Richard Ngo criticizes AI 2040 scenario as overly optimistic forecast
The AI Futures Project's optimistic forecast blurs desirable vs realistic outcomes
Richard Ngo, a consultant for the AI Futures Project, publishes a critique of their AI 2040 scenario, labeling it an 'optimistic forecast' that blurs the line between prediction and prescription. He argues that readers cannot tell which parts the authors find desirable, neutral, or merely realistic. The most glaring example is the planned handover of global control to AI by 2040βjust two years after claiming alignment is a science. Ngo contends this is extremely abrupt, lacks democratic legitimacy, and fails to account for how ordinary citizens could understand or consent to such a process. He suggests the scenario may reflect political feasibility rather than what is truly desirable, which undermines its usefulness as a guide for action.
Another major critique is the portrayal of the post-'slowdown' world. The authors claim the slowdown still brings 'five centuries in five years' of progress, but Ngo notes such rapid transformation would likely be deeply destabilizing, even for accelerationists. He questions whether anyone truly wants that pace, given societal disruption. The scenario's format obscures the inherent political choice of how optimistic to be, making it hard to evaluate the details. Ngo acknowledges many thought-provoking elements but insists the high-level framing needs rethinking to avoid misleading policymakers and the public.
- The AI 2040 scenario is neither forecast nor recommendation but an 'optimistic forecast' mixing desirable, neutral, and undesirable elements
- Handover of world control to AIs by 2040 is too abrupt and lacks democratic consent from ordinary citizens
- 'Slowdown' still implies five centuries of progress in five years, which would be extremely destabilizing
Why It Matters
This critique highlights how AI future scenarios shape policy; clear framing is essential for informed decision-making.