Intelligence Is Not the Main Bottleneck
Many commenters argue that in fields like medicine, biotech, and climate policy, raw intelligence—human or artificial—is not the limiting factor; instead, progress is constrained by regulation, organizational incentives, human nature, and the difficulty of coordinating many actors. Others counter that these very governance and coordination problems are themselves failures of applied intelligence, and that smarter systems or leaders could in principle redesign institutions and processes. The thread also explores adjacent worries about AI’s role in labor (from human “avatars” guided by headsets to fully embodied robots) and challenges the idea that a superintelligent AGI could easily “solve” complex societal problems such as climate change or healthcare.
Intelligence vs. Actual Bottlenecks
- Many argue raw intelligence (human or AI) is not the main constraint; access to data, permission to act, regulation, and institutional incentives dominate.
- Others counter that intelligence underlies governance, regulation design, and resource allocation, so “the bottleneck is still intelligence, just at different nodes in the system.”
- Several comments stress that even when we already know what to do (e.g., health, climate), the gap is execution, coordination, and power, not smarter plans.
Human Nature, Communication, and Design
- A recurring theme: human nature and communication are primary bottlenecks.
- Building systems for machines is seen as straightforward; building for humans is hard due to variability, bias, and idiosyncrasies.
- Communication quality, signal-to-noise, and organizational culture are viewed as critical constraints on “collective intelligence.”
AI Embodiment, Labor, and Dehumanization
- One thread suggests embodiment as the next bottleneck: AI plans vs. physical execution.
- Proposal: humans as AI-guided “avatars” for physical work via AR/VR headsets.
- Many find this dystopian or “slave-like,” arguing it reduces humans to cheap actuators and formalizes existing dehumanization in low-wage work.
- Some note similar systems already exist (remote robots, AR task guidance) and warn about inhuman framing of “solutions.”
Biotech, Medicine, and Regulation
- Multiple comments highlight clinical trials, human trials, and regulatory processes as major bottlenecks.
- Even with better tools or AI, data release delays, endpoint validation, and misaligned incentives stall progress.
- Debate over whether “neolabs” and better tools show intelligence is still a core constraint, versus regulation and system design being the real blockers.
Persuasion, AGI, and Hyperrational Myths
- Skepticism toward claims that superintelligent AI will be “hyperpersuasive” or god-like problem solvers.
- Some link this to earlier rationalist/EA ideas about AI gaining control via persuasion, criticizing this as magical thinking.
- Others concede AI can scale propaganda and bots, but note that deeply held beliefs and social dynamics limit one-shot persuasion.
Climate Change, Governance, and Systems
- Several argue climate change is already “technically solvable” but blocked by politics, interests, and social systems.
- AGI-based salvation narratives are likened to belief in a rational, benevolent “machine god.”
- Governance, power asymmetries, public opinion, and groupthink are emphasized as central bottlenecks that more intelligence alone may not fix.