Anthropic publishes an interactive model of how AI could reshape the US economy by 2030
The Econ Scenario Explorer from Anthropic's economics team models three futures, from a modest +1.6% of GDP to an extreme +32.4%, and compares them with the expectations of 10,980 surveyed Americans.
Anthropic's economics team has released the Econ Scenario Explorer, an interactive model of how artificial intelligence may affect growth, jobs and wages in the United States by 2030. The AI Hub Azerbaijan LinkedIn page shared the release this week.
How the model works
The model breaks jobs into bundles of tasks. For each task, AI can help a person do it faster, do it autonomously, leave it untouched or create new tasks. Users enter their own assumptions about capability and adoption, and the tool projects GDP, wages, labour share and unemployment.
Three scenarios
| Scenario | AI's role by 2030 | GDP vs. no-AI path |
|---|---|---|
| Modest | Roughly the impact the internet had | +1.6% |
| Substantial | AI can do about half of knowledge work, mostly autonomously; unemployment around 5% | +8.3% |
| Extreme | AI more productive than humans at most knowledge tasks; annual growth reaches 15%, unemployment beyond recession levels | +32.4% |
In the extreme scenario the labour share of income falls to 45.2%, with 54.8% going to capital.
What the public expects
Anthropic surveyed 10,980 US adults. The median respondent's answers produce an outcome close to the "substantial" scenario. The economy grows in all three scenarios; the difference is how the gains are shared.
The authors note that, like all economic models, this one simplifies reality. Its value is less in the point estimates than in making assumptions explicit. That is a useful discipline for any government or company writing an AI strategy: state what you expect AI to do to tasks, then check the numbers that follow.
