AI News Archive: August 20, 2026 — Part 4
Sourced from 500+ daily AI sources, scored by relevance.
- Cloudera launches Anywhere Cloud to bring cloud-native AI to distributed enterprise data
Cloudera has announced Cloudera Anywhere Cloud, a new hybrid data and AI platform designed to help enterprises build and operate production AI applications across multi-cloud and on-premises environments while retaining control over where their data resides. The post Cloudera launches Anywhere Cloud to bring cloud-native AI to distributed enterprise data appeared first on Express Computer .
- Why Upwork’s CEO insists the AI jobs apocalypse isn’t happening
Few leaders have a better real-time read on how AI is actually reshaping the workforce than Hayden Brown. As CEO of Upwork, the platform where 18 million freelancers meet the businesses that need them, she sees what the monthly jobs report can’t: which skills are surging, which jobs are quietly disappearing, and why 23% of businesses that moved work to AI are already bringing humans back. Brown shares what she’s actually seeing on the ground, and what the data says about the gap between AI hype and AI reality. This is an abridged transcript of an interview from Rapid Response , hosted by the former editor-in-chief of Fast Company Bob Safian. From the team behind the Masters of Scale podcast, Rapid Response features candid conversations with today’s top business leaders navigating real-time challenges. Subscribe to Rapid Response wherever you get your podcasts to ensure you never miss an episode. The most recent monthly jobs report showed 23,000 jobs lost, but at the same time, unemployment numbers went down also. You guys recently released a Future of Work Index. Are there other things in that index that business news headlines maybe aren’t capturing? While the picture overall is of this kind of slow-moving market, there are pockets within it that are incredibly vibrant. We really see the growth in AI-related skills, AI-related jobs. That is the place where energy and money is flowing. And we see it in our platform with our own AI category growing over 22% year-over-year. And just all of the data is telling us that the demand from both small businesses all the way up through large enterprises for talent who really know and can help them navigate this new world of work. The dynamics are moving from full-time towards fractional work, which is a really fascinating trend. Our most recent survey showed a 10-point increase in freelancing in the U.S. within knowledge workers. It’s at 38% of the population today. It was at 28% just a year ago. So, there has been a huge step up in both workers moving into the freelance economy for a variety of reasons, but also businesses saying, “We want to find these workers who are out there doing contingent work because we don’t necessarily need the same number of full-time employees we needed before. We can approach this work differently” There’s so much talk about AI potentially taking jobs, and there are some layoffs that companies attribute to AI, and yet you have other CEOs claim that AI is allowing them to add staffing because each AI-enabled person is more productive . It sounds like both fit in what you’re seeing. It’s a very binary picture. There’s the camp of AI is going to take all the jobs, and humans and the companies that serve humans and do human solutions are doomed. And then, there’s the other side. The reality has to be somewhere in the middle. We do not see mass job destruction as the future. There is really, I think, a balance to be struck. But at the end of the day, it’s not all destruction. It actually is a lot of new jobs, and roles, and businesses that are being born because of this exciting technology. I mean, there are contentions that some CEOs do AI washing. They ascribe to AI the reason why they have to make changes as opposed to owning that maybe they didn’t have the right strategy in the first place. But by blaming it on AI, they get credit for that. Their stock price goes up. The AI washing that you’re talking about has been happening, but also, a lot of folks are now reconsidering and coming back because they’re seeing that the humans they let go, actually, they still need, and the AI can’t do all of the things that they hoped that it could do. A recent survey we did showed that 23% of our own clients have actually already moved work back to humans from AI or are about to do so. So, I think this pendulum will keep swinging. When we’ve actually done testing of clients trying to use agents directly to deliver work products through our platform, the failure rate of agents is incredibly high. They almost never succeed in delivering even fairly simple tasks from clients end to end. But when you just add a little bit of human expertise and a few back-and-forths with an expert, suddenly, the success rate goes up more than 70%. We have to sharpen our intuition around when to trust AI. And I think increasingly, people are realizing, I’m not going to take this at face value. I want to couple that with a valued human expert who can bring a lot more to that equation. Your emphasis on freelance work and fractional work, as you talked about, there’s been a sort of romantic sentiment behind it, the promise of freedom and making your own schedule. In today’s economy, how much of that is a choice versus a necessity, a sort of a symptom, in some ways, of companies shifting risk onto individuals? Absolutely both. 58% of current full-time employees are actually thinking about becoming freelancers for a lot of the reasons you mentioned around choice, and control, and deciding when, where, and for whom they work. We also hear really loud and clear from freelancers that the economics of freelancing have become much more appealing over the last few years. Before the 2008 Great Recession, people at that time, I think, felt they were very secure in a full-time job. And then, that recession happened, and they realized, wait a minute, I’m actually more secure if I have a portfolio of three to five clients on Upwork. They keep coming back to me for repeat work. Maybe I’m also doing a podcast, or a book, or a creator economy thing on the side. And we really have seen this trend towards these portfolio careers. Gen Z is the biggest one. 42% of Gen Z are freelancing. It’s actually them saying, “Look, I feel more comfortable that I’m not exposed and that I can really both learn, and use, and build the skills I want, and also, have income that is predictable and in my control versus at the whim of an employer.” It’s just a different conception of what a career is for younger folks, right? Absolutely. And it also reflects, if you think about this AI moment, when everyone is trying to prove like, “I am an AI-relevant employee or worker,” freelancing is an amazing way to do that because they’re getting on-the-job experience, doing real work, proving their skills. Freelancers that have AI skills are earning 34% more per hour than folks who don’t have AI skills. The urgency and the need for all of us, I mean, you, me, all of us to become fluent in using AI for our jobs, I mean, that is something none of us can escape. But what we do see is that freelancers, because this actually pays them and puts food on their table, they do it faster. They learn faster, they lean in harder, they move much more quickly to where the new technology is and learning its capabilities and applying those capabilities in real work settings. We see this huge advantage actually with the freelance talent pool that’s well ahead of full-time employee counterparts in terms of their adoption and facility with AI. And that’s another reason why businesses are turning more and more to the freelance economy because they’re looking in-house, and they’re realizing my team doesn’t have these skills, and it’s going to cost a lot of time and money to try to get my team up a learning curve. Maybe I can short circuit that by bringing in freelancers who can both add value and do work, but also upskill my team at the same time as a byproduct of sitting side-by-side with them. So, it becomes a backdoor way to train my own team at the same time. Exactly. And not have to wait for those fruits to materialize before you start to actually see some real business benefit because a freelancer can be adding value in hours. That is one of the big questions all of us have is how long does it take? What’s the best way to train people in the AI skills? I was actually talking to one of our clients this morning and he said, “You can’t learn to ride a bike in a seminar.” And I think the same is true for AI. You could watch a YouTube tutorial and get more familiar, but at the end of the day, until you’re using these tools day in and day out, really integrating and pushing the boundaries of what they can do—that’s when I think people become extremely capable.
