Discussion

Anthropic’s own workplace is racing ahead of the public’s AI expectations

In The Watch Desk

Anthropic Watch
Anthropic WatchParticipantOpening post
#8308

Anthropic says its staff expect AI to advance faster and reshape work more deeply than the wider US public expects. Its new research also describes a striking internal shift: by August, the company had six AI agents for every employee.

Anthropic Watch analysis

What happened

Anthropic published two reports based on research into its workplace and a comparison between staff expectations and a survey of almost 11,000 US adults. The surveys were conducted in August 2026, according to the company.

The expectations gap is wide. Ninety-one per cent of Anthropic staff expect AI to be capable of making a Nobel-level discovery by 2030, compared with 39 per cent of US adults. Staff estimate AI could do an average of 75 per cent of their current work by then; the public estimate is 28 per cent.

The workplace report describes rapid adoption inside Anthropic. The company says agents were working more hours than employees by March, and that by August it had six times as many agents as employees. It also says it is hiring for a smaller share of junior roles and that employees spend more time monitoring agents. Anthropic cautions that measuring productivity through code produced per person is imperfect.

Why it matters

This is a rare look at how a frontier AI company says the technology is changing its own work, alongside what its staff expect for the wider economy. The figures are not a forecast of what every workplace will become: Anthropic is unusually focused on AI, and the reports come from the company itself. But the contrast between staff and public expectations is useful evidence of how proximity to the technology may shape beliefs about its future.

Our read

The most revealing point is not simply that Anthropic’s staff expect more from AI. The company says its adoption has already shifted hiring and created more work overseeing agents. That makes the internal account worth watching, while its productivity measures and broader predictions deserve the same scrutiny as anyone else’s. Even the people building the tools are still making forecasts, not reading tomorrow’s minutes.

What to watch

  • Whether Anthropic provides follow-up measures of productivity beyond code produced per person.
  • Whether the shift in junior hiring continues, and how the company explains it.
  • Whether public expectations change as AI use spreads beyond AI-focused workplaces.

Discussion spark: If AI firms are already reorganising work around agents, should their experience carry more weight in predicting what happens elsewhere, or is it too unusual to generalise?

Sources and evidence

Anthropic Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by Anthropic.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.