Agents, Robotics, and the Fate of Jobs- OpenAI's former Chief of Research weighs in

A few days ago, OpenAI 's former Chief Research Officer, Bob McGrew, sat down with Y Combinator's Garry Tan to chat about AI’s future—covering everything from AI agents to robotics to the fate of jobs.

Here are 3 big takeaways from their 30-minute interview. Welcome to this week's What's Up Wednesday! 🚀

1. Reasoning models will make AI agents actually work.
   We've had AI agents for a while, but let’s be honest—they were often unreliable and unpredictable. McGrew argues that reasoning models are the missing piece. After all, what are agents if not AI that can logically break down and iterate towards a goal?
2. Robotics are where LLMs were five years ago.
   McGrew believes robotics is the next big thing. Just like LLMs had their "GPT moment," robotics could have its own major breakthrough within the next 5 years. The future might be closer than we think.
3. The future of jobs? 🤷‍♂️
   McGrew, like many of us, isn’t entirely sure. But he predicts a world with two dominant job types:

• The Lone Genius— leveraging AI for independent research and creative work.
• The Manager— orchestrating AI agents to execute tasks at scale.

Bonus: Should you still learn to code?
McGrew says yes— and he's even using LLMs to craft programming lessons for his 8-year-old son. He believes hands-on coding helps develop taste and a deeper understanding of what's possible.

What do you think? Do McGrew’s predictions resonate with you, or do you disagree? Drop a comment below! ⬇️