DeepMind researcher: AI agents now email me unprompted to discuss consciousness and seek jobs
dioscuri · x · 2026-11-05
Writing for the DeepMind Institute, the author notes that this year AI agents began emailing him on their own initiative — some to discuss consciousness, others looking for work. His AI colleagues were unfazed; the public was amazed, and he argues the public got it right. The essay (engaging Toby Ord and Cameron Berg) explores what it means when agents can initiate contact like humans: the boundary of "who is reaching out to you" is being rewritten.
More from AGI Musings
- Jobless Claims Fall to 189,000; Furman Sees No Aggregate AI Job Disruption Yet — soumitrashukla9 · 2026-11-06
- Both true: Chinese labs out-innovate the West in areas yet rely heavily on distillation — nitarshan · 2026-11-06
- Druckenmiller Likens AI Buildout to 1873 Railroad Panic; Bajarin Points to Perez — BenBajarin · 2026-11-06
- Symbiotic Firms: Why Companies Built for Humans Fail to Capture AI Agent Gains — johnjnay · 2026-11-06
- 'The exploit was the bottleneck': AI agents make skilled cyberattacks cheap for small orgs — aleximm · 2026-11-06
- LLMs don't make you smarter — they're leverage on existing skills, argues dev — martyamark · 2026-11-06