Is OpenAI in a Death Spiral?

Hey Everyone,

As always for the best reading experience read this article on a web browser. [You can always do this by clicking on the title].

To continue yesterday’s topic a bit, I noticed some odd moves in talent departures from OpenAI that amount to them losing out in the AI war for talent to Anthropic and Google.

A major Talent exodus is taking place at ChatGPT maker OpenAI. In the last 24 hours:

This is striking at the core-team of OpenAI:

• Cofounder John Schulman is heading to rival Anthropic

• Cofounder Greg Brockman is taking a sabbatical

• Product leader Peter Deng is also departing

Greg Brockman

I’m taking a sabbatical through end of year. First time to relax since co-founding OpenAI 9 years ago. The mission is far from complete; we still have a safe AGI to build. - Source.

John Schulman

I shared the following note with my OpenAI colleagues today: I've made the difficult decision to leave OpenAI. This choice stems from my desire to deepen my focus on AI alignment, and to start a new chapter of my career where I can return to hands-on technical work. I've decided to pursue this goal at Anthropic, where I believe I can gain new perspectives and do research alongside people deeply engaged with the topics I'm most interested in.

To be clear, I'm not leaving due to lack of support for alignment research at OpenAI. On the contrary, company leaders have been very committed to investing in this area. My decision is a personal one, based on how I want to focus my efforts in the next phase of my career.

I joined OpenAI almost 9 years ago as part of the founding team after grad school. It's the first and only company where I've ever worked, other than an internship. It's also been quite a lot of fun. I'm grateful to Sam and Greg for recruiting me back at the beginning, and Mira and Bob for putting a lot of faith in me, bringing great opportunities and helping me successfully navigate various challenges. I'm proud of what we've all achieved together at OpenAI; building an unusual and unprecedented company with a public benefit mission.

I am confident that OpenAI and the teams I was part of will continue to thrive without me. Post-training is in good hands and has a deep bench of amazing talent. I get too much credit for ChatGPT -- Barret has done an incredible job building the team into the incredibly competent operation it is now, with Liam, Luke, and others. I've been heartened to see the alignment team coming together with some promising projects.

With leadership from Mia, Boaz and others, I believe the team is in very capable hands. I'm incredibly grateful for the opportunity to participate in such an important part of history and I'm proud of what we've achieved together. I'll still be rooting for you all, even while working elsewhere. - Source.

Upgrade to Premium

OpenAI has bled talent due to its internal schisms before, and those that aren’t with Sam Altman’s leadership style and focus have been shown the door, but this time it feels different. There are significant financial difficulties and a poor probability path to profitability this time. Google and Anthropic’s rivals have ‘caught up’ and OpenAI is no longer seen as some ‘generational company’ in spite of being far more revenue and product driven than its peers.

OpenAI’s Culture and Priority Shift

“The fundamental problem with Open AI is that it rapidly transformed from a premium AGI research company into a product company.” - Bindu Reddy.

Now that Microsoft has signaled OpenAI to be a direct competitor, things are about to change in the talent wars. This was signaled by Google’s $2.5 Billion deal for access to Character.AI’s models and executives recently.

That many of OpenAI’s best people are now working at Google would have been unthinkable even just two years ago. Anthropic’s more Enterprise customer focus and alignment priority just appeals more to deep learning researchers and the highest level machine learning engineers in Generative AI.

I’m not sure how Chinese AI startups, xAI, Cohere or Mistral and their peers compete in such a battle for the world’s best talent. While OpenAI pays well, their internal culture is full of drama and uncertainty due to their gambling on AGI. The snake-oil salesmanship of Sam Altman is also being criticized, hindering more than its helping OpenAI’s progress. He was a great CEO for the first few years of OpenAI when hype was good, but now as it enters a next phase of revenue generation, they badly need a better aligned professional CEO. But he is buddies with the Board so that’s unlikely to happen anytime soon. Talent can see the writing on the wall as their shares will get even more diluted. Google has significantly improved its competitiveness in the talent wars in 2024.

This is not the first time we’ve worried about an OpenAI exodus. There have been many waves, most notably nearly its entire superalignment team, which reduced its credibility as a serious research lab. At that time we didn’t know how massively wasteful financially OpenAI was. They likely literally couldn’t afford to do what they had promised in superalignment. They simply didn’t have the resources. Which each star Researcher that leaves OpenAI, a bit of the magic is gone. All that we are left with is a product called ChatGPT, and diminishing returns of a plethora of new products, and new features for those products.