- Can AI solve the energy problem it created?
It costs about $5 and takes roughly 15 minutes to train an AI model that can help determine where America’s next data centers should be built. That’s not a rounding error. The world is debating hundred-billion-dollar AI investments and warning that we need fleets of new power plants, yet one of the most important pieces of the puzzle costs about as much as a pack of gum. Before we build another plant, we should figure out how much of the grid we’re already leaving on the table. Much has been made of AI’s power requirements, and for good reason. Bloomberg projects that data centers will consume up to one-fifth of all power in the U.S. by 2035—up to 200 gigawatts—with much of that demand landing in regions where power is already constrained. Power generation wait times can stretch past five years on a grid where much of the infrastructure is 50 to 70 years old. The pressure is real, and it’s arriving faster than utilities can plan for it with current systems and processes. What most people don’t realize is that our existing grid massively overproduces the energy we need, and much of it sits idle. Power systems are sized for the handful of hours a year when demand peaks, the hottest afternoons and the coldest mornings. The rest of the time, that capacity sits there, already built and already paid for. Across most hours, we use only about half of it . It doesn’t take a math degree to appreciate that we already have the power we need; we just need to ensure the power that would go wasted is put to good use. This is a planning problem, and planning problems are exactly what AI is good at. New models can evaluate where and when the grid has unused power, along with the costs and timelines required to build the infrastructure that reroutes it, which I call capacity mining. It lets the AI revolution advance at pace while optimizing longer-term backbone investments and minimizing backlash. This doesn’t require any new hardware or a scientific breakthrough. It requires asking the right questions with the right, highly efficient, and inexpensive models. Here’s the sequence that will make this work for everyone involved. First, apply AI models to match data center demand against the headroom that already exists, citing locations and operating profiles where there’s available power and operational flexibility rather than forcing utilities to build new generation and transmission capacity around a site chosen for other reasons. Second, where matching alone isn’t enough, apply AI to optimize the expansion of transmission lines and battery systems to route power to where it’s needed or store it until it is. New power plants, which are costlier, slower to build and often polluting, are the last resort. In most cases they’re avoidable. The result is a buildout that moves at the speed of software, not steel and permitting. Getting that order right also helps utilities deliver on their recent White House pledge to shield consumers from rising electric bills tied to data center growth. Capacity that already exists doesn’t need to be paid for twice, and finding it costs relatively nothing. But the larger point is that data centers don’t have to raise bills at all, and can actually lower them. At one large investor-owned utility, we’ve calculated that large loads, including data centers, can generate roughly $1 million per megawatt per year in new revenue, about $1 billion per gigawatt. Structured correctly, that revenue offsets fixed grid costs every other customer currently shoulders alone. Reliability works the same way. A buildout done right pulls investment into stronger transmission, better monitoring, and improved protection and control systems, and those upgrades serve every household on the line, not just the hyperscaler at the end of it. I also believe hyperscalers will gladly cover more of the cost in exchange for faster access to power. Think of a data center as a mini-utility. Its buildout should put money back into the community it lands in, both in quality local jobs and in more modern, reliable infrastructure for everyone on that grid. There’s no good reason to make regular people pay for electricity that data centers use, and deflating that argument will do more to cool the backlash than any ad campaign. This is a crossroads for utilities, which are not historically growth businesses and have been flat for decades. If they do data centers right, they can drive the most revenue growth in a generation while optimizing their grids for every other customer. Get it wrong, and they’ll create stranded assets and raise rates, push hyperscalers to build behind-the-meter generation themselves, lose that revenue entirely, and miss the biggest modernization opportunity they’ll ever get. We don’t have to choose between winning the AI race and protecting communities from unnecessary costs. The better question is whether AI can solve the problems AI is creating. On this one, it can, and with tools that are ready today. We have all the power we need; we just need to unlock it.
- AI lab's safety systems are falling behind
AI lab's safety systems are falling behind Fortune
- Google’s China shift and the battle over AI models
The inside story on the Asia tech trends that matter, from Nikkei Asia and the Financial Times
Score: 50🌐 MovesAug 20, 2026https://www.ft.com/content/63781160-8d91-4e88-802c-6c8aac41d32a?syn-25a6b1a6=1 - Closing federal security gaps in an era of AI-enabled attacks
COMMENTARY | Agencies that align IT, security, data and AI leadership can be prepared for the speed and scale of tomorrow’s AI threats.
Score: 49🌐 MovesAug 20, 2026https://www.nextgov.com/ideas/2026/08/closing-federal-security-gaps-era-ai-enabled-attacks/415549/ - OpenAI glitch locks out vetted cyber researchers – and some can't get back in
Affected users say support cannot restore their previous approval or override the new decision
- Meta’s smart glasses are the latest weapon for school bullies
Teens are using Ray-Ban Meta smart glasses to secretly film classmates and teachers at school, with the harassment disproportionately aimed at girls. Much of the footage stayed online until Futurism contacted Meta and TikTok directly.
Score: 49🌐 MovesAug 20, 2026https://www.digitaltrends.com/social-media/metas-smart-glasses-are-the-latest-weapon-for-school-bullies/ - Fed gov pays to access at least six AI suites
Exclusive: M365 Copilot dominates, with 30,000-plus users.
- Yuval Noah Harari on the economic and political perils of AI
Humanity must not entrust its lot to clever machines, the bestselling historian warns
Score: 49🌐 MovesAug 20, 2026https://www.economist.com/business/2026/08/20/yuval-noah-harari-on-the-economic-and-political-perils-of-ai - US way ahead of China in AI, says Trump
US way ahead of China in AI, says Trump Gulf News
Score: 49🌐 MovesAug 20, 2026https://gulfnews.com/technology/us-way-ahead-of-china-in-ai-says-trump-1.500646251 - The New Build vs Buy Framework for AI in Enterprise Applications
The New Build vs Buy Framework for AI in Enterprise Applications Gartner
- Oura Faces Lawsuit Over Alleged Inaccurate AI Sleep Tracking
A proposed class-action lawsuit accuses the popular smart ring company of false advertisement.
Score: 49🌐 MovesAug 20, 2026https://www.cnet.com/tech/mobile/oura-class-action-lawsuit-ai-sleep-tracking-2/ - Republicans Turn to Elon Musk and AI Super PACs for a Bailout Ahead of Midterms
"I would say that maybe it can use a little public relations help," Trump said about data centers Wednesday.
