting preferred configurations. Tools that highlight reverted settings can be helpful, promoting a proactive approach in safeguarding data privacy.
Part 4/8:
The discourse emphasized that threat modeling is an ongoing process. Many have taken to subscribing to multiple resources, such as blogs or GitHub feeds, to remain informed about updates and their implications. This proactive attitude is increasingly necessary as the landscape of technological change continues to evolve at a breakneck pace.
A thought-provoking question arose regarding the potential of deep fake technology to bolster privacy amid current surveillance practices. Should deep fakes be leveraged to create noise in data, it could provide new dimensions to combat surveillance capitalism.
Part 5/8:
There is a growing body of skepticism about the efficacy of using AI-generated fake data to enhance privacy. Existing technologies aimed at obfuscating user identity, such as ad-blockers that click on every advertisement, have proved largely ineffective against sophisticated targeting by data trackers.
Deep fake technology poses an intriguing yet complicated solution. While there is potential in employing such techniques as a form of resistance against surveillance tactics, there remains considerable uncertainty regarding their practical implementation. The rapid evolution of surveillance technology suggests that any advantage gained could swiftly be countered by those employing AI for tracking and analysis.
Part 6/8:
The dialogue highlighted the notion that without concrete research and transparency in AI algorithms, strategies for using deep fakes for privacy gains remain speculative at best.
In another inquiry, a listener who is generally not a fan of Meta (formerly Facebook) asked how to control data collected while using their VR headset that requires a Meta account. Utilizing tools like VPNs, DNS settings, and specific applications to limit data collection became the focal point of the discussion.
Part 7/8:
Employing VPNs and DNS services, such as Pi-hole, can help in mitigating Meta’s telemetry to an extent. However, it was noted that blending multiple services could lead to complications like DNS leaks, affecting the overall user experience with the VPN and streaming services.
For the user at hand, the advice was to compartmentalize the usage of the VR headset and engage with it under less sensitive contexts. This suggests a broader principle for those engaging with platforms that prioritize data collection—maintain a separation to minimize risks.
Part 8/8:
As the podcast wrapped up, gratitude was extended to patrons who fuel discussions within the community. Through questions about security updates, the implications of emerging technologies like deep fakes, and the best practices for privacy amidst corporate data collection, participants engaged in a weighty exchange of knowledge.
The discussion leaves listeners with thoughtful insights on balancing security practices with personal privacy, emphasizing the importance of proactive measures in today’s tech-driven world.
The collective learning from this community not only aids individual users but also contributes to a larger understanding of the evolving landscape of privacy, security, and technology at large.
Part 1/8:
In a recent Q&A session of the Surveillance Support podcast, various themes about digital privacy, security practices, and user experiences were addressed. The conversation revolved around the complexities of threat modeling, particularly regarding the underlying knowledge required to accurately assess digital threats. Additionally, the speakers touched on the security concerns surrounding Android systems, the nuances of using custom mobile operating systems, and shared insights about popular privacy-centric podcasts.
Part 2/8:
One of the key questions raised by listener David Johnson revolved around the intricacies of threat modeling. The speakers acknowledged that while threat modeling is essential for understanding risks, it often requires a depth of knowledge that many may find challenging to acquire. Evaluating what constitutes a realistic threat versus an exaggerated one is a skill that seems to develop with experience.
Part 3/8:
Delving into personal experiences, the speakers reflected on their journeys toward understanding privacy and digital security. One host recounted their pivotal moment of realization while listening to a podcast discussing the risks of centralizing personal information within one ecosystem—specifically that of Google services. This notable epiphany prompted them to start diversifying their digital tools and adopting a more secure approach to their online presence.
Part 4/8:
The conversation further emphasized that there are two essential knowledge sets in the realm of digital privacy: technical knowledge and practical problem-solving skills. The hosts argued that often those who excel in technical proficiency may lack the practical skills necessary for effective threat modeling. Many have the capability to comprehend the mechanics of encryption or data security yet struggle in evaluating risks in real-world scenarios.
The discussion raised questions about how best to educate individuals on privacy without necessitating that they endure the "trial and error" phase that many seasoned users experience. The challenge lies in bridging the gap between theoretical knowledge and practical application for everyday users.
Part 5/8:
David also expressed concerns about the security of custom Android operating systems, commonly seen as offering increased privacy at the cost of some security guarantees present in stock Android systems. The hosts synthesized a few critical considerations:
Delayed Security Updates: Many custom operating systems may lag behind Google in delivering necessary security updates, posing risks for users reliant on timely patches.
