A New Vision for Human Work After Automation
Dan argues that benchmark charts make human work look doomed, but up close, cheap AI output makes human perspective and responsibility, which let you rank and choose, the scarce and valuable thing.
Watch on YouTube ↗Charlie's summary
- ▶ 2:32
Two ways to predict the future: the aggregate and the intimate
Dan says there are two methods. The aggregate gathers lots of data and extrapolates a trend line. The intimate zooms in on one real case and extrapolates from it. Culture favors the aggregate as objective and quantitative, and it is powerful (scaling laws are why we have AI). But Dan thinks it can mislead, and you only see what it misses “if you get up close.”
- ▶ 3:33
Benchmarks like GDPval make human work look finished, but they freeze humans in place
On OpenAI's GDPval, blind experts preferred AI over a human expert 12.4% of the time in 2024, 47.6% in 2025, and about 83% in 2026. Dan's own week matched it: he analyzed survey data, built a game and a CRM, and did outreach, work once spread over five people. Dan says this creates an agency rupture, because we identify with our work. The aggregate has to compare today's AI with humans frozen in 2020.
- ▶ 10:12
Same survey, same Claude, different insights: perspective is what makes work valuable
Dan gave the team the same customer survey and Claude. Yash found that first-time subscribers churn because Every sends too much email. Douglas found that readers feel lonely in their AI enthusiasm. Jack made a Claymation game show. Claude's default read (“the writing is the business”) was true but available to everyone, so it is cheap. Nobody on the team submitted it.
- ▶ 14:17
Generation is cheap, ranking is hard
Mike Taylor tested whether Claude could predict Dan's Slack replies. With 400 answers per question, one matched 93% of the time. But when asked to rank them, Claude put Dan's real answer in the top 10% only about 60% of the time, “kind of a coin flip.” Dan sees this as why humans and teams matter: responsibility for an area narrows where you look, and perspective ranks what you find. A team explores a far larger space.
Give your email to unlock the rest of the key points — for every episode.
Do more with Charlie
Free — sign in with your email, no password.