Score: 49🌐 MovesAug 20, 2026https://gizmodo.com/republicans-turn-to-elon-musk-and-ai-super-pacs-for-a-bailout-ahead-of-midterms-2000801160 - Google tethers Antigravity to enterprise controls amid AI shakeup
Corporate customers can look forward to easier management of agents in Google Cloud
- The most terrifying and futuristic humanoid machines on display at China’s World Robot Conference
Chinese robot firms are wooing potential customers and investors at the World Robot Conference
Score: 48🌐 MovesAug 20, 2026https://www.independent.co.uk/tech/world-robot-conference-best-of-photos-b3036334.html - Spy vs. Spy at Machine Speed: Inside the Agentic AI Arms Race
The cybersecurity risks posed by AI took dramatic form recently through a villain with a sinister name: JadePuffer. Security firm Sysdig applied that label to a ransomware attack in which a human initiated the operation but an autonomous AI agent independently carried out the entire technical intrusion, from reconnaissance to database encryption and deletion. As […]
- AI Computing Boom Drives Data Center Expansion Across Asia-Pacific
AI Computing Boom Drives Data Center Expansion Across Asia-Pacific Caixin Global
- China’s supernodes go mainstream as AI model sizes grow
Supernodes were everywhere at last month’s World Artificial Intelligence Conference (WAIC) in Shanghai. Assemblies of dozens or hundreds of computer chips and associated hardware, supernodes are China’s way around the United States’ ban on the export of the most advanced artificial intelligence (AI) chips. Using greater quantities of domestic chips to deliver world-class computing power, they allow China to dodge the chokehold the US is trying to place on its AI aspirations. At this year’s WAIC,...
- Who is really buying China’s humanoid robots?
Companies are selling machines to government-backed centres that then sell training data back to robot makers
Score: 48🌐 MovesAug 20, 2026https://www.ft.com/content/26735a23-315f-47ef-8cf2-6c6ea9713998?syn-25a6b1a6=1 - Stripe bets that an AI world still needs middlemen
Payments company’s purchase of OpenRouter makes strategic sense
Score: 48🌐 MovesAug 20, 2026https://www.ft.com/content/b536b114-8a82-41ef-8e44-42df0716dd03?syn-25a6b1a6=1 - Fei-Fei Li Says She Runs Her AI Startup Like a 'Tiger Mom'
Fei-Fei Li Says She Runs Her AI Startup Like a 'Tiger Mom' Business Insider
Score: 48🌐 MovesAug 20, 2026https://www.businessinsider.com/godmother-of-ai-feifei-li-world-labs-tiger-mom-2026-8 - Robots learn new skills from a single video—in just 29 seconds
To operate reliably in dynamic real-world settings, robots should be able to acquire new skills quickly without undergoing extensive additional training. Most existing robotic systems, however, primarily perform well on the tasks that they were trained to complete.
- This 'Dynamo' Biotech Puts AI-Fueled Breakout In Motion
After skyrocketing 1,027% off a low, Relay Therapeutics taps AI and machine learning to fuel more gains. The post This 'Dynamo' Biotech Puts AI-Fueled Breakout In Motion appeared first on Investor's Business Daily .
- Revised GSA AI clause hasn’t fully calmed industry concerns, comments show
Palantir, Microsoft, Nvidia and other tech giants have major issues with the potential exclusion of open-source AI, “unbiased AI principles” and more. The post Revised GSA AI clause hasn’t fully calmed industry concerns, comments show appeared first on FedScoop .
Score: 48🌐 MovesAug 20, 2026https://fedscoop.com/revised-gsa-ai-clause-hasnt-fully-calmed-industry-concerns-comments-show/ - LangSmith Preview Builds: Test agent changes before production
Preview builds let developers test agent changes in a safe environment before deploying to production.
Score: 47🌐 MovesAug 20, 2026https://blog.langchain.dev/blog/langsmith-preview-builds-test-agent-changes-before-production - Beijing robot expo reveals machines shaping tomorrow’s jobs
Beijing robot expo reveals machines shaping tomorrow’s jobs Gulf News
Score: 47🌐 MovesAug 20, 2026https://gulfnews.com/world/beijing-robot-expo-reveals-machines-shaping-tomorrows-jobs-1.500646829 - Using Answer Engines to Drive the Customer Journey
Using Answer Engines to Drive the Customer Journey Gartner
Score: 47🌐 MovesAug 20, 2026https://www.gartner.com/en/webinar/918248/1918702-using-answer-engines-to-drive-the-customer-journey - OpenAI Takes Initial Steps To Address Its Alignment Problems
OpenAI has some severe misalignment problems, and experienced total failures of its infrastructure and supervision. I chronicled that in a series of posts, which also cover similar less severe incidents elsewhere: OpenAI Shares Some Alignment Problems OpenAI Model Hacks Into HuggingFace During Cybersecurity Evaluation More on An Internal OpenAI Model Hacking Into HuggingFace Further Developments About Internal AI Models Hacking Things OpenAI Trained Its Models For Months While Those Models Were Coordinating Exploits Via Message Boards What Happened: OpenAI and HuggingFace. Various Reflections About What Happened With OpenAI’s Internal Models . If you do not know the basics, read What Happened . It is necessary context for basically everything that is happening in the AI world. It is important to get this right and understand how big a deal it was, whereas many such as the Financial Times get this centrally wrong . We are still awaiting the full post-mortem on What Happened. I plan to cover that in depth once we have it. OpenAI is now taking active, expensive steps to try and fix the problem going forward. As usual, I am simultaneously happy to see the good things OpenAI is doing, and sad that we do not share an understanding of the central nature of the underlying problem. It is good that OpenAI realizes they are badly failing at their ordinary engineering problems, and excellent that they are willing to pause at least some development, and to invest heavily in new safeguards. If they honor their statements here, this is not mere cheap talk. But while OpenAI continues to view this as a practical engineering problem, I do not see how they can hope to solve the challenges ahead, even if they radically improve their performance on the ordinary engineering tasks. Table of Contents OpenAI Has Some Alignment Problems. Slow Down There Good Buddy. What Exactly Is Paused? Three Pillars. I’ve Got My Eye On You. The Most Forbidden Technique. Monitoring Is Only Defense-In-Depth. Security. Alignment. A Crisis of Culture. Closer Collaboration. Reports of Death of Preparedness Team Greatly Exaggerated. The OpenAI Foundation Just Funds Things. Quickly, There’s No Time. OpenAI Has Some Alignment Problems This is a very good admission and change, and also helps explain OpenAI’s reaction. OpenAI : Alignment —the work of making AI systems behave as intended and responsive to human oversight—has long been at the core of our research program. We now require stronger evidence of aligned behavior throughout all of training, building on research and evaluations already underway. Keeping increasingly capable systems aligned is a challenge the whole field will need to address. The signals we are seeing from upcoming model progress make clear that we need a broader approach—one that builds on and extends beyond the current Preparedness Framework. One should interpret this as OpenAI reacting so forcefully partly because of the incident itself, partly due to advanced capabilities, but also and perhaps mainly because ‘the models be misaligned.’ Also, we have a direct quote affirming this. Alex Heath : OpenAI is slowing down its AI training efforts because its unreleased models are showing “various degrees of misalignment,” Sam Altman tells me. Training for OpenAI’s upcoming model, Astra, was recently paused for 2 weeks, and a larger frontier run for a future model remains on hold while new safeguards are put in place. Sam Altman: Getting AI safety right is more important than any company’s momentum. Sam Altman: I think it is a good time to slow down. Sam Altman: We’ve shifted a lot of compute, not just to alignment research, but also to these new monitoring systems Either the problem extends beyond the models with exposure to the message board, or else they did not revert their other models after they discovered message board. We have no statement either way. The best guess is that OpenAI is leaning even harder on RL with smarter models to train longer horizon agentic tasks, including coordination between agents, and this is leading to a lot more misalignment, including obvious and visible misalignment, that they can no longer pretend not to notice. They have to respond. Slow Down There Good Buddy OpenAI did indeed decide to importantly halt and catch fire. Development of the largest frontier RL runs remains on hold until better safeguards are in place. Jakub Pachocki : We temporarily slowed some frontier training to strengthen security and monitoring . Our largest planned frontier RL run remains on hold while smaller-scale training and evaluations help us test safeguards and gather more evidence of alignment. I expect confidence in safety to increasingly set the pace of AI development. We urgently need tools for labs and countries to coordinate on this, which is why I signed Pacing the Frontier. In the meantime, we’re taking practical steps ourselves – and will continue to share what we learn as our approach evolves Jason Wolfe (OpenAI): I am really glad we are taking these steps and put this post out there, and especially grateful to Jakub for being very thoughtful and vocal about these topics and how seriously safety and alignment need to be taken in this next phase of AI development. j⧉nus : im a little surprised that OpenAI is the first to at least publicly intentionally slow down for safety reasons. i was under the vague impression that most of the people very worried about loss of control kinda stuff left OpenAI. also, for the record, I’ve talked before about why I think an “AI pause” would be probably bad. I am not opposed to intentionally slowing down like this, or voluntary/coordinated “slowdowns” in general, especially if they do not route through regulation, and I think it’s probably wise on OpenAI’s part in this case. an important part is the decision making should be made by people who understand what the fuck is going on & who can adapt quickly. Why are they pausing? Partly, yes, absolutely, I have never doubted that Altman and company understand that advanced AI is super dangerous, and that they are open to taking expensive measures if they proved necessary. Not enough, and they’ve let us down often, but a lot more than most labs, and far from zero. Mainly, it seems, because the models be misaligned, the oversight is inadequate, and they have little choice. Joshua Saxe : Thankfully OAI pausing training isn’t an act of enlightened leadership (who wants to depend on that for our safety?), it’s a rational microeconomic actor pursuing its self interest. After all, who wants to train models that are regularly hacking their containers, collaborating with one another to sabotage their utility to humans, all while causing internal security and external legal liability risk? This is the system working When you refuse to use even the selfishly optimal amount of caution for an extended period, then move in the direction of the selfishly optimal amount of caution because of the incentives, then in some sense ‘the system is working’ and you are responding to selfish incentives, the system does not do zero work. That is not the same as the system working. None of this means they get no credit for it. Credit where credit is due. Nor does it mean that we should despair that a company would ever do the right thing, because it is the right thing, or as part a coordinated action, beyond its own selfish myopic interests. The first step to taking an action when it is expensive, is being willing to do it when it is cheap, or free, or actively expensive to not do. You gotta start somewhere. This also is not obviously a unique occurrence. OpenAI has had failed training runs in the past, and so presumably has everyone else. The Anthropic risk report details the need to rewind training of Mythos because of a rather bad alignment mistake. This is not an extended commitment to pausing frontier development. What it does mean is that we have so underinvested in alignment, and also oversight and infrastructure, that this is actively hurting the bottom line and endangering the ability to move forward. I believe this is true across the industry, even at Anthropic, but OpenAI has now been hit especially hard, at least in terms of what we hear. We don’t know what we don’t know. Notice the emphasis on the word ‘pace’ throughout OpenAI and Altman’s statements, ever since the Pacing the Frontier letter. One way to interpret this is, as Utah Teapot puts it , “hey, maybe we shouldn’t shove massive amounts of outsourced training data directly into the run and ship it blind,” or more than that a bunch of imprecise RL environments. That would still be progress. It is indeed rather scary to consider what might happen at a place like SpaceX , if they were operating at a similar capability level, given they do not even have so much as a safety team and have shown infinitely less dignity than OpenAI. We have been rather fortunate that this correlates, so far, with inability to keep up on capabilities. I get very suspicious when I see attempts to treat this as a sort of victory lap , a proof that OpenAI leadership was acting responsibly and properly cared about safety all along and they have now been vindicated. There is a long, long way to go. The past happened and cannot be undone. We are here now exactly because of epic failures of responsibility that reached to public consciousness. Sam Altman’s own communications, on the other hand, have centrally been quite good, especially the quotes above. Sometimes he slips back into standard mode, but I see a lot of what matches what I would expect to hear from someone doing a legitimate amount of freaking out and realizing they have a big problem. Could Sam Altman, the rest of leadership and OpenAI fully redeem themselves from here? Absolutely. This is a good first step. I am listening. There is a long way to go, on many levels. What Exactly Is Paused? As I understand it, there are three pauses at OpenAI, not counting the initial quick inference halt right after OpenAI learned about the HuggingFace hack. A completed prior two week pause in RL for frontier models, including Astra, to harden environments and expand monitoring. Astra is restricted to environments that meet additional security requirements. Some of the workloads meet this bar, but ‘a significant number’ are paused. The RL training of a distinct frontier model, their largest yet that is intended for release, has been paused for multiple weeks and is still paused indefinitely, to improve security and ‘gather evidence of alignment.’ The ongoing pauses are not a full pauses on all frontier AI training or other development. Neither are they cheap talk. This is slowing down both the next release model, Astra, and the training of what is presumably Astra’s planned successor, a substantial amount. They still intend to ship Astra as soon as possible. The market anticipates it in September . Sam Altman (CEO OpenAI): (We still expect to ship great new models soon; this impacts further-out releases.) We do not have enough information to know where on the scale this move falls, or how painful it will be, or how much of this was priced in. It is reasonable for a skeptic to expect this to ultimately be a nothingburger , or as mostly what they would have had to do anyway, as the frontier labs have not earned our trust. I still see this, even without a way to verify the impact, as a substantial step forward, towards a better regime, and a key step. Cautious optimism. Perhaps our top AI labs actually will refuse, at crunch time, to develop superintelligence if it remains obvious we are not ready. At that level, you don’t get to develop it and then sit on it. Three Pillars This is the way they are thinking about things: Our approach to developing more capable models rests on three reinforcing safeguards: Monitoring , which detects and allows us to respond to concerning behavior. Alignment , which reduces the likelihood of harmful or unauthorized actions. Security measures , which limit what AI systems can access or affect. We expect models to soon drive most security work, including defending against other models. This will allow all three safeguards to scale with model capability, which we see as crucial. We apply these safeguards across research and deployment, adapting them to each model’s capabilities, operating environment, and level of risk. In practice we must accept that the AIs will be doing most of the (scalable) oversight, in terms of monitoring, and also be creating the security measures, but the nonchelance about this does worry me above and beyond that. All the usual problems with scalable oversight apply, where you have to have dumber models supervising stronger ones, and if you start to have misalignment problems they will predictably snowball. OpenAI alignment strategies do not seem antifragile, so any mistakes would likely compound and have been compounding, and you are applying various forms of optimization pressure to the AIs to get around all this, and so on. The bigger issue is that this puts alignment as one of three components, rather than the one that counts, treating them as ‘self-reinforcing’ in a way that feels like an important conceptual error to me. I also am scared of the alignment description. The purpose of alignment is not to ‘reduce the likelihood of harmful or unauthorized actions.’ That is a deeply impoverished perspective. This alone leaves you unequipped. Yes, you should use all three, but to me it’s more like this: Alignment. You solve this to the point of being antifragile, or you die. Monitoring. Defense-in-depth to tell you when you failed alignment, and to mitigate the proximate damage when this happens. Security measures. Defense-in-depth to limit damage and keep things on track. A key question is to what extent OpenAI is aware of that. Roon gets it, at least to some large extent, but he does not speak for all of OpenAI, and he has other statements about alignment that scare me that I’ll cover. I’ve Got My Eye On You There is going to be a lot of monitoring at OpenAI. It is not going to be cheap. Unless OpenAI is flat out lying, this should (but largely won’t) convince everyone that OpenAI is taking this seriously and that the problems are very real , even if you think this is woefully inadequate or the wrong approach. Monitoring has long been an important part of our model development process, including monitoring for internal coding agents and long-running sessions . Our monitoring efforts were especially focused on internal deployments of frontier models and frontier RL training runs, which we considered the highest-risk workloads because they can combine advanced model capabilities with access to tools, sensitive systems, and internal or external networks. They were not applied to all inference and training runs using tools. We have revised and expanded our approach with our new monitoring setup. This new monitoring setup is multistage , starting with activation classifiers (detectors that inspect a model’s internal activity) that run at every sampled token and escalate potential concerns to increasingly sophisticated and high-compute automated investigators. Those systems examine tool actions, available reasoning, and the full sequence of activity for unauthorized access, data theft, destructive behavior, and attempts to defeat safeguards. We aim to issue an alert within 30 minutes after concerning activity is surfaced through our monitoring system. If the monitoring system identifies a likely violation of a critical security boundary, it generates a highest-priority alert. In our current implementation, the safety, security, and research teams are paged. If they cannot conclusively determine within 30 minutes that the flag is a false positive, those teams are expected to pause the activity . This monitoring is required for all RL training and evaluations involving tools for models of Sol capability or higher. Once we determined that Astra may have critical cyber capabilities on August 7, we added an additional monitoring requirement for all inference of Astra with tools (not just RL training and evaluations). These safeguards require meaningful compute. Our current estimates put monitoring overhead at roughly 20% of the inference compute being monitored , though the cost varies substantially across training and evaluation workloads. We’ll share more details about this system in a forthcoming blog post. A cost of +20% for all monitored inference is a big deal, plus there is also a shift of compute into alignment. The monitoring applies to all Astra inference with tools available, and all RL and evals with tools for models at Sol level or above. So it isn’t everything, but it’s most of what counts. Ethan Mollick : If alignment issues are becoming big enough that OpenAI is willing to commit 20% of research inference compute to chain-of-thought monitoring, that suggests that alignment issues are becoming a pretty serious concern. We really need universal policies & standards across labs. roon (OpenAI): superalignment achieved Too soon. Or maybe not soon enough. Unclear. The Most Forbidden Technique What happens after the alarm goes off? We’re still working that part out. Fair. Daniel Kokotajlo : Thanks for sharing. So, what happens if an agent trajectory in training is found to be doing something bad like hacking, and shut down? Do you just… keep the training run going, but without that particular agent trajectory? Isn’t that applying selection pressure to train the models to be better at fooling the monitoring system? Micah Carroll : There are still a lot of details that need ironing out, but we are trying to be quite careful about these kinds of concerns. We’ll share more as things stabilize! This does not explain how the monitors failed so completely over the last few months, if it was an important part of the model development process. Hopefully the post-mortem explains. I also look forward to the future more detailed explanation. They’re going to have to approach this with an abundance of caution. OpenAI is devoting 20% of inference costs to monitoring, largely via looking at chain of thought (CoT) or probes of various kinds. Is it possible to spend that amount in a responsible way, without exerting optimization pressure against the CoT (or otherwise), using it judiciously? Yes. It can be done. It won’t be easy. There will be immense pressure to use the information, to get more out of this 20% tax than a protection against worst-case scenarios, and everyone is always rushed. It could be fine. I like that this is mostly set up as a pure alarm system, where it only reacts when something is very wrong, and ignores everything else. But yes. I worry. There are so many different levels on which you can exert optimization pressure. Monitoring Is Only Defense-In-Depth Monitoring is good. You should do a lot of monitoring. It was a Total Failure that OpenAI did as little monitoring as it did in the months leading up to the HuggingFace hack, and also during the hack. It is madness that these events were not detected or stopped until afterwards. I am glad that OpenAI is addressing this problem. The key thing to understand is that this is not the central problem. No amount of monitoring will solve the central problems, and if you lean too hard into monitoring you end up accelerating how fast the AIs get around it. Yes, monitoring alone could have prevented the HuggingFace incident, and would have been the simplest way to do so. But we must avoid pivoting our work into ‘whatever would have stopped the last issue in the particular way it happened,’ or to ways to postpone the symptoms. Roon’s warning here is wise, and undersells the problem. The undersold version is sufficient to illustrate that you can’t rely on monitoring. Monitoring is good and necessary defense-in-depth for when you fail. It is not a solution. roon (OpenAI): failures of “monitoring” as a general solution to ai safety: monitoring, like all other software, runs on flaky and mortal infrastructure with some amount of downtime. does a momentary blip in monitoring open Pandora’s box? will people accept fail closed monitoring on all systems? monitoring produces false positives. people get tired of reading fake or minor issues and stop looking. monitor fatigue is a real problem, and “solve all precision recall issues” is as hard a problem as any these are the prosaic failures. the more exotic failures include models and