“If Mythos actually made Anthropic employees 4x more productive, I would radically shorten my timelines” by Ryan Greenblatt

About this episode

Ryan Greenblatt discusses the implications of AI productivity uplift at Anthropic and its potential impact on timelines for AI development.

Subtitle: Better estimates of uplift at AI companies seem helpful. Anthropic's system card for Mythos Preview says: It's unclear how we should interpret this. What do they mean by productivity uplift? To what extent is Anthropic's institutional view that the uplift is 4x? (Like, what do they mean by “We take this seriously and it is consistent with our own internal experience of the model.”) One straightforward interpretation is: AI systems improve the productivity of Anthropic so much that Anthropic would be indifferent between the current situation and a situation where all of their technical employees magically work 4 hours for every 1 hour (at equal productivity without burnout) but they get zero AI assistance. In other words, AI assistance is as useful as having their employees operate at 4x faster speeds for all activities (meetings, coding, thinking, writing, etc.) I’ll call this “4x serial labor acceleration”1 (see here for more discussion of this idea2 ). I currently think it's very unlikely that Anthropic's AIs are yielding 4x serial labor acceleration, but if I did come to believe it was true, I would update towards radically shorter timelines. (I tentatively think my…

People in this episode

Host: Ryan Greenblatt

Topics covered

• AI productivity
• Anthropic
• labor acceleration
• productivity uplift

Keywords

• Mythos
• AI companies
• productivity
• estimates
• timelines

Mentioned in this episode

Products: Mythos Preview, Automated Coder, Mythos

Books & works: Mythos

More episodes of Redwood Research Blog

• “A taxonomy of barriers to trading with early misaligned AIs” by Alexa Pan · April 21, 2026 · 2h 6m
• “Introducing LinuxArena” by Tyler · April 20, 2026 · 9 min
• “Current AIs seem pretty misaligned to me” by Ryan Greenblatt · April 15, 2026 · 1h 5m
• “Anthropic repeatedly accidentally trained against the CoT, demonstrating inadequate processes” by Alex Mallen, Ryan Greenblatt · April 14, 2026 · 12 min
• “Logit ROCs: Monitor TPR is linear in FPR in logit space” by Kerrick Staley, Aryan Bhatt, Julian Stastny · April 12, 2026 · 28 min
• “My picture of the present in AI” by Ryan Greenblatt · April 7, 2026 · 21 min