Introduction

> The first time I witnessed the power of immersive gaming, it was transformative; I remember that moment when a player encountered a monster for the first time—it was like tapping into a primal, reptilian reaction. It was clear to me then: "This is going to make a difference; this is going to be powerful."

> Transitioning to the world of virtual reality, I've been on a journey to create portals into new dimensions; I'm fascinated by the interplay between advancing VR and the possibilities of artificial general intelligence. The future holds immense potential, and I’m driven by the idea that “this is what it's all about.”

Programming languages

> My early fascination with computers ignited when I began tinkering with simple programs on TRS-80s, such as "PRINT John Carmack." Discovering the capabilities of programming felt magical, compelling me to dive deeper into the intricacies of coding and develop more complex games, even extending to innovative hacks like efficient screen scrolling on Apple II.

> Programming languages have distinct strengths and weaknesses, and while modern languages like Python offer convenience—especially in AI/ML work—they also impose significant performance trade-offs. For serious programming, I've always gravitated back to C and C++ for their control and efficiency, despite their potential pitfalls with memory management, as they still compel the programmer to be disciplined and thorough, fundamentally shaping robust software engineering practices.

> The philosophy of efficient programming and optimization has evolved significantly. While the constraints of older hardware required meticulous, low-level optimizations, today's software often doesn't need to push those boundaries. However, areas like VR development still demand precision optimization, echoing the exigencies of past game development where unlocking performance was paramount to achieving smooth, immersive experiences.

> Reflecting on future technologies, such as VR or even nuclear energy, it’s evident that understanding the limits and potentials of current systems is crucial. Effective programming is not just about executing tasks but creating adaptable, efficient systems over their lifespan. Innovation often lies in leveraging and rethinking these boundaries, whether pushing computing power for VR or considering radical engineering efficiencies in energy production.

Modern programming

> Building things that provide value to users is crucial: "Everything that we're doing really should flow from user value... we're solving problems for people or we're providing entertainment to people... that's how you make the world a better place."

> Balancing leadership and independence in product development: "You're at your best when you've got a leader that is at their limit... deeply cares about not letting anything slip through the cracks... but we also want independent teams bubbling up ideas... that's just not the company that they want the team to be."

Day in the life

> It's intriguing how my early days shaped my approach to productivity. I found a rhythm that worked for me, typically working "60 hours a week, a 10-hour day, six days a week," but I also realized that long sessions can sap your effectiveness. I don't believe in pushing myself to the limit like some do; I need my "eight hours of sleep" and a refreshed mind to perform well. The truth is, your marginal productivity may dip after long hours, but that doesn’t mean you're not getting anything done—it's about finding the balance and optimizing time.

> The dedication to hard work has deeper implications too. It's not just about the hours logged; it's about the growth that comes with it. I liken it to the principles behind "Jiro Dreams of Sushi"—immersing yourself in your craft day in and day out not only shows commitment but also builds your mastery over time. For me, putting in those hours didn't just lead to better results; it sincerely tempers your skills for the long haul. If you're passionate about achieving something, it's clear: "if you want to accomplish something, working longer and harder is the path to getting it accomplished."

Hard work

> Striking a work-life balance varies greatly among individuals, and for me, it's about recognizing what keeps you engaged and passionate. I've always thrived on having multiple interesting projects to switch between, which keeps me from feeling burnt out. This approach helps maintain enthusiasm in all aspects of my work and life, allowing me to stay productive and avoid stagnation.

> It's crucial not to dismiss the value of being obsessive about certain pursuits. For some, including myself, diving deeply into work can be immensely fulfilling and lead to significant progress. This intense focus can achieve outcomes that might not be possible with a more cautious or balanced approach. It's important to understand that everyone is different, and what works for one person might not work for another.

Pizza and Diet Coke

> One thing that defined being rich for me was being able to buy all the pizza I wanted. I used to have a pizza delivered every single day, and I even found out years later that I was getting a discount without knowing it, all because I was such a good customer. It was like a daily ritual that brought me joy.

> Despite my love for pizza, I've always been good about not overeating and staying healthy. Working at Facebook Meta provided me with good food options, and I get a decent amount of exercise, which has kept me in good health. I've managed to maintain a balance and avoid things that are not good for me. My health has held up well over the years, and I'm feeling good.

Setup

> There's a stark difference between how hardcore game developers and big Silicon Valley companies approach programming tools, particularly debuggers. "It's amazing what you can do with sometimes much better tools," and having a debugger at hand isn't just about fixing bugs; it's an essential part of the coding process, allowing developers to interactively explore code and understand systems that are too complex to grasp through sheer reading alone.

> My experience with analyzing massive code bases was eye-opening and humbling. "Anything that can be syntactically allowed in your language, if I'm... it's going to show up eventually in a large enough code base," which highlights the importance of automated tools and guardrails in software development. Good intentions aren’t enough; we all generate errors, and the right tools can catch those before they escalate into real problems.

