George Hotz, the hacker known as geohot, has transitioned from breaking consumer electronics to building the future of autonomous vehicles and artificial intelligence. His philosophy emphasizes radical simplicity, the decentralization of power, and a deep-seated belief that the universe is fundamentally computational.

Part 1: The Art of Hacking & Engineering

  1. On the Essence of Hacking: "Hacking isn't about getting what you didn't pay for, it's about making sure you do get what you did." — Source: Velvet Cache
  2. On Consumer Rights: "This case is about a lot more than what I did and me. It's about whether you really own that device that you purchased." — Source: Eurogamer
  3. On Code Refactoring: "You have never refactored enough. Your code can get smaller, your code can get simpler, your ideas can be more elegant." — Source: Lex Fridman Podcast
  4. On Code as Liability: "Code is a liability. Every line you write is a line you have to maintain." — Source: SE Daily
  5. On System Exploitation: "Hacking is figuring out how to speak to the device, then persuading it to obey your wishes." — Source: iMore
  6. On Software Quality: "Software engineering today is mostly trash; it's just translating business requirements into React code without understanding the computer." — Source: YouTube
  7. On the Hacker Mindset: "I'm not looking for a job. I'll ping you when I crush Mobileye." — Source: Bloomberg
  8. On Simplicity: "If you have a function with 5 arguments, you have a problem. Your abstractions are wrong." — Source: YouTube Archive
  9. On Artistic Code: "Efficiently writing idiomatic code 100% is artistic." — Source: Reddit AMA
  10. On Motivation: "It wasn't to be rich. I wanted to use it [the iPhone] with T-Mobile." — Source: Reddit

Part 2: Deep Learning & the tinygrad Philosophy

  1. On tinygrad's Vision: "Tinygrad is the RISC of the machine learning stack—extreme simplicity that allows anyone to understand it." — Source: Data Driven Investor
  2. On Hardware Control: "Fundamental limitation of cloud is who owns the off switch." — Source: Podcast Notes
  3. On AI Efficiency: "Neural networks primarily use static memory accesses and ADD/MUL operations; you don't need a complex stack." — Source: Latent Space
  4. On the Four Operations: "Everything in deep learning can be reduced to four types: Unary, Binary, Reduce, and Movement." — Source: tinygrad Documentation
  5. On LLM Disruption: "When someone makes an LLM that's capable of citing its sources, it will kill Google." — Source: Lex Fridman Podcast
  6. On AI Safety Hysteria: "I don't think AI kills the human species. I think it kills society as we know it, which is much harder to do than people think." — Source: Lex Fridman Podcast
  7. On Intelligence Scaling: "Intelligence doesn't 'go critical' in the catastrophic sense; we’ve already been in a loop of recursive self-improvement with tools." — Source: Dwarkesh Patel Podcast
  8. On AI vs. Human Drivers: "Self-driving cars don't need to be perfect; they just need to be better than humans." — Source: Goodreads
  9. On End-to-End Learning: "Feature engineering methods will ultimately be surpassed by comprehensive AI solutions, like AlphaZero beat Stockfish." — Source: Medium
  10. On Neural Net Complexity: "If your AI coding assistant is frequently helpful, your framework is probably too complex." — Source: YouTube

Part 3: Learning, Education & the Digital Mind

  1. On Self-Education: "Don't go and try and do something 'substantial in the space.' Just go do SOMETHING." — Source: Substack
  2. On the Purpose of School: "The school system is a factory for compliance; it doesn't align with learning useful skills." — Source: YouTube
  3. On the Internet as a Resource: "With the internet, everyone has the same resources; what matters is where you spend your time and what you do with it." — Source: YouTube
  4. On Practicality: "We don't do whys here. We only do hows." — Source: Substack
  5. On Intrinsic Motivation: "Working hard for other people is not important. You gotta work hard on things you care about." — Source: Substack
  6. On Hacking Nature: "Viruses are just code; biology is just another system to be understood through programming." — Source: Reddit
  7. On Deep Understanding: "If you don't fully understand the code that your agent is writing, you're backing yourself into a corner." — Source: YouTube
  8. On Reading Code: "The best way to learn how to code is to read good code. Books are mostly filler." — Source: YouTube Archive
  9. On Mastery: "I don't think I'm special; I just spent more time looking at the registers than you did." — Source: Bloomberg
  10. On Iteration: "The process of writing good software is iterative, requiring multiple rewrites until the idea is elegant." — Source: SE Daily

Part 4: Business, Startups & comma.ai

  1. On Big Tech Waste: "Waymo is not a product, it's a press demo. $250,000 to outfit a car that still needs a driver is not viable." — Source: Third Law Reaction
  2. On the Comma.ai Strategy: "We are the Android of self-driving cars, while Tesla is the iOS." — Source: YouTube
  3. On Revenue and Funding: "If you are pre-revenue and you've raised $10 billion, it's against everything Silicon Valley stands for." — Source: Medium
  4. On Making Driving Chill: "The point of comma.ai isn't to solve level 5; it's to make driving chill in measurable, valuable increments." — Source: Comma.ai Blog
  5. On Market Disruption: "The winners of the self-driving race won't be traditional car makers, just like Nokia didn't win smartphones." — Source: YouTube
  6. On Capital Efficiency: "The point of comma.ai is to solve an AI problem, not just to make money. We value capital efficiency above all." — Source: SE Daily
  7. On Open Source Trust: "Open source is the only way to build trust in a system that controls your car." — Source: Wikiquote
  8. On Regulators: "I'd rather spend my life building amazing technology than dealing with regulators and lawyers." — Source: Forbes
  9. On Decentralization: "My central thesis is that systems that centralize power are bad, and those that decentralize it are good." — Source: Wikiquote
  10. On Human Drivers: "Humans are really, really good drivers—absurdly good. Replacing them entirely is much harder than people realize." — Source: CleanTechnica

Part 5: Philosophy, Reality & the Future

  1. On Consciousness: "I don't think I'm conscious. I think I'm just a computer program." — Source: Lex Fridman Podcast
  2. On the Universe: "Everything running in the universe is computation. I believe the extended Church-Turing thesis." — Source: Lex Fridman Podcast
  3. On the Fermi Paradox: "I think intelligent civilizations existed elsewhere in the universe, but they've blown themselves up." — Source: Lex Fridman Podcast
  4. On Immortality: "Immortality is the only good objective function. Once you have that, you can do whatever else you want." — Source: Lex Fridman Podcast
  5. On Capitalism: "I'm a fan of capitalism, but the game we're playing right now is rigged; it's not real capitalism." — Source: Lex Fridman Podcast
  6. On Social Media: "Diversity in humanity is decreasing because we are all interconnected through social media, converging on the same ideas." — Source: Lex Fridman Podcast
  7. On Hard Problems: "I don't think hard things are actually hard. I don't think P equals NP, but most 'hard' problems are just poorly understood." — Source: Lex Fridman Podcast
  8. On Human Alignment: "The idea of a single unaligned superintelligence is less concerning than the existing lack of alignment among humans." — Source: Dwarkesh Patel Podcast
  9. On Reality vs. Games: "Stop playing these fake human games. There's a real game here—the game where nature wrote the rules." — Source: Lex Fridman Podcast
  10. On the Future of Power: "The amount of power in the world is increasing. What matters is the distribution of that power." — Source: Dwarkesh Patel Podcast