Ilya Sutskever

2.5 months ago:

“After almost a decade, I have made the decision to leave OpenAI.  The company’s trajectory has been nothing short of miraculous, and I’m confident that OpenAI will build AGI that is both safe and beneficial under the leadership of Sam Altman, Greg Brockman, Mira Murati, etc… It was an honor and a privilege to have worked together, and I will miss everyone dearly.   So long, and thanks for everything. I am excited for what comes next — a project that is very personally meaningful to me about which I will share details in due time.” - Source.

Sam Altman’s strategy is clearly to have more junior employees who are like ‘little lieutenants’ he can groom, that you can pay less and who ask fewer serious questions. They do your bidding, and Microsoft’s, to build products - with the sole goal of revenue generation.

Sam Altman guided his PR/comms team to a ‘theatrical approach’ to public relations, and boy has it backfired. It’s painful to watch.

Peter Deng, a product manager who joined OpenAI last year after leading products at Meta, Uber, and Airtable, also exited some time ago, the company confirmed. On his LinkedIn two months ago he introduced OpenAI for Non-profits. Radio silence on his Twitter account.

But former OpenAI employees aren’t just joining Google, they are joining Anthropic, OpenAI’s main direct competitor and a spin-off from OpenAI itself a few years ago. It’s one thing for Ilya and Andrej to start their own companies, like we assumed they might. But John Schulman and more serious alignment focused people joining Anthropic is concerning if you are an OpenAI fan. OpenAI’s once insurmountable first-mover advantage has faded in 2024, badly.

How OpenAI has spent its supposed $13 Bn investment from Microsoft is unclear at best, and wasteful at worst. It’s been a glorified and bought research lab for Microsoft and some of that Copilot Era enthusiasm there is not meeting the requirements of Copilot customers. Google is badly outpacing Microsoft now in Generative AI products. Even as they plundered Inflection AI for talent. The talent wars is also hurting how American AI firms and startups can actually innovate. As Generative AI startups are already heading for the exits, failing, or consolidating - it’s a gigantic mess where revenue is not on the other side of the Generative AI rainbow 🌈.

Nobody can convince me Apple, Amazon or Microsoft are destined to be winners in Generative AI. OpenAI’s deal with Microsoft likely locked them into a death spiral and had Sam Altman been more experienced, he would have realized this. But I think he felt there was no other choice. As time would have it, Elon Musk was only to be among the first rival to be kicked out of the core team, the first of many.

Elon Musk’s New Attack on OpenAI

“Musk’s new lawsuit takes aim at OpenAI’s deal with Microsoft and the definition of artificial general intelligence.” - The Verge.

Read the Lawsuit

A new lawsuit has appeared in August, 2024. Elon Musk filed a federal lawsuit against Sam Altman alleging that the Tesla boss was “manipulated” into co-founding OpenAI.

Adn this is all super relevant to how OpenAI became what is it today. Musk — who co-founded OpenAI in 2015 — sued the company in February, accusing the ChatGPT maker of abandoning its original nonprofit mission by reserving some of its most advanced AI technology for private customers.

Musk claims he invested in OpenAI on the basis it would be a nonprofit, but that later, Altman and others, along with Microsoft, “established an opaque web of for-profit OpenAI affiliates, engaged in rampant self-dealing.” What I found most interesting about the lawsuit is what Elon Musk has to say about the AGI snake-oil narrative. “The perfidy and deceit is of Shakespearean proportions,” the suit adds.

Musk says OpenAI “flipped the script” after its technology approached Artificial General Intelligence (AGI), a type of technology that matches or surpasses human capabilities.

“OpenAI’s focus shifted from its advertised charitable purpose—to benefit the public and protect humanity—to a vehicle for Altman and his partners’ self-enrichment. This is most clearly evidenced in OpenAI’s partnership with Microsoft and the proliferation of an opaque web of for-profit OpenAI affiliates, recently valued at a whopping $100 billion,” the suit reads.

“This is a much more forceful lawsuit,” Musk’s lawyer, Marc Toberoff, said to The New York Times. The new suit argues that OpenAI broke federal racketeering laws in a conspiracy to defraud Musk and that its contract with Microsoft would revoke the tech giant’s rights to OpenAI’s technology once artificial general intelligence (AGI) had been achieved.

When Sam Altman is fighting other tycoons, these are battles and lawsuits he doesn’t need. This is not the time to push the gas on the “AGI” narrative. OpenAI’s PR/Comms team really have discredited “AI” to wide swaths of the general public, lowering trust in these systems significantly, even as students and young workers feel pressured into using these products.