> Reflecting on my journey, I realize that some of the most valuable lessons came from my earlier struggles. "There are mistakes that we made that we couldn't really have known how things would have worked out," but I’ve learned the importance of user value generation and team dynamics. Balancing technical mastery with effective management could have led to even greater successes, proving that there's always room for growth and understanding in this ever-evolving industry.

id Software

> Reflecting on the early days, I remember the pivotal role of games like Wolfenstein 3D, Doom, and Quake, which showcased the power of computers for graphics and performance. These games sparked a deep understanding of what machines were capable of, shaping my path profoundly.

> The journey began with humble beginnings, delving into contract programming to make ends meet. Softdisk, with its quirky monthly subscription software model, provided a platform to develop and port games across various systems, leading to pivotal encounters with talented programmers like John Romero and Lane Roth.

> The rapid iteration and pressure-cooker environment at Softdisk's Gamer's Edge department were crucial in honing our skills. Crafting numerous games monthly laid the foundation for our future successes, akin to how the Beatles honed their skills through early works before fame.

> The advent of scrolling technology on PCs presented a game-changing opportunity. Through innovative solutions like adaptive tile refresh, we could achieve smooth scrolling, bridging the gap between console and PC gaming experiences. This breakthrough eventually led to the development of games like Commander Keen and the evolution of shareware with Apogee Software's trilogy model.

Commander Keen

> The inception of Commander Keen really was a collaboration, with Tom Hall driving the creative vision while I handled the programming side. It all came together quickly, and we were surprised by how much success we saw right off the bat, earning around thirty thousand dollars a month from it—much more than we'd anticipated.

> Innovating with technology often comes down to simple ideas, like the scrolling technique I developed that wrapped memory instead of sticking to standard limits. “So many of the great ideas of technology are things that are back of the envelope designs,” and that method allowed us to push the boundaries creatively without being constrained by how things were typically done.

Hacker ethic

> The hacker ethic deeply influenced my career, emphasizing sharing information, collaboration, and taking joy in others’ accomplishments. Releasing the source code for id Software’s game titles was a big step towards embodying this ethic. It was about creating a community where knowledge was freely shared, ensuring that innovation could flourish collectively rather than being stifled by individual possession or fear of competition.

> The nuances of credit and recognition in programming often conflict with the hacker ethic. Despite being credited with various innovations, I try to focus on the joy of creation and community contribution over asserting primacy. As a technical culture, we should prioritize building cool stuff and sharing it. Over time, true contributors will be recognized through the natural evolution of history, rather than through immediate, sometimes misplaced, credit.

Wolfenstein 3D

> Wolfenstein 3D was a breakthrough in game development as it explored the power of changing perspectives to create an entirely different player experience. By adapting the same gameplay logic from 2D overhead games like Catacombs into a first-person shooter view, it triggered a significant impact on players, leading to a more immersive and reactive gaming experience.

> The technical innovations in Wolfenstein 3D, such as the raycasting approach for rendering 3D environments and the compiled scalar technique for fast sprite scaling, were pivotal in achieving a seamless and glitch-free game experience. The optimizations made, like using small assembly code for raycasting or pre-generating specific code for varying sprite sizes, ensured the game's performance even with limited hardware resources.

> The key takeaway from Wolfenstein 3D's development process is the concept of harnessing past programming experiences and techniques to address new challenges effectively. By drawing parallels between different programming tasks and optimizing data structures based on prior knowledge, it allows for efficient problem-solving and innovation across various gaming platforms and development scenarios.

Doom

> The leap from Wolfenstein to Doom was not just about graphics; it marked a shift in creative freedom. "In a game like Doom... there's kind of a touring completeness level of design freedom," allowing for endless levels to be made and experienced, a stark contrast to the more limited creativity of Wolfenstein.

> Introducing technologies like binary space partitioning was pivotal for enhancing rendering speed and visual complexity. The challenges of precision in 3D, particularly with "epsilon problems," pushed me to innovate and find new methods to manage complex geometry efficiently.

> Doom's design fostered a strong community and modding culture that encouraged user creativity, allowing players to create their own levels without destructively altering the primary game files, which was revolutionary. "We had this whole setup... people could build things and add it to your game without destructively modifying it."

Quake

> "Quake was a revolutionary project that pushed me to my limits, both technically and creatively. We aimed to create a true 3D engine with six degrees of freedom, complex lighting models, and an entirely new programmable aspect for modding. It was an ambitious leap from Doom, and in hindsight, splitting those innovations into two phases might have been more pragmatic."

> "Quake’s development was a high-stress environment but it taught me invaluable lessons about perseverance and the importance of incremental progress. I've always been able to cope with stress by bearing down on my work and knowing that a little more effort will lead to a breakthrough. It's a mindset that's kept me from slipping into darker places others might struggle with."

