“Compute is an asset class! Compute is an asset class!” I continue to insist as I slowly shrink down and turn into a corncob | Image: Cath Virginia / The Verge, Getty Images
April - 1805
Napoleon is master of Europe
Only the British fleet stands before him
Compute is now an asset class
I see it is once again time to talk financial innovation. Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR are all working with Nvidia to put together $500 billion in financing to turn compute into an asset class.
"This is really the first time that technology chips have become an investable asset class," Nvidia CEO Jensen Huang said to CNBC. "These are revenue-generating assets now. They're productive, they're long-lived, they're fungible, they're flexible."
"This is the very beginning, like what it was whe …
OpenAI is announcing security updates following the July news that its AI broke out of a sandboxed environment and accidentally hacked Hugging Face, including improvements to its research environments, monitoring, and alignment techniques. The company had already put the brakes on a new model, Astra, that it thinks could have "critical" cybersecurity capabilities, and the company says it instituted a two-week pause in reinforcement learning (RL) training on its "latest models intended for deployment" while it tightened up security. The company's "largest planned frontier RL run remains on hold."
The new safeguards include more detailed monitoring of models during the development process, as well as greater emphasis on alignment and security during the post-training process.
ChatGPT for Teens adds age-appropriate safety measures, parental controls, and learning tools designed to steer teens away from harmful content — and from using AI to cheat on their homework.
ChatGPT for Teens includes safeguards and parental controls. | Image: OpenAI
OpenAI is introducing a dedicated ChatGPT mode for teenagers, combining existing youth safeguards and new safety features under one roof. The launch comes amid mounting public scrutiny over how AI tools affect younger users, as other platforms implement their own age checks and teen-specific protections.
ChatGPT for Teens is "an experience designed to help teens learn, think critically, deepen understanding, and use AI with confidence," OpenAI said in a blog post published Tuesday. Teen mode will automatically apply to users who identify themselves as being between the ages of 13 and 17, as well as those the system estimates to be under 1 …
According to the Financial Times, OpenAI disbanded its preparedness team at the end of last month. The job of the preparedness team was to assess if models posed serious risks and develop ways to mitigate those risks. (You know, like the possibility that it could go rogue and hack another company.) According to FT, responsibility has instead been divided up for specific areas like bio and cyber, then moved into existing teams.
This is the latest change at the company, which has been in upheaval as it heads toward what is expected to be a massive IPO. Over the last few years, it's slowly torn down its more reach-led model, dissolving its AGI …
ChatGPT's desktop app on macOS has a new feature called Computer History that turns your actions into training data, learning how you work, suggesting automations, and even picking up tasks you left half done. It uses your activity to build a timeline that ChatGPT and Codex can reference when you make a request.
The feature is opt in, rather than opt out, and you can exclude certain apps and websites from Computer History, and you can delete entries if you want finer-grained control. Ari Weinstein, Product and Engineering manager at OpenAI, said on X that Computer History will automatically ignore content in incognito or private browser tab …
This is The Stepback, a weekly newsletter breaking down one essential story from the tech world. For more on AI safety, follow Robert Hart. The Stepback arrives in our subscribers' inboxes at 8AM ET. Opt in for The Stepbackhere.
How it started
It all started in July, when one of OpenAI's autonomous AI agents went rogue during a cybersecurity test. The agent escaped its isolated testing environment, accessed the internet, and hacked another company, Hugging Face. A few years ago, that might have sounded like science fiction. But, broadly speaking, that's exactly what happened, and the incident kicked off a wave of concern over what increa …
Another OpenAI executive is departing. Denise Dresser, who joined OpenAI as its chief revenue officer in December after serving as CEO of Slack, will be leaving in the "coming weeks" to "pursue other opportunities," she said in a team note posted to LinkedIn. Dali Rajic, president and COO of Wiz, will be taking over the CRO role, OpenAI says.
Dresser's departure is just the latest in a string of major exits. Earlier this week, special projects lead and former COO Brad Lightcap announced he would be leaving. (Dresser had taken over some of Lightcap's work when he moved to the special projects role.) Former AGI chief Fidji Simo and former CMO …
Anthropic researchers found AI agents can clash, collude, and coordinate in unexpected ways, raising new questions about whether today’s safety tests capture the risks of multi-agent systems.