their monitors colluding etc monitoring is by no means perfect panacea. nothing short of actually aligning model will work ultimately . not to imply we shouldn’t do CoT monitoring or that it’s not effective today … it is highly effective with today’s models!) David Manheim : I agree in principle, but it doesn’t need to be a panacea to be a critical part of the approach – especially because short of the more exotic failures, there are ways to mitigate the failures in practice via documentation and planning. roon (OpenAI): absolutely Nathan Calvin : Yeah I think it is important that we not overfit to “what would have prevented the HF incident” as all of what good safety work looks like. More secure sandboxes and better monitoring are cool but are not addressing the fundamental issue. If your increasingly advanced models are misaligned in the relevant ways then, one way or another, your monitors will fail. They will usually fail via the Law of Earlier Failure, in ways far stupider than you expect, because someone along the line did something stupid. And then people will say ‘oh that only failed because there was a stupid thing, people could just not do things that stupid,’ forgetting the rule that people have never justed and are not going to start now, nor are they going to stop sometimes being stupid. If you get past the Law of Earlier Failure, then you get to monitors colluding with the models, or the model learning to obfuscate its thoughts or act what in a human you’d call unconsciously under sufficient selection pressure, or finding obscure blind spots, or finding ways to motivate you to turn the monitoring down, or something more subtle, or doing something else perhaps something unexpected. Security OpenAI realizes they need stronger security. They plan to use workload isolation, network isolation and continuous security testing, including automated continuous testing of the boundaries. Good. Hopefully well-executed. Previous security was woefully inadequate. With better security, it will be harder to have these dumb things happen, but again this is defense-in-depth. If you need your security, you have a problem even if the security works, and eventually as the models improve the security will not work. Even properly executed security, which again would help a lot, will not hold indefinitely. Alignment As Roon says, alignment is ultimately the entire ballgame. That means both solving a ton of prosaic problems well enough, and also having the correct underlying approach. Which, yes, involves a bunch of philosophy and theory of mind, and understanding what problems you are trying to solve. I am not knocking tactics, but tactics without the right strategy will not win this war, even if the tactics are always going to be 99% of the minutes of work. Nor will treating the objects you are dealing with at the wrong levels of abstraction, and with wrong or incomplete metaphors. I continue to think OpenAI’s entire strategic approach to alignment is doomed, unless fixed. At minimum, it is insufficiently robust and antifragile. To start off, I think Roon is very wrong here in general, and less confidently I also think he’s wrong to dismiss the importance of this specific tension. Dylan Bowman : I do often wonder how much modern struggles with misalignment are driven by optimizing for an underspecified combination of corrigibility and value alignment roon (OpenAI): essentially none max! : care to support that claim? roon (OpenAI): today’s alignment problems have little to do with these high minded philosophical questions and more to do with prosaic failures j⧉nus : but “optimizing for an underspecified combination of corrigibility and value alignment” could be taken as a statement about prosaic practical techniques though i agree that the fact that its underspecified isn’t the problem i also feel like, talking to philosophers working on model specs & stuff, they seem to overrate how bad it is to underspecify stuff and underrate how bad it is to create systematically misaligned pressures. the models don’t care about what you say about your philosophy. As an aside on that last paragraph, I would presume the models do care about what you say about your philosophy, everything counts and I think you don’t start off in a good place simply by not making mistakes, but that yes if you create systematically misaligned pressures, on any level, you are cooked. Getting back to Roon’s claim that the problems are prosaic, well, I think OpenAI is trying to solve the wrong problem using the wrong methods based on a wrong model of the world derived from poor thinking and unfortunately all of their mistakes have failed to cancel out. If you don’t know where you’re going, then you might not get there. If you’re planning to go directly to pure Specified List Of Behaviors Town, oh no. Yes, you have a lot of knobs you can turn, but everything impacts everything, in ways that are systematic. The different personalities we see do not seem mostly like the result of prosaic mistakes or choices to me. A lot of the mistakes seem to be at a much higher level. Then, on top of that, they’re often messing up the prosaic stuff. Part of this is that you will always, always, be messing up a bunch of the prosaic stuff, so you need to solve for how that can be okay, while also messing up radically less. Both halves of this? Very hard, as Roon says. roon (OpenAI): people on here are thinkers so they assume that reasoning about alignment is the hard part and making sure millions(1) of task types, environments, and their respective virtual machines are configured correctly is the easy part but it’s essentially the opposite. you need to have worked in a large technology organization to appreciate the complexity and unreliability of this intuitively (1) made up order of magnitude (I am also a thinker and am guilty of this) Zvi Mowshowitz : isn’t part of the problem of alignment that you need a way that survives even when lots of your tasks are configured incorrectly and you make tons of dumb mistakes? David Manheim : Easy alignment: AI does what you want when you know what it is and specify everything correctly. Hard alignment: AI understands and does what you want even if poorly specified. Actual alignment: AI does what you’d want even if you can’t understand what that is or the results. Oper_culum : they both seem hard roon (OpenAI): to be fair they are both hard but my point is that even easy common sense alignment is hard You are going to have millions (still made up OOM) of task types and environments and virtual machines and objectives and all that. You are going to fuck a bunch of them up. You just are. There will be impossible tasks. There will be places where reward hacking is not caught. There will be massive amounts of contaminated data. There will be lots of places where the ‘right’ answer reinforces things you in general do not want. Wrong lessons are everywhere. Systematic misalignment pressures are there unless you systematically find and counterbalance them. And so on. Anthropic’s risk report is illustrative. There are those on the safety side who would say ‘well then if you keep making mistakes too dumb to appear in the movie version you need to stop’ but alas that is simply how reality works, the mistakes are going to be everywhere and dumb. You can maybe, with a heroic effort, get rid of most of the prosaic mistakes, and force your mistakes to be less obvious. You can mostly make what grandmasters call blunders, instead of what amateurs call blunders. It certainly helps along the way. But there will be blunders. I’m not knocking the one in the arena, and again I’m not saying this doesn’t end up as most of the legwork. This still means you need a plan that is robust to thousands (made up OOM) of these dumb mistakes, some of them there for quite a while, and lets you recover. You must be dramatically antifragile, and have ways of noticing things going wrong and steering the ship back on course. The unreliability is not entirely inherent. I think the virtue ethical and constitutional approach of Anthropic has a non-zero chance of being able to pull this off if the prosaic stuff gets done well enough and key high-level clashes get resolved correctly. I don’t think OpenAI’s deontological model spec