Hello, folks. Welcome to TechCrunch's regular AI newsletter.
This week in AI, OpenAI lost another co-founder.
John Schulman, who was instrumental in the development of ChatGPT, OpenAI's AI-powered chatbot platform, has departed the company for competitor Anthropic. Schulman revealed the announcement on X, stating that his decision was motivated by a desire to focus more on AI alignment — the science of ensuring AI behaves as intended — and do more hands-on technical work.
However, one cannot help but question if Schulman's resignation, which coincides with OpenAI president Greg Brockman's prolonged vacation till the end of the year, was opportunistic.
On the same day Schulman announced his departure, OpenAI said that it intends to change the style of its DevDay event this year, opting for a series of on-the-road developer engagement events rather than a big one-day conference. A representative told TechCrunch that OpenAI would not reveal a new model during DevDay, implying that work on a replacement to the company's current flagship, GPT-4o, is moving slowly. (The delay in Nvidia's Blackwell GPUs could stall the pace much further.)
Could OpenAI be in trouble? Did Schulman notice the writing on the wall? Well, the future for Sam Altman's enterprise is undoubtedly bleaker than it was a year ago.
Ed Zitron, a public relations professional and all-around tech pundit, recently described in his newsletter the several challenges that OpenAI faces on its route to continuing growth. It's a well-researched and extensive post, and I won't do it a disservice by repeating it. But Zitron's points regarding OpenAI's growing pressure to succeed are noteworthy.
OpenAI is expected to lose $5 billion this year. To fund the escalating costs of staff (AI researchers are extremely costly), model training, and model serving at scale, the company will need to raise a large sum of money over the next 12 to 24 months. Microsoft would be the apparent beneficiary; it owns a 49% investment in OpenAI and, despite their occasional competition, works closely with OpenAI's product teams. But, with Microsoft's capital expenditures increasing 75% year on year (to $19 billion) in anticipation of unrealized AI returns, does the company really want to pump untold billions more into a long-term, hazardous bet?
.This reporter would be shocked if OpenAI, the world's most recognized AI business, did not eventually find the funding it requires. However, there is a very real risk that this lifeline will come with less favorable terms, as well as a long-rumored change to the company's profit-cap structure.
To survive, OpenAI will most certainly deviate from its initial aim and venture into unexplored and uncertain territory. Perhaps that was too much for Schulman (and others) to swallow. It's difficult to blame them; as investor and company cynicism grows, the entire AI sector, not just OpenAI, faces a reckoning.