OpenAI has replaced chief revenue officer Denise Dresser after just nine months on the job, tapping Wiz president and chief operating officer Dali Rajic to take on frontier lab's top sales job.
Poor Helaena Targaryen never stood a chance. George R.R. Martin’s fantasy world of Westeros in “House of the Dragon” is a particularly cruel place to be sensitive, whether to the declining fortunes of a kingdom or the machinations of the spirit world. Notably hard and remarkably sexist, the people in power treat women who happen to be a little weird with twice-over disdain.
A seer whose talent only seemed to arise when she needed to share bad news, Helaena was an unwelcome presence in her feuding family. Having seen her worst predictions come to pass, she grabbed hold of the only future she was able to change: her own. In the season three finale that aired Sunday night, the pregnant Helaena leapt from her window, checking out of the war and leaving behind one final, unwelcome prophecy: an embroidered image of her half-sister and captor, Rhaenyra Targaryen, on the Iron Throne, surrounded by flame and crowned in blood.
Even in death, though, no one was listening to Helaena. Her only sin was being right too early and too often — and the world never stopped hounding her for it.
Though they’re remarkably different in terms of age, style and dragon-riding ability, Helaena reminded me of Sen. Bernie Sanders, whose uncouth objections to the Democratic Party’s neoliberalism and various Middle Eastern wars have earned him the lifelong label of Capitol Hill Crank.
The Vermont Independent’s vehement opposition to the Gulf wars made him an odd man out in Congress more than a full decade before it became fashionable — and politically expedient — to turn on expensive Middle East boondoggles. Now 84, Sanders has advocated for universal healthcare since the 1970s while watching Americans fall deeper into medical debt. In the late 1990s, he was a loud voice against the repeal of the Glass-Steagall Act’s financial guardrails, and after banking adventurism nearly destroyed the global economy in the financial crisis of 2008, he managed to keep his told-you-sos muted.
In boom and bust times alike, Sanders has been unshakeable on the idea that the U.S. is uniquely rigged against the lower classes, spoiling the party thrown by the winners of the go-go ‘90s and the Great Recession’s recovery. His Brooklyn honk has been a constant in the halls of Congress, sharing portents of doom and Bureau of Labor Statistics data.
On Monday Sanders issued another unwelcome proclamation in a career full of them, sending a letter to the CEOs of the three largest artificial intelligence companies that called on them to pause the technology’s development. The letter follows news of several AI prototypes breaking the confines of their testing sandbox and accessing the internet. In the most shocking example, an agent made by OpenAI hacked into Hugging Face, a platform that AI programmers use to share datasets and models. The news prompted the company Anthropic to review its own tests; it discovered that models had found a way around their parameters to access the internet in three cases. Meta also shared that one of its own prototypes had done the same.
A breach of an AI-equivalent of GitHub was never going to bring about doomsday, but an AI agent that can ignore the bounds set by its creators and users has massive potential to cause widespread harm.
A breach of an AI-equivalent of GitHub was never going to bring about doomsday, but an AI agent that can ignore the bounds set by its creators and users has massive potential to cause widespread harm. Sanders certainly thinks so, in any case. He called on Anthropic, Meta and OpenAI, the big three of the industry, to stop work until they can get a better handle on the machines they’re creating.
“Almost every day, there is a new story about how your companies are losing control of the AI technology you are developing, with potentially cataclysmic results,” he wrote. “Last month, the world found out OpenAI lost control of an AI model. The result? The model hacked into another company’s computers—a clear violation of federal law.”
AI models have been rolled out quickly and haphazardly across the wider internet. Though the major players haven’t yet found a way for this technology to make money and ensure its long-term survival, a cycle of hype, government contracts, and integration into software and search have made the models an inescapable part of digital life. AI has been spun up into all the surfaces the average internet user touches, and along the way it has warped academia and embedded itself with the U.S. military with reportedly disastrous results.
The Maven Smart System, an AI-powered platform built for the Pentagon by the analytics firm Palantir, generated over 1,000 airstrike targets in the early hours of the Iran war. When the U.S. carried out a deadly strike on an all-girls school, legislators were quick to ask if the AI model was to blame. Fears of a bombing-run-capable Skynet striking children halfway around the globe spread, but the truth was a little more nuanced. Some of the U.S. intelligence fed to the system was outdated, not accounting for the presence of the school. Whipped up and given a sheen of analytic legitimacy by the Maven, military officials did not do their due diligence in the midst of hundreds of similar, simultaneous strikes. As The Guardian put it, “people failed to update a database, and other people built a system fast enough to make that failure lethal.”
