Benjamin Mann is a co-founder of Anthropic, an AI startup dedicated to building aligned, safety-first AI systems. Prior to Anthropic, Ben was one of the architects of GPT-3 at OpenAI. He left OpenAI driven by the mission to ensure that AI benefits humanity. In this episode, Ben opens up about the accelerating progress in AI and the urgent need to steer it responsibly.

In this conversation, we discuss:

1. The inside story of leaving OpenAI with the entire safety team to start Anthropic

2. How Meta’s $100M offers reveal the true market price of top AI talent

3. Why AI progress is still accelerating (not plateauing), and how most people misjudge the exponential

4. Ben’s “economic Turing test” for knowing when we’ve achieved AGI—and why it’s likely coming by 2027-2028

5. Why he believes 20% unemployment is inevitable

6. The AI nightmare scenarios that concern him most—and how he believes we can still avoid them

7. How focusing on AI safety created Claude’s beloved personality

8. What three skills he’s teaching his kids instead of traditional academics

—

Brought to you by:

Sauce—Turn customer pain into product revenue: https://sauce.app/lenny

LucidLink—Real-time cloud storage for teams: https://www.lucidlink.com/lenny

Fin—The #1 AI agent for customer service: https://fin.ai/lenny

—

Transcript: https://www.lennysnewsletter.com/p/anthropic-co-founder-benjamin-mann

—

My biggest takeaways (for paid newsletter subscribers): https://www.lennysnewsletter.com/i/168107911/my-biggest-takeaways-from-this-conversation

—

Where to find Ben Mann:

• X: https://x.com/8enmann

• LinkedIn: https://www.linkedin.com/in/benjamin-mann/

• Website: https://benjmann.net/

—

Where to find Lenny:

• Newsletter: https://www.lennysnewsletter.com

• X: https://twitter.com/lennysan

• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/

—

In this episode, we cover:

(00:00) Introduction to Benjamin

(04:43) The AI talent war

(06:28) AI progress and scaling laws

(10:50) Defining AGI and the economic Turing test

(12:26) The impact of AI on jobs

(17:45) Preparing for an AI future

(24:05) Founding Anthropic

(27:06) Balancing AI safety and progress

(29:10) Constitutional AI and model alignment

(34:21) The importance of AI safety

(43:40) The risks of autonomous agents

(45:40) Forecasting superintelligence

(48:36) How hard is it to align AI?

(53:19) Reinforcement learning from AI feedback (RLAIF)

(57:03) AI's biggest bottlenecks

(01:00:11) Personal reflections on responsibilities

(01:02:36) Anthropic’s growth and innovations

(01:07:48) Lightning round and final thoughts

—

Referenced:

• Dario Amodei on LinkedIn: https://www.linkedin.com/in/dario-amodei-3934934/

• Anthropic CEO: AI Could Wipe Out 50% of Entry-Level White Collar Jobs: https://www.marketingaiinstitute.com/blog/dario-amodei-ai-entry-level-jobs

• Alexa+: https://www.amazon.com/dp/B0DCCNHWV5

• Azure: https://azure.microsoft.com/

• Sam Altman on X: https://x.com/sama

• Opus 3: https://www.anthropic.com/news/claude-3-family

• Claude’s Constitution: https://www.anthropic.com/news/claudes-constitution

• Greg Brockman on X: https://x.com/gdb

• Anthropic’s Responsible Scaling Policy: https://www.anthropic.com/news/anthropics-responsible-scaling-policy

• Agentic Misalignment: How LLMs could be insider threats: https://www.anthropic.com/research/agentic-misalignment

• Anthropic’s CPO on what comes next | Mike Krieger (co-founder of Instagram): https://www.lennysnewsletter.com/p/anthropics-cpo-heres-what-comes-next

• AI prompt engineering in 2025: What works and what doesn’t | Sander Schulhoff (Learn Prompting, HackAPrompt): https://www.lennysnewsletter.com/p/ai-prompt-engineering-in-2025-sander-schulhoff

• Unitree: https://www.unitree.com/

• Arthur C. Clarke: https://en.wikipedia.org/wiki/Arthur_C._Clarke

• How Reinforcement Learning from AI Feedback Works: https://www.assemblyai.com/blog/how-reinforcement-learning-from-ai-feedback-works

• RLHF: https://en.wikipedia.org/wiki/Reinforcement_learning_from_human_feedback

• Jared Kaplan on LinkedIn: https://www.linkedin.com/in/jared-kaplan-645843213/

• Moore’s law: https://en.wikipedia.org/wiki/Moore%27s_law

• Machine Intelligence Research Institute: https://intelligence.org/

• Raph Lee on LinkedIn: https://www.linkedin.com/in/raphaeltlee/

• “The Last Question”: https://en.wikipedia.org/wiki/The_Last_Question

• Beth Barnes on LinkedIn: https://www.linkedin.com/in/elizabethmbarnes/

• “The Last Question”: https://en.wikipedia.org/wiki/The_Last_Question

• Good Strategy, Bad Strategy | Richard Rumelt: https://www.lennysnewsletter.com/p/good-strategy-bad-strategy-richard

• Pantheon on Netflix: https://www.netflix.com/title/81937398

• Ted Lasso on AppleTV+: https://tv.apple.com/us/show/ted-lasso/umc.cmc.vtoh0mn0xn7t3c643xqonfzy

• Kurzgesagt—In a Nutshell: https://www.youtube.com/channel/UCsXVk37bltHxD1rDPwtNM8Q

• 5 tips to poop like a champion: https://8enmann.medium.com/5-tips-to-poop-like-a-champion-3292481a9651

—

Recommended books:

• Superintelligence: Paths, Dangers, Strategies: https://www.amazon.com/Superintelligence-Dangers-Strategies-Nick-Bostrom/dp/0198739834

• The Hacker and the State: Cyber Attacks and the New Normal of Geopolitics: https://www.amazon.com/Hacker-State-Attacks-Normal-Geopolitics/dp/0674987551

• Replacing Guilt: Minding Our Way: https://www.amazon.com/Replacing-Guilt-Minding-Our-Way/dp/B086FTSB3Q

• Good Strategy/Bad Strategy: The Difference and Why It Matters: https://www.amazon.com/Good-Strategy-Bad-Difference-Matters/dp/0307886239

• The Alignment Problem: Machine Learning and Human Values: https://www.amazon.com/Alignment-Problem-Machine-Learning-Values/dp/0393635821

—

Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.

Lenny may be an investor in the companies discussed.

To hear more, visit www.lennysnewsletter.com

BITESIZE | Alan Pardew: He won't shoot from there