The Directions of Technical Change
Miklós Koren Zsófia L. Bárány Ulrich Wohak CEU, KRTK, CEPR, CESifo CEU, CEPR, HSOE, ROA CEU [email protected] [email protected] [email protected] · 2026
Read The Directions of Technical Change by Miklós Koren Zsófia L. Bárány Ulrich Wohak CEU, KRTK, CEPR, CESifo CEU, CEPR, HSOE, ROA CEU [email protected] [email protected] [email protected] (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-12958.txt.
Description
Generative AI is directional: it excels at some tasks and flops at others, and workers can meet job requirements with different task mixes. Miklos Koren and two co-authors model AI adoption in high dimensions, where a worker adopts a tool only when output rises. Both worker and tool perform tasks described by convex production sets, and because the tool needs supervision from the worker's own time, adoption resembles hiring a coworker.
The key statistics are the worker's pre-AI shadow values, the output gain from relaxing each task direction slightly, which need not match observed activity. As AI capability grows, adopted directions expand in a cone around those prices. Near the entry threshold, small capability gains trigger large extensive expansions, while between entry and all-in thresholds, use stays partial. The flexible model nests most existing task-based accounts of technical change.
Catalog record
| CREATOR | Miklós Koren Zsófia L. Bárány Ulrich Wohak CEU, KRTK, CEPR, CESifo CEU, CEPR, HSOE, ROA CEU [email protected] [email protected] [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-12958