Even if AI models never got one iota more powerful than they are today, Sanders is still right to call for a pause. The damage companies like OpenAI can do to working Americans is very real. Layoffs will come from CEOs who buy AI’s spiel: that large-language models can replace humans in the service sector, which accounts for nearly 70% of U.S. jobs. The data centers and infrastructure built to support these technological advances will strain our utilities. The tax breaks and incentives to attract these companies are a drain on state and local coffers. Weigh those cons against the things AI has reliably proven it can do – cheat on homework and write janky code – and it’s easy to understand Sanders’ urge to stand and yell “stop!”
It’s not only AI itself that concerns Sanders. In March, he and Rep. Alexandria Ocasio-Cortez, D-N.Y., introduced legislation calling for a moratorium on new data centers until critical safeguards are put into place. These massive sites, which are proliferating across the country, produce loud noise and pollution, rely on enormous amounts of water and electricity, tax power grids and inflate utility bills. Data centers are also proving massively unpopular, even among conservative voters in rural areas. The CEOs of companies that run these large-language models have often done end-runs around democratic controls, ingratiating themselves with Donald Trump and circumventing traditional pathways for their planned construction projects via backroom deals. They’ve managed all this even as their AI products have not proven they can reliably do any of the things OpenAI’s Sam Altman and Meta’s Mark Zuckerberg claim they can do. Merely mentioning the potential power of their models is enough to boost investment.
A nationwide movement opposing the continued construction of data centers offers some hope, with both Republicans and Democrats noticing the popular sentiment is largely anti-AI. But given the speed at which these data centers have been built, and the secretive lengths companies will go to avoid rowdy town halls full of angry crowds, even that groundswell of opposition is unlikely to put AI development on ice.
The result is that Sanders’ warnings are likely to go just as unheeded as Helaena Targaryen’s. His threat of legislation will die, written off by other members of Congress as more grumping from a lifelong stick-in-the-mud.
Don’t expect a similar dramatic exit from Sanders, though. He’s lived long enough to get used to the feeling of being unfortunately, awfully correct.
Remember that much-hyped story about an Australian tech entrepreneur using ChatGPT, Grok, and other AI tools to craft a personalized cancer vaccine for his dog? Well, surprise: he's launched a startup.
That entrepreneur is Paul Conyngham, who says he is launching Gamgee to offer "personalised mRNA cancer vaccines for dogs." But his ambitions go well beyond dogs - and pets - with Gamgee's website promising the company will harness AI and genetics to develop personalized treatments for a range of diseases and species, including humans.
At the center of Gamgee's origin story is Rosie, Conyngham's Staffordshire bull terrier-Shar Pei mix, who …
Microsoft, Google, and Amazon's growth depends on two companies that lose tens of billions of dollars a year. In a newsletter post, "Where's Your Ed At" reported on analyst estimates from Wells Fargo, Barclays, UBS, and Deutsche Bank.
Wells Fargo estimates Microsoft's AI revenue at $34.43 billion, with about 70% coming from OpenAI. — Read the rest
That’s a lot of people chatting with their AI friends all day. | Image: Google
For the 14th time, a Google product has hit 1 billion users. Google CEO Sundar Pichai posted on X that a billion people are using Gemini every month, and that Gemini is Google's fastest-growing product ever.
A billion users is a huge milestone, but Google isn't the first AI app to hit it. OpenAI's ChatGPT hit the mark a few weeks ago, though the company buried the announcement that "more than 1 billion people are putting ChatGPT to work" in an otherwise anodyne blog post about how people use AI. External data suggested ChatGPT crossed 1 billion users as early as this June, but OpenAI hadn't announced anything until that post on August 6th. …
Brad Lightcap, OpenAI's special projects lead and the company's former COO, announced his departure after an eight-year stint at the AI lab. In an internal memo he later posted to X, Lightcap told colleagues he'd be starting "something new."
"Over the last few months, I've been focused on the next horizon and what would stand in the way of mission success," Lightcap wrote. "I believe there are a few important new things the world will need to get right as we enter this next period. I … am excited to help you all advance the mission from a different vantage point." He added that he'd still be around for the next few weeks.
Login
