OpenAI ex-chief scientist planned for a doomsday bunker for the day when machines become smarter than man
Months before he left OpenAI, Sutskever believed his AI researchers needed to be assured protection once they ultimately achieved their goal of creating artificial general intelligence, or AGI. 'Of course, it's going to be optional whether you want to get into the bunker,' he told his team
If there is one thing that Ilya Sutskever knows, it is the opportunities—and risks—that stem from the advent of artificial intelligence.
An AI safety researcher and one of the top minds in the field, he served for years as the chief scientist of OpenAI. There he had the explicit goal of creating deep learning neural networks so advanced they would one day be able to think and reason just as well as, if not better than, any human.
Artificial general intelligence, or simply AGI, is the official term for that goal. It remains the holy grail for researchers to this day—a chance for mankind at last to give birth to its own sentient lifeform, even if it's only silicon and not carbon based.
But while the rest of humanity debates the pros and cons of a technology even experts struggle to truly grasp, Sutskever was already planning for the day his team would finish first in the race to develop AGI.
According to excerpts published by The Atlantic from a new book called Empire of AI, part of those plans included a doomsday shelter for OpenAI researchers.
'We're definitely going to build a bunker before we release AGI,' Sutskever told his team in 2023, months before he would ultimately leave the company.
Sutskever reasoned his fellow scientists would require protection at that point, since the technology was too powerful for it not to become an object of intense desire for governments globally.
'Of course, it's going to be optional whether you want to get into the bunker,' he assured fellow OpenAI scientists, according to people present at the time.
Written by former Wall Street Journal correspondent Karen Hao and based on dozens of interviews with some 90 current and former company employees either directly involved or with knowledge of events, Empire of AI reveals new information regarding the brief but spectacular coup led to oust of Sam Altman as CEO in November 2023 and what it meant for the company behind ChatGPT.
The book pins much of the responsibility on Sutskever and chief technology officer Mira Murati, whose concerns over Altman's alleged fixation on paying the bills at the cost of transparency were at the heart of the non-profit board sacking him.
The duo apparently acted too slowly, failing to consolidate their power and win key executives over to their cause. Within a week Altman was back in the saddle and soon almost the entire board of the non-profit would be replaced. Sutskever and Murati both left within the span of less than a year.
Neither OpenAI nor Sutskever responded to a request for comment from Fortune out of normal hours. Ms. Murati could not be reached.
Sutskever knows better than most what the awesome capabilities of AI are. Together with renowned researcher and mentor Geoff Hinton, he was part of an elite trio behind the 2012 creation of AlexNet, often dubbed by experts as the Big Bang of AI.
Recruited by Elon Musk personally to join OpenAI three years later, he would go on to lead its efforts to develop AGI.
But the launch of its ChatGPT bot accidentally derailed his plans by unleashing a funding gold rush the safety-minded Sutskever could no longer control and that played into rival Sam Altman's hands, the excerpts state. Much of this has been reported on by Fortune, as well.
Ultimately it lead to the fateful confrontation with Altman and Sutskever's subsequent departure, along with many likeminded OpenAI safety experts aligned with him who worried the company no longer sufficiently cared about aligning AI with the interests of mankind at large.
'There is a group of people—Ilya being one of them—who believe that building AGI will bring about a rapture,' said one researcher Hao quotes, who was present at the time Sustekever revealed his plans for a bunker. 'Literally a rapture.'
