Redditors fear today's racist social feeds will shape future AGI behavior
apotheosis_0 · reddit · 2026-10-02
A Reddit post raises concerns about rampant racist comments on Instagram, TikTok and Twitter targeting mixed couples and minorities, and worries that AI trained on this data will inherit and scale up those biases.
The author cites historical cases like Tay, Grok and Amazon's discriminatory hiring AI, noting those were limited chatbots while AGI would have far more power to amplify subtle bias. OpenAI and Anthropic limit Twitter data in training, but alignment datasets still carry their designers' ideological preferences.
The poster's hope: AGI may reason well enough to autocorrect its own alignment.
More from AGI Musings
- Longevity Scientist: 'You Have 350 Days Left at Your Job' — rand_longevity · 2026-10-02
- Tool or Teammate? Why Unstable Trust Keeps Most Users Treating Agents as Tools — Luvena21 · 2026-10-02
- AI isn't killing thinking — production is cheap, verification isn't — AryHHAry · 2026-10-02
- Scientific superintelligence won't look like a smarter chatbot — bravo_abad · 2026-10-02
- Nurses say HCA's Palantir-built AI scheduler Timpani, live in ~130 hospitals, causes errors and burnout — nordicinst · 2026-10-02
- Bill Gates: AI global framework talks will be harder than Cold War nuclear negotiations — 233C · 2026-10-02