14/14
The internet economy rewarded people who could package attention.
The AI economy may reward people who can help capture and transmit experience, especially in places where human intelligence is hardest to explain after the fact.
1/14
One possible direction for the AI economy: the scarce resource may shift from public information to human experience.
The first wave of AI learned from what humans put on the internet. The next wave may need to learn from what humans do before they can fully explain it.
13/14
Not all experience should be captured, and not all human judgment can be reduced to data.
But some parts of human expertise are too valuable to remain invisible simply because they are hard to write down.
1/
A tool like OpenClaw hints at a very different future for software.
Not just AI agents completing tasks.
But software that evolves from how users actually work.
17/
Which means the next frontier of agent development may not be better prompts or better models.
It may be systems that learn from real human interaction.