person
Stanford economist and former Acemoglu colleague who projects substantially greater AI productivity gains and is described as the main scholar showing potential AI job losses.
No evidence-backed relationships are recorded.
Entry-level hiring can be switched back on in a quarter. The repetitions juniors used to earn on the way to competence cannot, and the shortfall will only be legible when this year's cohort should be running things.
Reality
Adoption has outrun the personal computer and the internet, yet the technological and institutional foundation stays unsettled. The move is to buy optionality, not lock in architecture.
Reality
A new cost analysis puts OpenAI's frontier model at half Anthropic's price per benchmark task. The retry and cleanup arithmetic behind that number is less settled than the price sheet.
Reality
A July 13 statement of eighty-eight words has collected close to 2,000 signatures and 17 Nobel laureates. The measurement work it implies has not been done inside a single company.
Reality
The MIT economist told Fortune that AI is neither a stochastic parrot nor a guaranteed win, and that adoption evidence outside coding is thin. Both halves bear on how you size spending.
Reality