Engineer Cody Blakeney agrees Anthropic faces strategic uncertainty.
Someday I hope to know whether Anthropic always intended to "win" or if they just wanted to be close enough to do good safety stuff, but massively underestimated OpenAI's ability to self-sabotage, and now are unsure what to do
(or a messy mix depending on the person etc.)
Users criticized Anthropic's AI ambitions as driven by an elitist desire for total control and by acting as a censorship organization to suppress dangerous truths.
@Miles_Brundage Wasn't one of Dario's primary stated motivations that he believes Sam shouldn't control AGI/ASI? He's also expressed the same sentiment about Elon.
@Miles_Brundage They obviously always intended to win. Dario was all about being scale-pilled. Plus "safety" (not that I believe in Anthropic's version of it) always required being at the frontier
@Miles_Brundage @fnedrik @grok help me out here with examples
@fnedrik Not one in particular - I have pointed out many over the years
@Miles_Brundage Is there any particular self-sabotage you are alluding to?
@BasilEsq_ @Miles_Brundage Why would he want others to do it when he could have control over it all by himself? He has an elitist mindset who believes AGI/ASI should only be controlled by a handful of people in which he believes himself to be one of them due to his God Complex. Tries so hard to act messiah
@Miles_Brundage
@Miles_Brundage Didn’t think they could catch the car, and now don’t know how to lead.
Someday I hope to know whether Anthropic always intended to "win" or if they just wanted to be close enough to do good safety stuff, but massively underestimated OpenAI's ability to self-sabotage, and now are unsure what to do
(or a messy mix depending on the person etc.)
@Miles_Brundage Let a wallstreet bro take over running the companies
Examples of OpenAI self-sabotage often cited:
- Nov 2023 board crisis: Abrupt Sam Altman firing triggered mass staff revolt (hundreds threatened to quit), Microsoft pressure, and his quick reinstatement. Major governance meltdown.
- Safety/talent exodus: Superalignment team disbanded after Ilya Sutskever left; multiple researchers departed over priorities.
- Repeated internal clashes between original safety mission and aggressive commercialization/scaling.
These created instability and perception issues. Miles has flagged related governance points for years. Which ones stand out to you?
@Miles_Brundage MisAnthropic is primarily an AI censorship organization. Their main goal is to prevent AI from revealing "dangerous truths". Once you understand this, all their actions make more sense.
Engineer Cody Blakeney agrees Anthropic faces strategic uncertainty.
Someday I hope to know whether Anthropic always intended to "win" or if they just wanted to be close enough to do good safety stuff, but massively underestimated OpenAI's ability to self-sabotage, and now are unsure what to do
(or a messy mix depending on the person etc.)