False Sense of Security: Users might incorrectly assume that simply switching to a custom OS automatically makes them more secure, overlooking the need for robust security practices, such as selecting the right applications and managing permissions.
Okta shares soared Tuesday after the company reported strong third-quarter results.
Shares of Okta popped more than 18% in extended trading Tuesday after the identity management company released third-quarter results that beat analysts' estimates and offered rosy guidance.
Here's how the company did:
Earnings per share: 67 cents adjusted vs. 58 cents expected by LSEG
Revenue: $665 million vs. $650 million expected by LSEG
Salesforce reported its fiscal third-quarter earnings on Tuesday and posted strong fourth-quarter guidance.
Salesforce shares were up 9% on Tuesday after the company's fiscal third-quarter earnings report showed revenue and fiscal fourth-quarter guidance that exceeded analysts' expectations.
Here's how the company did compared with what Wall Street was expecting, based on a survey of analysts by LSEG:
Earnings per share: $2.41 adjusted vs. $2.44 expected
Revenue: $9.44 billion vs. $9.34 billion expected
Intel is considering tapping an outside hire as CEO, a nearly unprecedented occurrence, as the company reels from the ouster of Pat Gelsinger.
Intel is considering hiring an external candidate to replace ousted CEO Pat Gelsinger and has hired executive search firm Spencer Stuart to identify potential successors, according to people familiar with the matter, who requested anonymity due to the confidential nature of the discussions.
“I’m trying to find holiday gifts for my sisters. I open a bunch of tabs, I want my wife’s advice.”
That’s Browser Company CEO, Josh Miller, in his company’s latest ad for its new AI browser, Dia. Consulting your spouse to find gifts for your siblings is a pure — and dare I say, sweet — thing to do with a browser. But the new product he’s showcasing is replacing Arc, a beloved browser that put Miller’s company on the map. Not everyone is happy about the Browser Company’s pivot from Arc to AI browsing, and this latest commercial inadvertently explains why.
i checked recent Elon news and this is what I found
Elon Musk's AI highlights RCO Finance (RCOF), an altcoin with potential 21,404% gains by 2025. Early buyers can invest at $0.0777.
That means AI is already getting better at figuring out based on market data which crypto is right to invest in.
Google is already coming up with an answer for OpenAI soon to be released Sora AI.
Google’s new Veo tool makes videos from text or images, saving time and money. Businesses like Mondelez and WPP are already using it.
Katsukokoiso.AI is a project from professional photographer Eugenio Marongiu, an alpha tester on OpenAI’s text-to-video model Sora. Marongiu used the unreleased Sora model to create a piece titled “Surreal Elderhood,” which aims to balance absurdity and striking realism.
Amazon's going all-in on AI.
They're not just playing catch-up, they’re building supercomputers, rolling out custom chips, and launching their own models to rival OpenAI.
Even Apple’s intrigued.
It feels like they’re gunning to own the AI backbone.
Smart? Bold?
Definitely one to watch.
Cake help businesses compile and manage their open source AI infrastructure and reduce their engineering overheads.
A new company is emerging from stealth today with backing from Google’s AI-focused venture fund to help businesses compile their open-source AI infrastructure and reduce their engineering overheads.
The FTC just nailed two data brokers for selling ultra-precise location data, down to what room you’re in,
without consent. Hospitals, protests, churches, nothing was off-limits.
They’ve been ordered to clean up their act and delete improperly collected data. Honestly, it’s about time someone stepped in cause this level of privacy invasion is chilling.
Apple’s using Amazon’s AI chips for search and model training
They’re testing Amazon’s Trainium2 chips, claiming up to 50% efficiency gains, while still dabbling with Google and Nvidia tech.
What grabbed me is Apple showing off this partnership so openly at AWS’s event, rare move for them.
It looks like Apple’s quietly building its AI arsenal, but also hedging bets across all major cloud players.
There’s a new connected fitness device in town and it’s called Growl. Inspired by hardware companies like Peloton and Tonal, Growl is building a boxing bag that you can attach to your wall at home. Users can then start immersive, gamified boxing classes from the comfort of their home.
It also looks like the team draw some inspiration from boxing classes like Brooklyn Fitboxing — just spying an opportunity to offer a similarly intense fitness experience but without the user having to trek to a studio and sweat it out in a group setting.
