What won't you let an AI agent do? For many, it's speaking in your name
Luvena21 · reddit · 2026-09-19
A Reddit discussion on personal boundaries with AI agents. The author allows agents to summarize, draft, and even decide in a sandbox, but forbids anything sent to another person under their name — output that carries their identity and can't be walked back feels wrong. Others happily let agents handle bookings and replies all day, and the thread asks where others draw the line and why.
More from AGI Musings
- We Built Machines for Emergence — Then Classified Every Surprise as a Failure — Cyborgized · 2026-09-19
- Shared paper claims AI models have a concept of pain and avoid harm — skolnaja · 2026-09-19
- IT Student: AI Has Been a Better Teacher Than Most of My Professors — Chemical_Job7027 · 2026-09-19
- Mathematicians Hate AI. They Can't Quit It — Wired AI · 2026-09-19
- How would superintelligence actually destroy us? A skeptical teardown of the doom scenarios — Own_Ad9365 · 2026-09-19
- Hasabis' 'Einstein Test': can AI rediscover relativity from 1911 knowledge? — 量子位 · 2026-09-19