READING · SELECTION
An evolving timeline of research, signals and experiments for tracking how work, organisations and the economy are changing.
EVOLVING COLLECTION · 2025—
7 texts shown
01 / 07
ESSAY · POLITICAL ECONOMY
The central argument is that if AI eventually replaces human labour at scale, the issue goes beyond unemployment. People could gradually lose economic value and therefore political and social bargaining power, while capital and power concentrate among those who control AI systems.
02 / 07
POLICY ESSAY · ANTICIPATION
Amodei argues that AI capabilities are advancing much faster than political institutions. He calls for anticipating consequences for security, employment, taxation, civil liberties and geopolitics instead of waiting until problems become obvious.
03 / 07
RESEARCH PAPER · MANAGEMENT
The paper, dated 30 June and posted on 16 July, finds that managers often overestimate AI usefulness, including for tasks where it does not improve performance. Making its limitations visible improves deployment decisions.
04 / 07
PERSONAL ACCOUNT · MEANING OF WORK
A deeply personal response to recent AI progress in mathematics. Beyond employment, the author asks what happens to the meaning of a profession when a machine can make the discoveries that gave the work its purpose.
05 / 07
ECONOMIC RESEARCH · JOBS
AI is already blurring occupational boundaries: 43.5% of the specific workplace uses analysed involve tasks historically associated with another profession. It is changing not only how people work, but also who does what.
06 / 07
PUBLICATION · SCIENTIFIC RESEARCH
OpenAI reports ten results produced by an internal model on open research problems, several of them longstanding. This is a concrete example of AI moving from research assistant to direct producer of new scientific knowledge.
07 / 07
FIELD EXPERIMENT · MANAGEMENT
After several months in which five employees were genuinely managed by AI agents, the results are mixed: the managers were often accommodating and generous, but made concrete mistakes involving schedules, confidentiality, labour law and organisation. AI management is already more than a theoretical prospect.
No text matches this theme.
END OF SELECTION · TO BE CONTINUED