My partner @leiysky said:
> Once you understand the boundaries of what agents can do, they’re incredibly liberating. You can focus your own energy on learning and leave the experiments to the agent.
I’ve always had more ideas than I could implement. Each one used to require an exclusive afternoon or a dedicated week to research, prototype, wrestle with tradeoffs, and handle the work I enjoy least: benchmarking, bisecting, etc.
Now I’ve essentially shaped GPT into an extension of myself. It operates within the boundaries of what I already know how to do, while bringing AI’s endless patience for debugging and testing.
That means I can delegate much of the execution—and that’s where the productivity gain comes from.
Agents don’t replace learning or judgment. They make experimentation much more cheaper.
That's right, GPT-5.6 Sol is awesome and can be used pretty much anywhere, including in the CC harness.
To celebrate this, together with the fact that I'm not going anywhere... I have reset usage limits for all paid users of ChatGPT Work and Codex.
Have fun out there!
Here’s my resume:
- B+ in Science
- JV soccer team (2 years)
- Can eat >10 Oreos in one sitting
- Owner of several Dogecoins
- Can burp the alphabet
- Can run fast (top 25% of class)
- Spend majority of my life on here
- Been to 23/50 states
- Born in the USA