Dev questions AI ethics: if education can't make humans moral, why trust aligned AI
luislamb · x · 2026-10-31
luislamb argues it is deeply risky to assume we can build AI that behaves as planned, given we can't even guarantee humans become ethical through education. He asks whose morality underlies the systems built by those claiming AI omniscience, tagging Yoshua Bengio and Gary Marcus in this skeptical take on alignment feasibility.
More from AGI Musings
- Security researcher: 'one-person APTs' likely already exist, like one-person companies — cyb3rops · 2026-10-31
- Why AI won't rewrite the OS: decades of accumulated edge cases are the real moat — AryHHAry · 2026-10-31
- V建议大学生记「agency 日志」练主动性，网友：需要日志说明已被系统驯化 — rickasaurus · 2026-10-31
- Why AI can't just build the next Opus: the bottleneck is verification, not generation — AryHHAry · 2026-10-31
- DeepMind's Gemini agent pipeline cracks rules for 11 unresolved BB(6) Turing machines — rvp · 2026-10-31
- Snowflake exec to teams: empathy is the top AI skill, stop sharing 90-minute videos — marlene_zw · 2026-10-31