Microsoft is worried about SearchGPT

Microsoft will force OpenAI to make SearchGPT a more minor product due to it using elements of Bing. Microsoft’s Bing already makes a fair amount of revenue in Search Ads.

Microsoft called OpenAI a “competitor” for the first time in an SEC filing recently. On one hand Microsoft wants SearchGPT to be the next cash-cow for them earning 75% of the profits, but on the other they are concerned that OpenAI is outcompeting them for shared spend of customers in the Copilot Era. This all makes both Microsoft and OpenAI look particularly bad.

There aren’t just conflicts of interest, but very poor revenue generating potential in these sort of a situations with conflicting incentives and legal headaches. It would be easier just to tank the startup and acquire what is left of it. A bit like Google refusing to fund Character AI (when they were expected to do so), then making an alternative “licensing” deal for its model. The unfortunate thing is how BigTech controls the future products and talent in the Generative AI space, will drastically limit its ability for real innovation in the space.

Then it’s not just a question of how OpenAI Survives, but about how Microsoft and OpenAI’s first mover advantage is fading fast. But Anthropic for all of its high moral ground, is mostly just funded by Amazon and Google, major Cloud rivals of Microsoft. So essentially all these people are in bed together, and it doesn’t matter which side of the bed they sleep or wake up on. The results are going to be the same at the end of the day.

OpenAI on the Brink of Bankruptcy

You don’t take a sabbatical and OpenAI’s most critical juncture if you are confident. OpenAI might need another round of funding to remain afloat. Greg Brockman quit for a few days when Sam Altman was fired, but now it seems he is quitting when OpenAI needs him the most.

The startup spends $7 billion on training its AI models and $1.5 billion on staffing. It has drastically overpaid talent compared to its revenue potential. OpenAI’s $80 Billion valuation is already unrealistic, but as they get more funding it will dilute shares to an unimaginable degree. Who would want to work there in such a financial situation? You’d have to be fairly young in your career or have a major stake.

If you are really on the brink of bankruptcy singing the AGI tune is probably not a great idea. But Sam Altman is not a professional CEO, he’s scattered across a host of Venture Capital projects, billionaire networks and spinning PR/Comms fairytales. Even Brockman likely realized this.

Greg Brockman has titles like President, Chairman and Cofounder of OpenAI on his LinkedIn. Why would he be taking a sabbatical now? He’s rich enough he can do whatever he wants, but it just signals to the world that OpenAI is a sinking ship. What does Greg know that we do not know yet? This means that the battle over who will own ChatGPT has likely already begun.

Mira Murati as the young Sheryl Sandberg of OpenAI’s last few months (or years) 🐞

The person you have to feel bad for in all of this is Mira Murati. She’s classy, capable and seems above reproach. With all the chaos of being an executive at OpenAI, Mira is the odd person out who has among the most challenging of jobs (and doesn’t give up).

But you cannot tame someone like Sam Altman. Sam is a misfit manipulator by personality type. Raised in St Louis, Missouri, Sam Altman was the Stanford dropout who had become the president of the massively successful Y Combinator startup incubator before he was 30.

He’s a venture capitalist at heart, able to project the most genuine concern around his favorite and GPT-4 inspired talking points.

“All through the summer of 2023, Altman was treated like a Beatle, stopping by DC as part of a world tour, meeting prime ministers and presidents around the globe. US Senator Kyrsten Sinema gushed: “I’ve never met anyone as smart as Sam… He’s an introvert and shy and humble… But… very good at forming relationships with people on the Hill and… can help folks in government understand AI.” Glowing portraits at the time painted the youthful Altman as sincere, talented, rich and interested in nothing more than fostering humanity.” - Source.

But how will he be remembered? What will be his legacy? Besides the person who imploded the maker of ChatGPT?

Leave a comment

Is OpenAI in a Death Spiral? 💀

Sorry Ilya, Andrej and sorry to you Mira, you all have your reasons. This talent exodus at OpenAI speaks of a terrible uncertainty at the startup. Of course the outcome has also been predicted by many of us analysts and AI watchers. The death spiral at OpenAI has begun.

Loading...

A startup isn’t just a product. It’s compromised by the team that gave it life.

Credit: Linus.

At least Ilya has SSI now. Whether you agree or disagree with me on my speculations here or my characterization of OpenAI, the company is already a shadow of what it once was. It barely matters how good GPT-5 or GPT-6 are in such a sad and brutal state of affairs. The AI startup is widely overvalued and it can’t end well, and everybody knows it.

Thanks for reading!