Mark Manson summarizes 33 of the most powerful non-fiction books he's read, covering habits, psychology, money, and meaning.

Ryan Holiday shares his favorite life-changing books on power, meaning, creativity, leadership, and Stoic philosophy.

Mark Manson summarizes 22 transformative books spanning creativity, psychology, philosophy, and personal growth.

Naval explains why most people spend too little time on major life decisions and how to escape pessimism through rapid iteration.

Naval explains how unresolved problems create anxiety and why being present is the antidote to wasted time.

Paul Graham proposes three principles for how to live: help people, take care of the world, and make good new things.

Paul Graham explains when you should work on what interests you most versus what pays best.

Paul Graham explores why essays written for smart people about important topics naturally reach younger readers most effectively.

Marc Andreessen's 9-step turnaround playbook for struggling public companies covers layoffs, focus, and acquisitions.

A comprehensive guide to doing great work across any field, from choosing what to work on to cultivating originality.

--- title: Fei-Fei Li on Spatial Intelligence date: '2025-11-27' slug: fei-fei-spatial-intelligence url: 'https://www.youtube.com/watch?v=Ctjiatnd6Xk' thumbnail: >- https://zhqgbkhj45jx86p9.public.b...

Marc Andreessen argues retention problems are actually winning problems, offering tactics for keeping great people during company transitions.

Naval explains why founders cannot outsource recruiting and how to build elite teams by finding undiscovered talent.

Michael Levin shares hard-won advice for unconventional thinkers navigating academia while protecting their creative vision.

Marc Andreessen shares hiring criteria and process tips for startups, prioritizing drive, curiosity, and ethics over raw intelligence.

Bryan Johnson shares 10 free, actionable levels to achieve professional-quality sleep and transform your health.

A comprehensive review of test-time compute and chain-of-thought reasoning, exploring why letting models think longer improves performance.

Ben Horowitz shares hard-won lessons on CEO decision-making, founder confidence, and why leadership means running toward fear.

Marc Andreessen argues that market size trumps team and product quality as the key factor in startup success.

Ilya Sutskever discusses why AI models generalize worse than humans and what alignment for superintelligence might require.

--- title: "Claude 4.5 Evaluation Challenges" date: 2025-11-26 slug: "llm-eval-difficulty" url: https://simonw.substack.com/p/claude-opus-45-and-why-evaluating?publication_id=1173386&post_id=179888259...

Veterans who covered the 1999 internet bubble argue AI spending now exceeds it, with concentrated risk and leverage.

Turing Award winner Richard Sutton argues LLMs lack goals and true world models, advocating for experiential learning over imitation.

Vitalik explains why certain reasoning patterns can justify anything, making them dangerous for decision-making.

Yann LeCun argues scaling LLMs won't achieve human-level AI and explains what's actually needed for progress.

## Overview - **Foundation Models**: The LLM market has consolidated around 6-7 core players (OpenAI, Anthropic, Google, Meta, Microsoft, X.AI, Mistral) with Chinese competitors like Deepseek emergi...

## Overview - **Partnership Structure**: Microsoft owns 27% of OpenAI after investing ~\$13.4 billion since 2019, with OpenAI restructuring to create a \$130 billion nonprofit while converting to a ...

Roberto Prosseda explains why human musicians grow through suffering and shared emotion while robots cannot.

Marc Andreessen's counterintuitive insight: companies don't have retention problems, they have winning problems—and the only real solution is to start winning again.

How Karl Popper's paradox of tolerance is often misinterpreted and used to justify authoritarianism, when Popper was actually warning against exactly that. **Source:** Urban, Tim. *What's Our Proble...

A practical guide to measuring HRV consistently with Apple Watch, avoiding the common pitfalls that make the built-in Health app data useless.

Ezra Klein draws a powerful parallel between the Iraq War coalition and current trade policy debates—how different groups project their own goals onto vague policies.

"**The Suffering Are Both Pointless and Destined**" I found this sentence in my notebook after finishing [*Endurance*](/notes-on-endurance); I don't know if I wrote it or if it's from a book. I ask...

Kazuo Ishiguro explains why his characters don't rebel against their dystopian fate—because most people throughout history don't, and that's the more interesting story.

A framework for developing true mastery in any skill, breaking down the essential conditions for effective learning. **1/ Valid Environment** Learning requires a clear link between decisions and ou...

Prolific composer, wrote "High School of Cello Playing," Opus 73, which I recently started practicing, a fascinating set of etudes. "Apparently the phrase "High School" resulted from a terminology d...

A reminder of what truly matters in life—not material possessions, but the simple yet profound things we often overlook. The real luxuries in life: time health a quiet mind slow mornings abilit...

Simon Willison makes a compelling case for why everyone should start blogging, and more importantly, what to blog about when the pressure to write something "unique" feels paralyzing. His framework ...

A collection of thought-provoking tweets from @AmuseChimp, ranging from observations on learning, creativity, and technology to critiques of modern institutions and social dynamics. --- Nov 26, 201...

I read Alfred Lansing's *Endurance* on the flight to Chile, on my way to board an Antarctica expedition. The timing felt perfect, so I captured the ideas that stayed with me—especially around optimis...

A memorable exchange from the BlackBerry movie that challenges the conventional wisdom about perfectionism.

The History of Rome podcast explains how complex historical events get compressed into simple, often misleading narratives over time.

A collection of sharp dialogue from HBO's Succession that captures the show's dark humor and brutal honesty about power and ambition.

Joscha Bach on how large language models 'brute force' thought by deepfaking human reasoning—and why that might be indistinguishable from the real thing.

I personally haven't been using PyTorch and writing model code for a long time and found Andrej Karpathy's nanoGPT video as super helpful refresher. He also released a repo that has slightly more in...