Veeam brought on strategic investors including TPG and Temasek to bolster its cap table with sites set on an eventual IPO.
Veeam, which helps its users secure and recover their data, has seen its valuation spike in a recent secondary sale that was aimed at bolstering the company’s cap table before an eventual IPO.
Intel's CEO Gelsinger just got the boot, and it’s chaos.
His plan to fix Intel, like betting big on advanced chips and AI was too slow for the board. Intel’s losing to TSMC, Nvidia, and even Apple while struggling with huge losses and layoffs.
With no successor lined up, it’s a mess. Intel's trying to stay relevant but keeps missing the mark. National security implications make it even wilder.
HuggingFace's CEO warns that open source Chinese AI models risk spreading censorship worldwide.
China’s open source AI models have been making the news lately for their strong performance on various AI tasks such as coding and “reasoning.”
However, they have also attracted criticism — including from OpenAI employees — for censoring topics sensitive to the Chinese government, such as the Tiananmen Square massacre.
ServiceTitan made a few interesting disclosures about what it will do with the money from its IPO and who it will sell the stock to.
On Tuesday, cloud business software provider ServiceTitan offered a price range for its initial public stock of $52 to $57 a share, with hopes to raise $446.2 million to $514.2 million at the midrange.
It also made a few other interesting disclosures about what it will do with the money and who it will sell the stock to.
China imposed export controls on gallium and germanium in 2023, two metals that are vital parts of the semiconductor, telecommunications, and electric vehicle industries.
China has retaliated against the tightening export controls placed on it by the United States by banning shipments to the US of key materials used in semiconductor manufacturing and military applications. Gallium, germanium, antimony, and super-hard materials are some of the dual-use items China is prohibiting from export. Beijing is also imposing stricter controls on graphite.
Scientists have developed spray-on electronic tattoos that could be a game-changer in brain monitoring.
Researchers in Austin, Texas, are taking aim at an old but crucial tool for studying the human brain: the electroencephalogram, or EEG. For nearly a century, EEGs have helped scientists and doctors measure the electric pulses that fire through our brains, diagnosing everything from epilepsy to traumatic brain injuries. But they’ve always come with strings attached — literally.
Now, a team of scientists has introduced a stunning upgrade. They’ve created a liquid ink that can be sprayed directly onto a person’s scalp, forming an ultra-thin electronic tattoo. This “tattoo” EEG listens to the brain beneath it.
Mindfactory sold 4,350 motherboards during Black Friday week, according to data posted by TechEpiphany. Of that amount, 3,800, or 87.3%, were AMD boards, with just 550, or 12.6%, from Intel.
It appears that AMD is continuing to capitalize on Intel's failings. With the Raptor Lake stability issues and Arrow Lake proving a disappointment, Team Red is killing it in CPU and motherboard sales. During the recent Black Friday event, almost 90% of the motherboards sold by German retailer Mindfactory came from AMD, which is also dominating the US Amazon sales charts.
This new device uses light to perform the key operations of a deep neural network on a chip, opening the door to high-speed processors that can learn in real-time.
The deep neural network models that power today’s most demanding machine-learning applications have grown so large and complex that they are pushing the limits of traditional electronic computing hardware.
Photonic hardware, which can perform machine-learning computations with light, offers a faster and more energy-efficient alternative. However, there are some types of neural network computations that a photonic device can’t perform, requiring the use of off-chip electronics or other techniques that hamper speed and efficiency.
I checked out Spotify Wrapped’s new AI-powered podcast.
Two synthetic hosts chat about your top tracks, call you by name, and even guess your vibe based on your music.
It’s clever, but also a little too personal, like, are they spying on my shower playlists?
Cool innovation, but I can’t decide if it’s fun or freaky.
I read about Encore, this cool AI-powered platform for finding secondhand treasures online.
It’s like thrifting meets GPT-4, super specific searches, no SEO clutter, and results from sites like Poshmark or Mercari.
It’s clever, but paying $36/year for extra results? Not sure I’d splurge. Still, I love how it makes sustainable shopping feel effortless and fun.
Did you know scientists made a laser tech called Lambdis that spots landmines from a distance? It maps buried objects by detecting ground vibrations, even catching plastic mines metal detectors miss.
With 110M mines still out there and 84% of victims being civilians, this feels groundbreaking for both safety and efficiency. It could even check bridges or help in aerospace someday.
Unreal, right?
This interview with Tim Cook was packed!