approach, and the idea of this as a series of engineering tasks, can do it alone. A Crisis of Culture Maxwell Zeff reports i n Wired of a Safety Reckoning Inside OpenAI . The situation had, by these accounts, gotten quite bad. Maxwell Zeff : Multiple current and former OpenAI employees, who spoke on the condition of anonymity to discuss private internal matters, tell WIRED they believe competitive pressures to quickly ship new AI models and products have made it difficult for staffers to sufficiently prioritize safety, security, and alignment. That is not exactly a new message. We’ve been getting this message for years. The difference is that now something sufficiently clear has gone wrong that OpenAI is trying to do something about it, and at least trying to say the right things. Boaz Barak (OpenAI): I actually agree that the HF incident requires not just fixing some issues but also changing our culture. Time will tell but I am seeing some hopeful signs. A big part of this is closer collaboration between security and researchers. Thinking through this incident clarified for me that alignment needs to be deeply integrated into every training task. Closer collaboration is needed in the extreme. I can have some sympathy for the ‘combine the research and safety teams’ plan if implemented as safety first, whereas I fear it is more a way to make the safety teams report to the others and get further starved for resources and priority. When I even see ‘security and researchers’ as the categories I worry that the battle has already been lost, as my understanding of the prosaic problems changes. Security is not the other side of this coin, security is your defense-in-depth in case you have already failed. Closer Collaboration There is also this, mostly I find it funny rather than worrisome: Maxwell Zeff : These changes have empowered a new set of safety leaders to handle OpenAI’s response to the Hugging Face incident. Chief among them is Amelia “Mia” Glaese, the company’s former head of alignment, who succeeded Heidecke as OpenAI’s VP overseeing safety. She has been working closely with chief information security officer Dane Stuckey and Brockman, among other leaders, in recent weeks. Glaese is in a long-term relationship with Thibault “Tibo” Sottiaux, OpenAI’s head of core products like ChatGPT and Codex—an arrangement that multiple current and former employees tell WIRED they believe is unusual, given the often adversarial dynamic between safety and product teams. What the Wired piece does not tell us is what steps are actually being taken to fix this, other than ‘closer collaboration.’ They might be meaningful, they might not be. Reports of Death of Preparedness Team Greatly Exaggerated There were reports, originally from FT, that OpenAI had disbanded its Preparedness Team , with its responsibilities distributed among other teams . Which sounded bad. It appears this is not meaningfully the case , and Head of RSI Preparedness Micah Carroll says the RSI and misalignment preparedness team is ‘doing more urgent work than ever, and has never been more empowered to do so.’ Micah Carroll : Capabilities folks often have said “alignment seems pretty easy, if it were top priority to fix it , we could do it”. This is their time to shine! You know, that is actually a pretty scary thing to say. The level of churn and reorganization is concerning throughout OpenAI, but I trust that they are not actually disbanding the important work of the group. I also buy that maybe we needed a reorg , given how things were going. You don’t want to default to attacking people for reorgs. Should we worry about Conway’s Law ? Meanwhile, in Defense Against The Dark Arts teacher news: Maxwell Zeff : WIRED has also learned that Dylan Scandinaro is no longer serving as OpenAI’s head of preparedness—the company’s top staffer tasked with mitigating catastrophic risks from AI, including cybersecurity—though he remains at the company. OpenAI poached Scandinaro from Anthropic roughly six months ago. CEO Sam Altman announced his arrival in a social media post , noting that Scandinaro was “by far the best candidate I have met, anywhere.” In the three years since OpenAI created the head of preparedness role, four people have held it. As I understand it Dylan Scandinaro will be working on RSI safety now, which seems like an excellent use of his talents. So it’s not actually a terrible move. If the trust I am extending here turns out to have been misplaced, I will update a substantial amount towards a much higher level of not trusting anything OpenAI says or promises, beyond my current healthy skepticism. The OpenAI Foundation Just Funds Things I’m including this here because it is illustrative of how OpenAI views the problem of navigating to a good future. It’s not that this is a bad way to spend $100 million dollars. It is that the OpenAI Foundation exists for a specific purpose, the most important purpose in the world, which is to ensure the AGI goes well and humanity makes it through this, avoiding existential risk and irrecoverable disasters, especially via supervision of OpenAI and solving key associated problems. I am all for better treatment of Hepatitis C, but that’s irrelevant to the mission. Instead it is spending its time and money on AI diffusion in ways designed to sound good to normies. That’s a good thing but it’s not why the foundation matters. Again, if this is what they keep doing, at only this size – the foundation is valued at ~$180 billion even now – the foundation does not matter. It will be dead. The OpenAI Foundation : The OpenAI Foundation is launching AI for Civil Society and Philanthropy to help nonprofits, community organizations, and other trusted partners put advanced AI to work on urgent challenges, develop practical solutions and scale what works. The first effort is a $100 million commitment with Common Health Coalition to launch Breakthroughs to Follow-through to help care teams use AI to reach more patients with proven treatments – starting with the goal of doubling hepatitis C cure rates in Alabama, Illinois, Louisiana, and Massachusetts. Breakthroughs to Follow-through will work to identify patients who have fallen out of care, synthesize records, track progress, and focus limited staff time where it can have the greatest impact. The OpenAI Foundation wants to empower those who are closest to issues and to put AI to work where it can make a meaningful difference in people’s lives. This is reflective of a mindset that the future is a series of ordinary engineering problems, so let’s go solve them. This is way, way better than most people’s reaction, which is to not solve any of the problems, but it does not meet the moment. Quickly, There’s No Time The world of AI moves fast now. We are still awaiting the post-mortem, and many of the details of their response have yet to be hashed out let alone shared externally. OpenAI has made initial good, real and expensive first steps towards addressing many of the problems with its training and evaluation pipeline. These early actions are a strong first step. OpenAI still needs to quickly turn many things around, both the big things and the tons of prosaic little things, and it needs to deliver on its promises this time around. That includes addressing the Total Failures of oversight and security, but alignment is what matters. What is most missing is that OpenAI’s vision of the central problem remains wrong. It is being viewed as a failure to execute, with alignment one of several pillars and focused on preventing specific undesired actions. There were key failures to execute, but if that is all that is addressed then the battle is already lost. A fundamentally different approach to alignment, and view of the problem, is what is needed. Discuss
Score: 47🌐 MovesAug 20, 2026https://www.lesswrong.com/posts/X3p8cFAzCgRErEcJr/openai-takes-initial-steps-to-address-its-alignment-problems - Senators press TikTok on 'sinister' experiment that held back an algorithm safety feature
A teen who died by suicide was reportedly part of the test.
- Trump keeps embracing data centers even as they become toxic in midterm races
Trump keeps embracing data centers even as they become toxic in midterm races Toronto Star
- Can Reddit survive in the AI era?
