Jev founder's route replicated by his former employer two years after leaving
The post reflects on Diogo Almeida leaving OpenAI to build Jev, which took off, after which OpenAI released the Decisions API and Dots covering similar ground.
imagine being Diogo Almeida... you leave OpenAI. spend 2 years building a completely different kind of model. launch Jev. within DAYS: > ~40M views > developers everywhere start integrating it > Vercel + Cloudflare add support > investors are reportedly talking 10B+ valuations then 2 weeks later... your old company launches the Decisions API. a specialized model for: > classification > routing > choosing an agent's next action > returning probabilities instead of paragraphs sound familiar? lol. and OpenAI basically did the same thing with Dots today. Muse / OpenClaw / Grok Bot prove people REALLY want always-on personal agents... then OpenAI launches Dots. this is such a weird part of building in AI right now. you can spend years finding something nobody else is doing... prove there's a massive market for it... and within weeks the biggest labs on earth can replicate it.