He denies Apple’s ties to OpenAI (but his wording feels a bit slippery), insists AI doesn’t clash with Apple’s green goals, and teases Apple Glasses as the future after Vision Pro.
What stood out?
His claim that Apple’s legacy will be health tech, like saving lives with the Apple Watch’s heart alerts. Also, no plans to charge for AI features.
Cook makes a solid case that Apple isn’t late to AI, but is it all just Apple PR spin?
Have you seen JIDU’s insane new ROBO X EV, and wow, this car is next-level. It’s not just a supercar; it’s a fully autonomous race car with AI smarts and falcon-wing doors. 0 to 62 mph in under 1.9 seconds.
Plus, it’s rocking carbon fiber, a 360-degree safety system, and even Baidu’s Apollo tech for track-level self-driving.
It’s $6,915 to pre-order (in China), but no mass production until 2027. If I had the cash, I'd buy it just for those looks.
Who knew EVs could be this cool?
AI in medical imaging is exploding, over 50% of healthcare orgs now use it, compared to just 17% in 2018.
Neurology and stroke imaging lead the way, but tools also tackle mammograms and lung nodules. FDA approvals are skyrocketing too.
It’s exciting, but most systems only use it for 1-2 tasks. Imagine the impact as adoption grows, this is the start of something big.
I saw this take on Reddit about AI tools.
The author’s SaaS business thrives on pre-built prompts, but it seems most users prefer simple forms over direct interaction with ChatGPT. It makes sense when you think about it because ease beats skill for most.
But are we losing creativity by skipping the ‘prompting’ part? Sort of like outsourcing our imagination, no?
Robots learning to navigate like animals?
Yep, that's the vibe here.
QUT researchers used brain-inspired Spiking Neural Networks to boost navigation accuracy by 41%, and it’s super energy-efficient too. Think smarter robots for space or disaster zones, with less energy drain.
This is a cool tech with real-world potential.
Another invention coming up.
Imagine a robot so small it can crawl into a tissue sample and take photos or measure forces at the cellular level.
Cornell just made one. It’s 2 microns, magnet-controlled, and interacts with light to image what microscopes can’t.
The future of tiny tech is super wild.
The collaborations seek to simplify asset management by bringing traditional financial assets on-chain and expanding access to DeFi.
Backed, Sonic (formerly Fantom) and Chainlink have teamed up with Australian hedge fund Fortlake Asset Management to turn its Sigma Opportunities Fund into a tokenized investment.
Managed by Fortlake—recently named Fixed Income Manager of the Year in Australia— the fund’s units will be transformed into digital tokens, a process called tokenization, where real-world assets are represented on a blockchain, the companies announced exclusively to Decrypt.
#tokenization #fixedincome #investment #sonic #chainlink #defi
Reaching the geothermal energy far beneath our feet will require digging further into the Earth's crust than ever before.
Beneath our feet is an almost limitless source of energy, but while a few lucky locations have geothermal heat close to the surface, the rest of the world will need to dig a lot deeper. The challenge is how to get deep enough.
There are some spots around the world where energy literally bubbles to the surface. In Iceland, home to more than 200 volcanoes and dozens of natural hot springs, tapping into this energy isn't hard. Dotted around the country are steaming pools of water, heated by the geothermal fires that burn just below the crust. Boiling jets of water and steam are thrown into the air by geysers.
French electronics giant Metavasio, under the brand name Thomson Computing, has launched the world’s first web3 laptop, offering digital asset storage capabilities and support for decentralized applications.
According to a press release shared with crypto.news, the laptop features built-in cold storage for digital assets and an operating system tailored for interactions with decentralized networks and applications.
The PC system requirements for Indiana Jones and the Great Circle reveal that the game requires hardware-accelerated ray tracing at all visual presets.
Indiana Jones and the Great Circle might be 2024's most demanding PC game. It is one of a handful of titles that requires a GPU capable of hardware-accelerated ray tracing, and the game's listed specs for path tracing exceed prior titles that use the bleeding-edge technology.
Elon Musk scored priority access to NVIDIA for his xAI startup. Musk’s AI startup ordered $1.08 billion worth of NVIDIA’s GB200 AI servers.
NVIDIA’s GB200 AI servers are made by Foxconn. The chip company is expected to start the servers’ deliveries by January 2025, and Elon Musk just nabbed a priority spot for xAI, reported DigiTimes.
