Unsupervised Learning Ep 64: GPT 4.1 Lead at OpenAI Michelle Pokrass: RFT Launch, How OpenAI Improves Its Models & the State of AI Agents Today
394 snips
May 8, 2025 
Michelle Pokrass, a key player in launching GPT-4.1 at OpenAI, delves into the intricacies of AI model development and evaluation. She discusses how user feedback fuels improvements and the evolving challenges in AI training. The conversation highlights the shift towards Reinforcement Fine-Tuning for personalized applications and the importance of balancing conversational abilities with problem-solving as we look towards future models. Her insights reflect a deep understanding of AI's potential and the organizational growth at OpenAI.
AI Snips
Chapters
Transcript
Episode notes
Prioritize Developer-Centric Models
- Focus model development on what developers actually want, not just benchmarks.
- Build evaluation tests based on real API usage and user feedback to improve instruction following.
Gathering Meaningful Feedback
- Gather user feedback by engaging deeply to understand problem themes.
- Use a combination of internal and external users to identify key evaluation areas.
Request for Real-World Evals
- Contribute real-world, long-context, and instruction-following evals to improve models.
- These evals are hard to make but crucial for meaningful model advancements.