OpenAI Leaders Backed Full-Duplex GPT-Live Models
OpenAI researcher shares experience of sustained support for voice models despite initial skepticism.
OpenAI researcher Kundan Kumar posted that leadership backed full-duplex GPT-Live models for his 14 months at the company even when success looked deeply improbable. He called the approach typical for research projects there and noted few environments would sustain such bets. Greg Brockman quoted the post and said OpenAI has built the muscle of making long-term research bets. Replies on X praised the sustained conviction and rare willingness to back uncertain high-reward work.
In my ~14 months at @OpenAI, one of the most surprising and genuinely wonderful things about it, has been the full leadership support for full-duplex models (gpt-live series), even when it seemed deeply improbable that they would work. And yet, if they did work, it was obvious…
Reactions from ranked influencers
1 postOpenAI has built the muscle of making long-term research bets:
In my ~14 months at @OpenAI, one of the most surprising and genuinely wonderful things about it, has been the full leadership support for full-duplex models (gpt-live series), even when it seemed deeply improbable that they would work. And yet, if they did work, it was obvious…
