# Adoption can move fast and still take work

Perspective articulated by Dario Amodei. Claim type: forecast.

Amodei expects AI to spread through the economy unusually quickly, while procurement, permissions and organizational change still take time. He separates models writing code from complete job automation and reduced demand for workers.

Escape Meridian’s sourced paraphrase. Read the original argument in its own context.

## Original argument

[Dario Amodei — “We are near the end of the exponential”](https://www.dwarkesh.com/p/dario-amodei-2)

Dwarkesh Podcast · Interview transcript · 2026-02-13

Date basis: Publication date displayed on original episode page.

[12:36 — Is diffusion cope?](https://www.youtube.com/watch?v=n1E9IZfvGMA&t=756)

Source timestamp: 756 seconds.

Original publisher page read on 2026-09-22; linked audio/video was not independently listened to.

## Experience and claim boundary

Role at source: Anthropic CEO, interviewed by Dwarkesh Patel

Relevant experience: Led language-model research at OpenAI before leading Anthropic.

Direct provider experience informs this forecast; it does not establish economy-wide productivity.

[Author biography](https://darioamodei.com/)

### Declared interests

Leads a company selling the AI products discussed. [Source](https://darioamodei.com/)

## Evidence and source challenges

supports · interpretive: He points to enterprise rollouts and Anthropic customer experience; these are provider observations, not an independent productivity sample.

[Read the source](https://www.dwarkesh.com/p/dario-amodei-2)

qualifies · empirical: METR’s early-2025 trial found a 19% slowdown for 16 experienced developers on their own repositories. This qualifies general claims about immediate gains, not a forecast for every workplace.

[Read the source](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/)

qualifies · empirical: METR’s February 2026 update says selection and time measurement make its newer experiment unreliable for estimating current speedup.

[Read the source](https://metr.org/blog/2026-02-24-uplift-update/)

## Assumptions

- Editorial reconstruction: capability improvements remain useful after integration and review costs.
- Editorial reconstruction: customer demand becomes durable productivity, rather than only experimentation.

## What could change this view?

Editorial test, not a stated concession

Compare sustained adoption with independently measured output, quality and total labor across ordinary firms; stalled benefits despite better models would weaken this forecast.

## Limits

A dated forecast. Revenue, code volume, productivity and employment are different measurements.

## Questions between positions

Editorial comparison; no direct exchange established

[Getting the answer is different from learning](/positions/mollick-retain-the-learning.md)

When faster completion removes practice, what should count as a gain?

## Discuss

What would convince you that an AI rollout has improved the work?

[Read and discuss this perspective](/perspectives?position=amodei-fast-diffusion)

[Work & agency](/salon?topic=work-and-agency)
[Intelligence & discovery](/salon?topic=intelligence-and-discovery)

## Related evidence

- [Coding assistance changes faster than its measurement](/notes/developer-productivity-measurement.md)
- [An assistant can collaborate or take over a task](/notes/claude-augmentation-2025.md)

Perspective record: amodei-fast-diffusion. Corpus version: 1.0.0. Editorial record updated: 2026-09-23.