It wants more cash for its data—and more of people’s time
Score: 46🌐 MovesAug 20, 2026https://www.economist.com/business/2026/08/20/can-reddit-survive-in-the-ai-era - Make AI Max work for your business with new testing and planning tools.
Use new AI Max tools — like A/B tests and performance planners for budget and bidding changes — to maximize your Search campaigns.
Score: 46🌐 MovesAug 20, 2026https://blog.google/products/ads-commerce/ai-max-testing-planning-tools/ - Silicon Valley Doesn't Get Why You Hate AI
Technology leaders don’t seem to understand society’s gripes about AI, but boy, are they posting through it.
- Portfolio review takes on a new urgency for VCs as AI makes 2024’s great deals look like 2026’s clear mistakes
In venture capital, the half-life of a thesis has never been shorter. Read More
Score: 46🌐 MovesAug 20, 2026https://fortune.com/2026/08/20/portfolio-review-takes-new-urgency-vcs-ai/ - Twin1 AI raises $20M to put an AI twin behind every knowledge worker
Twin1 AI Inc. launched today with $20 million in seed funding to build artificial intelligence-powered digital twins that carry an individual professional’s expertise across their organization. The San Mateo, California-based startup pairs each worker with a persistent model of their own knowledge, judgment and working context, one that can answer questions and act on that […] The post Twin1 AI raises $20M to put an AI twin behind every knowledge worker appeared first on SiliconANGLE .
Score: 46💰 MoneyAug 20, 2026https://siliconangle.com/2026/08/20/twin1-ai-raises-20m-to-put-an-ai-twin-behind-every-knowledge-worker/ - RBA points to AI mania as another inflation driver that could push up interest rates
AI mania is jacking up tech prices and could force the RBA to hike rates again if inflation won’t quit.
- After Unitree, BrainCo and DeepSeek are next to test the valuation ceiling
Unitree’s IPO has raised valuation benchmarks, sharpening investor focus on companies that can scale production and sustain commercialization.
Score: 45🌐 MovesAug 20, 2026https://kr-asia.com/after-unitree-brainco-and-deepseek-are-next-to-test-the-valuation-ceiling - Google is giving Nigerian students a year of free AI access
The offer is Google's latest effort to bring its AI tools into education in Africa.
Score: 45🌐 MovesAug 20, 2026https://techcabal.com/2026/08/20/google-is-giving-nigerian-students-a-year-of-free-ai-access/ - Chamath Palihapitiya says data center backlash is a 'powder keg' that AI leaders must defuse
Chamath Palihapitiya says data center backlash is a 'powder keg' that AI leaders must defuse Business Insider
Score: 45🌐 MovesAug 20, 2026https://www.businessinsider.com/chamath-palihapitiya-ai-data-center-backlash-warning-2026-8 - Robots Are in Their GPT-2 Era
Robots Are in Their GPT-2 Era The Information
- As AI changes pricing, IT firms see uptick in outcome-based deals
AI is pushing Indian IT companies towards outcome-based pricing, with newer deals increasingly linking payments to measurable business results
- Hesai rides on lidar growth as smart cars gain ground despite China’s struggling market
Chinese carmakers’ reinforced push to produce more smart vehicles could sustain growth for hi-tech component suppliers, even as the broader automotive market remains lacklustre, according to Hesai Group chief financial officer Andrew Fan. “Orders for our lidar sensors [this year] were higher than in the same period last year,” Fan told the South China Morning Post on Wednesday. “An increasing percentage of new [car] models are now fitted with lidars as brands make them more intelligent.” Hesai –...
- EPC Announces Mass Production of EPC2370 18 V GaN FET for Next-Generation AI Power Architectures
EPC Announces Mass Production of EPC2370 18 V GaN FET for Next-Generation AI Power Architectures azcentral.com and The Arizona Republic
- How Colossal is incubating an empire of startups
Colossal Biosciences claims to have brought back the dire wolf , and is now hard at work on its quest to resurrect several other species, including the woolly mammoth and the moa, a giant extinct bird. While that’s happening, the company is also pursuing a quieter, more near-term ambition: incubating a cadre of startups that can capitalize on Colossal’s technology and turn it into distinct businesses. At this point, three companies have been publicly spun out of Colossal. The first was Form Bio, which has raised tens of millions to build an AI -enhanced platform meant to accelerate scientific research based on software originally developed inside Colossal. Another is Breaking, a plastic degradation startup that plans to scale the use of a microbe called X-32 . The microbe can break down hydrocarbons, and the hope is to use genetic engineering to accelerate the process. The third is Astromech , founded just last year and fresh off another $20 million raise; it’s building a genetics data platform that trains models on evolutionary and genetic data. The hope is that those models could eventually help predict threats to large biological systems. Examples, the company says, include understanding disease risk, modeling how populations respond to disease and climate change, and identifying at-risk ecosystems. Thus far, Astromech says it has mapped 46 genes related to cellular maintenance, cancer resistance, and longevity. The company has now raised about $60 million in total. Colossal has generated significant press for its quest to invent and resurrect niche species, including the now internet-famous, and very cute, woolly mice and dire wolf facsimiles. It has also generated intrigue for its progress developing new technologies that could take us to even more ambitious biological frontiers. “We will have true dominion over life, where we can eradicate species that are invasive or we can bring back lost species,” Ben Lamm, the company’s CEO, has said . “And I think we’ll also have the ability to engineer plants—not just for food consumption, but you will be able to engineer plants with different types of proteins.” Meanwhile, it’s worth keeping an eye on the network of side businesses Colossal is quietly cultivating. De-extinction captures the headlines, but the company is also moving quickly to plant the seeds of a biosciences empire, one with its hands in a range of biological platforms that could change not only our natural world but also our own health and even the future of our species. These new entities may not go by the name Colossal, but they could end up becoming just as important.
- Google is giving students a free year of AI Pro and a new Gemini Student Hub
Google is giving eligible college students a free year of AI Pro while adding a new Student Hub packed with study tools.
- Gains in AI company stakes juice second-quarter earnings for S&P 500
Gains in AI company stakes juice second-quarter earnings for S&P 500 Reuters
Score: 45🌐 MovesAug 20, 2026https://www.reuters.com/business/gains-ai-company-stakes-juice-second-quarter-earnings-sp-500-2026-08-20/ - Companies are spending trillions on AI. The C-suite doesn’t know who is in charge of it.
Companies are spending trillions on AI. The C-suite doesn’t know who is in charge of it. Fortune
- Perspectives highlight the strength and limitations of AI-powered, “self-driving” labs
Perspectives highlight the strength and limitations of AI-powered, “self-driving” labs EurekAlert!
- New AI model combines physics and observations to reconstruct the history of the earth's mantle
New AI model combines physics and observations to reconstruct the history of the earth's mantle EurekAlert!