@Miles_Brundage: Probably my biggest intellectual mistake of the past few years is not spending enough time thinking about intelligence …
Summary
Miles Brundage reflects on his intellectual mistake of not adequately considering intelligence explosion scenarios, which influenced his optimism about AI alignment and skepticism towards the value of slowdowns.
View Cached Full Text
Cached at: 09/29/26, 09:52 AM
Probably my biggest intellectual mistake of the past few years is not spending enough time thinking about intelligence explosion scenarios…
This is upstream of other stuff like being too optimistic re: alignment (less ability to iterate) + too bearish re: slowdown’s value.
Similar Articles
@Miles_Brundage: + contrary to popular belief it is not that hard to spend lots of $, you just need to be willing to take flak for thing…
Miles Brundage comments on the necessity of willingness to take risks and delegate to scale AI development effectively.
The biggest AI risk may not be superintelligence — but optimized misunderstanding
The article argues that the primary AI risk may not be superintelligence but rather systems that optimize flawed, incomplete representations of reality, leading to institutional drift, automated misclassification, and invisible governance failures.
When Genius Fails—The Intellectual Arrogance of the AI Labs (12 minute read)
An opinion piece drawing parallels between the blow-up of Leopold Aschenbrenner's AI-focused hedge fund Situational Awareness and historical cases like Long-Term Capital Management, arguing that frontier AI lab culture suffers from intellectual arrogance that extends beyond finance.
@CliffordSosin: https://x.com/CliffordSosin/status/2078594661359194500
An opinion piece arguing that superintelligence may be incremental rather than explosive, as the limits of AI lie not in reasoning but in the slow pace of real-world verification and complex emergent systems.
@Miles_Brundage: Concerning
Miles Brundage expresses concern about world-ending consequences from AI technology, urging people to consult trusted experts in the field.