'The real alignment problem is bureaucracy,' not LLMs, argues VC
jbarbier · x · 2026-08-29
- The discussion starts from a quoted post: bureaucracy's purpose is to create a simulated environment that slowly conditions you to obey orders for no reason, until someone yells "left" in the labyrinth and you instinctively go left.
- The original poster replies "Kafka was a prophet," noting the surveillance state is autonomous not in the AI sense but in the way described in Seeing Like a State.
- Their core claim: the real alignment issue is embedded in the bureaucratic colonization of goals; LLMs aren't the problem — they may even help us track what the bureaucratic state does in its own name.
More from AGI Musings
- AI doesn't mean the end of mathematics – yet — ArtificialOther · 2026-08-29
- METR's HF agent probe sparks debate: not a swarm of users, but one octopus-like agent — AccBalanced · 2026-08-29
- Gary Marcus: AGI Believers Confuse Movie AI with Reality — GaryMarcus · 2026-08-29
- Affordable L5 Autonomy Needs AGI, Still 5-10 Years Away — npew · 2026-08-29
- Multi-agent science world writes 125 cited papers, finds new 604-sphere record beating AlphaEvolve — progenitor414 · 2026-08-29
- Companies need plans for open/local models to avoid vendor lock-in in critical AI infrastructure — AJChadha · 2026-08-29