Gebru and Bender: this summer's AI breakthroughs are marketing, not magic
nordicinst · x · 2026-09-22
In an MIT Technology Review op-ed, Timnit Gebru and Emily Bender argue this summer's flurry of AI "breakthroughs" is largely marketing under scrutiny:
- Anthropic claimed Claude Mythos beats most security experts at finding vulnerabilities
- After the OpenAI–Hugging Face hacking incident, Anthropic (proudly) and Meta (reluctantly) disclosed similar incidents
- Both Anthropic and then OpenAI claimed mathematical breakthroughs by their models
- Anthropic engineer Jacob Coxon went viral leaving the company, saying it and OpenAI are "racing straight towards self-improving superintelligence and gambling with our lives"
The authors argue the press uncritically repeats companies' anthropomorphizing framings designed to paint the software as incipient AGI, and urge fact-checking and accountable, transparent AI over hype.
More from AGI Musings
- Economist links AI labor-market research in new Forbes companion piece — robseamans · 2026-09-22
- Commentary: AI's real problem is cost and reliability, not extinction — DavidLinthicum · 2026-09-22
- Who protects everyone from the people claiming to control AI? — Ray_Bayesian · 2026-09-22
- NYU's Robert Seamans publishes AI labor policy framework for the Aspen Institute — robseamans · 2026-09-22
- Opinion: Bill Gates' AI Predictions Are the Latest Chapter in an Old Story — DavidLinthicum · 2026-09-22
- Track 1 vs Track 2: the data science work AI is commoditizing, and what stays valuable — mdancho84 · 2026-09-22