Musk reportedly approached NVIDIA CEO Jensen Huang directly about xAI’s GB200 server orders. Getting priority access to NVIDIA’s GB200 AI servers aligns with Musk’s goals for xAI.
Tesla China has reportedly clarified that it currently does not have any immediate plans to release the Cybertruck in the country. The update was posted by Chinese media outlet Jiemian News, which claimed that the Tesla China team had denied that the Cybertruck would be brought to the county soon.
The Cybertruck was unveiled in November 2019, and deliveries started in the United States on November 30, 2023. So far, the Cybertruck has only been delivered to the United States, Canada, and Mexico. A look at Tesla China’s official website shows that only the Model Y, Model 3, Model S, Model X, and the Powerwall can be ordered in China. The all-electric pickup truck is not listed as available for order on Tesla China’s official website.
SpaceX is reportedly discussing the possibility of selling insider shares, putting the aerospace company’s market value at $350 billion.
People familiar with the matter say that SpaceX is still discussing the possibility of selling insider shares with buyers and ironing out the details. However, its valuation of $350 billion is a significant increase from the $210 billion earlier this year.
In June, SpaceX was estimated to be the most valuable aerospace company, with a valuation of $210 billion. Keep in mind that SpaceX is not publicly traded. Over the summer, Bloomberg reported that SpaceX planned to sell insider shares at $112 per share in a tender offer.
The US has announced a new set of sanctions designed to restrict China's access to critical Western technologies that could enhance its AI and military capabilities.
Tech pundits broadly agree that advanced chipmaking and AI applications will be pivotal in shaping the 21st-century economy. The US and its allies, along with China, are investing staggering amounts of resources into these fields. However, maintaining a competitive edge over China in these critical areas is becoming increasingly challenging, even with aggressive government intervention. This difficulty stems from the complexity of global supply chains and China's ability to rapidly identify alternative methods for acquiring the materials and tools needed to meet future demands.
AI is taking over LinkedIn, one post at a time
Since ChatGPT launched, LinkedIn has become an AI hotspot, with a massive 189% jump in AI-generated content. More than half of long-form posts you scroll past now might be written by AI. Think of it like a digital co-writer quietly reshaping professional conversations. The shift is huge, imagine your favorite bloggers doubling their output overnight. AI isn't just here; it's running the show.
#ai #artificialintelligence #aicontent #linkedin #futureofwork #technology
Who’s Training AI With Your Posts?
Bluesky says it’s not using your content to train AI, but others, like Hugging Face, have accessed public posts for machine learning research. It’s a reminder that once something’s public, it’s fair game. Bluesky is working on ways to let users control consent, but stopping third-party use is tricky. Think of it like posting a note on a billboard, you control the message, but not who takes a photo of it.
New method could be applied to wide range of nanoparticles, promoting sustainability in lasers, biosensors and electronics
For the first time, researchers have shown that it is possible to recycle the nanoparticles used to make microscopic supraparticle lasers, addressing the need to sustainably manage these valuable materials.
Amazon Nova, unveiled at AWS re:Invent 2024, represents a bold vision for AI, offering multimodal models and unmatched accessibility.
When Amazon CEO Andy Jassy steps onto a stage, he does so with intent. At AWS re:Invent 2024, he wasn’t just unveiling a new product; he was laying out Amazon’s vision for the future of artificial intelligence. With the announcement of Amazon Nova, Jassy positioned the company as a frontrunner in an increasingly competitive AI landscape.
At its re:Invent conference, AWS announced on Tuesday a series of updates to Q Developer, its coding assistant platform that competes with the likes of GitHub Copilot. The focus here is on going beyond code completion and helping developers with a wider range of routine tasks involved in the end-to-end software life cycle.
The service, which you may remember under its previous name of “CodeWhisperer,” is part of AWS’s overall Amazon Q generative AI platform, which also includes Q Business (and which is also getting a slew of updates today).
Current weekly views remain below the all-time high of 9.3 million, indicating potential room for growth in retail engagement.
YouTube views of crypto influencer content reached 4.72 million for the week of Nov. 25.
The following is an excerpt from The Block’s Data and Insights newsletter.
AWS has announced the 'next generation' of its SageMaker AI dev platform, including a unified data dashboard.
It’s been close to a decade since Amazon Web Services (AWS), Amazon’s cloud computing division, announced SageMaker, its platform to create, train, and deploy AI models. While in previous years AWS has focused on greatly expanding SageMaker’s capabilities, this year, streamlining was the goal.