This story was originally featured on Fortune.com
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Fox News
23 minutes ago
- Fox News
If you don't know about these video tools, you're already behind
I've said it before, and I'll say it again, AI is changing everything. And now, it's coming for video in a huge way. We're not just talking about deepfakes or filters anymore. This is next-level, movie-magic stuff. Enter to win $500 for you and $500 for your favorite person or charity in our Pay It Forward Sweepstakes. Hurry, ends soon! Let's talk about the wild part first. You don't need any editing software. You don't need a green screen. Heck, you don't even need a camera. Your keyboard is now the director, the producer, the editor, and your creative partner. All you have to do is type what you want to see, and boom, AI turns your words into high-def, studio-quality video clips. It's like having a Hollywood studio tucked inside your laptop. Two of the biggest names in this space right now are Sora from OpenAI and Veo from Google. And yes, you can actually try them both right now. I've tested them side by side, and you can watch my results at the end to see which one came out on top. Sora is built right into ChatGPT and is shockingly good. Type something simple like, "a golden retriever puppy playing on the beach with a tennis ball," and within seconds, Sora delivers a smooth, visually rich video clip. No stock footage. No need to mess with editing timelines or export settings. Just write what you want to see. Here's the breakdown: And here's a hot tip: Microsoft just announced that Sora is being integrated into their new Bing Video Creator, so you'll start seeing it pop up in more places soon. Now over to Google's Veo, and wow, this one is aiming for cinematic quality. Veo gives you up to 60-second videos in 1080p, and the realism is something else. You can include camera moves, lighting shifts, and smooth transitions, just by typing a sentence. For example: "A cat runs across a Paris rooftop at sunset." And it delivers exactly that, complete with lens flares and realistic shadows. To get your hands on Veo, you'll need to sign up for Google's Gemini plans: I wanted to see how each platform handled the same prompt, so I asked both to generate a video based on: "Show me a golden retriever puppy playing on the beach with a tennis ball." Now you can watch the videos side-by-side. My AI puppy Bella stars in both (Veo's on the left, Sora's on the right). My thoughts? Sora nailed the adorable close-up details, Bella's floppy ears, the texture of the fur, the light hitting the waves. But it fumbled a bit when it came to physics. The ball's bounce didn't feel quite right. On the other hand, Veo did a better job with the action. The puppy's movements were more lifelike but it wasn't as sharp on the visual details. We are entering a new age of video creation where your ideas are enough. No gear, no crew, no edits. Just imagination and a little typing. Whether you're a marketer, content creator, educator, or just someone who likes to experiment, this is the future of storytelling. Trust me, it's just getting started. Award-winning host Kim Komando is your secret weapon for navigating tech. Copyright 2025, WestStar Multimedia Entertainment.
TechCrunch
32 minutes ago
- TechCrunch
Why investing in growth-stage AI startups is getting riskier and more complicated
Making a bet on AI startups has never been so exciting — or more risky. Incumbents like OpenAI, Microsoft, and Google are scaling their capabilities fast to swallow many of the offerings of smaller companies. At the same time, new startups are reaching the growth stage much faster than they historically have. But defining 'growth stage' in AI startups is not so cut-and-dried today. Jill Chase, partner at CapitalG, said on stage at TechCrunch AI Sessions that she's seeing more companies that are only a year old, yet have already reached tens of millions in annual recurring revenue and more than $1 billion in valuation. While those companies might be defined as mature due to their valuation and revenue generation, they often lack much of the necessary safety, hiring, and executive infrastructure. 'On one hand, that's really exciting. It represents this brand new trend of extremely fast growth, which is awesome,' Chase said. 'On the other hand, it's a little bit scary because I'm gonna pay at an $X billion valuation for this company that didn't exist 12 months ago, and things are changing so quickly.' 'Who knows who is in a garage somewhere, maybe in this audience somewhere, starting a company that in 12 months will be a lot better than this one I'm investing in that's at $50 million ARR today?,' Chase continued. 'So it's made growth investing a little confusing.' To cut through the noise, Chase said it's important for investors to feel good about the category and the 'ability of the founder to very quickly adapt and see around corners.' She noted that AI coding startup Cursor is a great example of a company that 'jumped on the exact right use case of AI code generation that was available and possible given the technology at the time.' Techcrunch event Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Boston, MA | REGISTER NOW However, Cursor will need to work to maintain its edge. 'There will be, by the end of this year, AI software engineers,' Chase said. 'In that scenario, what Cursor has today is going to be a little less relevant. It is incumbent on the Cursor team to see that future and to think, okay, how do I start building my product so that when those models come out and are much more powerful, the product surface represents those and I can very quickly plug those in and switch into that state of code generation?'
Axios
an hour ago
- Axios
These businesses are big winners from the Musk-Trump breakup
Elon Musk 's breakup with President Trump undercuts his personal empire, but it's good news for a variety of corporate players. Why it matters: Without close ties to the White House, Musk is less powerful politically. State of play: The list of business winners from the Musk-Trump breakup includes: Chinese EV makers: Any sign of trouble for Tesla is good news for businesses like booming EV maker BYD, which already sells more vehicles than Tesla. Musk has cited Chinese EV companies as Tesla's biggest competition. OpenAI: Musk and OpenAI have been feuding over the latter's now-abandoned attempt to turn into a for-profit. The president may now be more inclined to side with the ChatGPT creator and its boss, Sam Altman, who made a U.S. investment announcement earlier this year that Trump touted and Musk questioned. Blue Origin: Jeff Bezos' space company could reap the rewards if NASA seeks alternative partners to replace Musk's SpaceX. Other winners include Tesla short-sellers, who made an estimated $4 billion on Thursday "in one of their biggest single day gains ever," WSJ reported, citing data from S3 Partners. lost myriad contracts, also stand to gain if deal scrutiny diminishes.