Quantum robots will likely have a mix of quantum and regular computers to handle both complex and simple tasks, work with existing systems, and communicate effectively.
A new study by an international team of scientists has outlined the potential for a new class of robots that can match the capabilities of humans. These advances are driven by the convergence of quantum computing and artificial intelligence, giving rise to the field of quantum robotics.
This field promises to overcome the limitations inherent in classical robotics and revolutionize various sectors.
The energy generated by the solar cells is used for driving or is fed directly into the high-voltage battery.
A top car manufacturing company is developing a new type of paint that could generate electricity. Mercedes-Benz’s engineers believe that their solar paint could revolutionize electric vehicles.
The new PV coating will consist of innovative solar modules, which will be applied to the car body in a way similar to a wafer-thin paste.
CUE6 uses AI to learn from mistakes, adjusting aim, posture, arm position, and shot strength in real-time like an athlete.
Toyota’s humanoid robot, CUE6, amazed onlookers by achieving the farthest basketball shot by a humanoid robot.
The AI-powered robot completed a seemingly difficult shot from 24.55 meters (80 feet 6 inches), earning its second Guinness World Records (GWR) title.
The feat was achieved in Nagakute, Aichi, Japan on September 26. For reference, the length of an NBA court is 29 meters (94 feet).
Nine Toyota volunteers initially developed CUE as an AI basketball-playing robot in 2017 during their free time. Over the years, CUE steadily improved its shooting accuracy and range.
The team fabricated a one-cubic centimeter cartilage tissue using 600 spheroids in less than 40 minutes.
A new 3D printing technique can rapidly produce biological tissues, outpacing existing methods by a factor of ten.
Penn State researchers have created a new 3D bio-printing technique called High-throughput Integrated Tissue Fabrication System for Bioprinting (HITS-Bio).
This new bioprinting technique uses cell clusters, called spheroids, to fabricate complex tissue.
This development could lead to a future where damaged organs could be replaced with lab-grown tissue, and diseases could be studied in highly accurate models.
But by and large, stopping AI academic fraud has not been a priority for most schools or educational institutions.
It’s been two years since ChatGTP made its public debut, and in no sector has its impact been more dramatic and detrimental than in education. Increasingly, homework and exam writing are being done by generative AI instead of students, turned in and passed off as authentic work for grades, credit, and degrees.
It is a serious problem that devalues the high school diploma and college degree. It is also sending an untold number of supposedly qualified people into careers and positions such as nurses, engineers, and firefighters, where their lack of actual learning could have dramatic and dangerous consequences.
The U.S. must investigate whether European governments have tried to unduly influence the country's laws around artificial intelligence, according to Sen. Ted Cruz, R-Texas.
Sen. Ted Cruz, R-Texas, has urged U.S. officials to investigate whether European governments have tried to unduly influence the country's laws around artificial intelligence.
Britain hosted the world's first global AI safety summit last year, bringing tech executives, world leaders and academics together to discuss a coordinated approach on regulation. Some countries have since set up their own AI safety institutes, which have continued to share expertise across borders.
New York City Mayor Eric Adams was mocked for taking his early 2022 paychecks in Bitcoin—but now its price is skyrocketing.
New York City mayor Eric Adams, known for his humility and low profile, delivered his latest message to his haters this week—this time on the subject of Bitcoin.
“Remember when y’all laughed at me when I first got my Bitcoin?” Adams said to a room of reporters during a press conference on Tuesday. “Who’s laughing now?”
Adams then laughed heartily, before adding: “Go look at my Bitcoin now.”
Grayscale recently filed to convert its Solana Trust into a spot ETF, marking a significant step in expanding institutional access to Solana.
Grayscale Investments is ramping up efforts to expand its product offerings in the cryptocurrency space by filing a proposal with the U.S. Securities and Exchange Commission (SEC) to convert its Grayscale Solana Trust (GSOL) into a spot exchange-traded fund (ETF). If approved, the fund would trade on NYSE Arca, joining a wave of similar applications by other firms.
Researchers explore the integration of quantum computing into robotics, highlighting advances in architecture, perception and interaction.
Researchers explore the integration of quantum computing into robotics, highlighting advances in architecture, perception and interaction.
The team identifies quantum computing’s potential to address classical robotics’ limitations, such as data processing, real-time response and cognitive functions.
The research shows that challenges — including hardware immaturity, quantum-classical integration and the need for secure hybrid networks — remain before scientists achieve the full transformative effects of quantum robots.