> "The experience with Quake's networking and graphics taught me that understanding the full technical stack is crucial for innovation. From high-level concepts to low-level hardware details, having that holistic view helps in making strategic sacrifices that deliver significant advantages. This approach applies across all fields I've worked in - from game development to VR, aerospace, and now AI."

John Romero

> When reflecting on the animosity with John Romero, I realized our corporate structure at id Software was flawed. Equal partnership without clear expectations led to tension when not everyone pulled their weight. Looking back, a different framework could have prevented pushing out valuable contributors.

> Despite our differences, I admired John Romero's versatility and initial impact on our projects. He was a talented programmer who brought creativity and flair to our work, leaving his mark on games like Doom and Quake. Our diverging focuses and perceptions on success created challenges, but I still appreciate his contributions and wish him the best in his endeavors, like his early foray into mobile gaming.

Metaverse

> The concept of the metaverse has long fascinated me, originating from Neil Stephenson's "Snow Crash." It embodies the limitless potential to craft virtual worlds that can provide incredible experiences, akin to how we approached game design during the days of Doom and Quake. Building an interconnected series of worlds isn’t overwhelmingly complex; it's about delivering something amazing and worthwhile.

> Success in the metaverse hinges on creating compelling, engaging experiences like Minecraft and Fortnite, rather than getting bogged down in overly technical capability-first approaches. If you make something extraordinary that users adore, you can evolve it instead of attempting to invent everything from scratch.

> The true essence of virtual reality lies in its ability to transport users to immersive worlds that outperform their reality—something I believe can significantly enhance lives. Virtual spaces can offer experiences that aren’t accessible in the physical world, democratizing aspects of life that are often reserved for the privileged, like beautiful landscapes or social gatherings.

> Despite being part of a large company, there's a need for agility and creativity that typically flourishes within smaller teams. My hope is that we unlock this potential by fostering an environment that encourages audacious thinking and innovation. I genuinely believe we're on the cusp of a significant transformation in how we experience connection and presence, both in virtual environments and beyond.

Elon Musk

> My friendship with Elon Musk has been enlightening, especially as we delve into detailed technical discussions. We've had conversations about things like engine material selection, propellant options, and even specific engineering details such as welds and sub-assembly processes. Despite being labeled by some as merely a business guy, Elon is profoundly involved in the technical aspects, and his ability to micromanage critical decisions for both Tesla and SpaceX is something I truly respect.

> The contrast between how Elon and I approached our aerospace endeavors is quite telling. While Elon went "all in" with significant personal risk on ventures like SpaceX and Tesla, my work with Armadillo Aerospace was more of a side project with finite resources that wouldn't impact my family if it failed. Elon's commitment to addressing massive problems like space exploration and electrification of transportation garners my deep admiration, especially as he was fully immersed in these projects from the start.

Mars

> I made a bet on boots on Mars by 2030, expressing less than 50% chance of success due to launch complexities and transit uncertainties. While infrastructure on Mars may emerge in the near future, human presence by 2030 seems less likely, but I believe in the possibility shortly after.

> The critical role of individuals, like Elon Musk, in ambitious endeavors like SpaceX aiming for human presence on Mars contrasts with self-driving cars, where multiple companies are actively working towards the goal. Personal passion drives Mars mission, while collaborative efforts shape the future of self-driving cars, a complex problem with multiple stakeholders and incentives.

Nuclear energy

> Nuclear energy, specifically fission, holds immense potential and simplicity that is often overlooked. “You put these two rocks together and they get hot all by themselves” – that’s a game-changer. The reserves of uranium could power the planet effectively, yet social perceptions cloud its promise, making it feel like the “smart money bet” for sustainable energy solutions.

> I've also been diving into solar photovoltaic systems, appreciating the “magical progress” in semiconductors, but acknowledging that they don’t address all energy needs. The disparity between technological capability and public sentiment is stark – it’s strange how non-technical factors overshadow advances that could fundamentally alleviate global energy crises, which underpin conflicts and unrest.

AGI

> Achieving artificial general intelligence (AGI) could be a pivotal moment in human history. The potential of AGI to fundamentally change the course of humanity is a driving force behind my focus on this field.

> AGI development doesn't necessarily require as many resources as commonly believed. I believe that the code for AGI could potentially be tens of thousands of lines, not millions, and that a few key insights may unlock its development.

> The path to AGI may involve a limited number of key insights. I speculate that less than six crucial breakthroughs, each simple enough to fit on an envelope, might be necessary to progress towards achieving AGI.

> The dichotomy between embodied and purely digital AGI forms is a key consideration. While some believe physical embodiment is essential, I maintain that simulated environments in the digital realm can provide substantial value and are not hindered by physical constraints.

> The trajectory of my work is shifting towards a deeper commitment to AGI research. After navigating various fields and experiences, I am now determined to delve more seriously into AI, emphasizing unique angles of attack and leveraging strategic investments to further this pursuit.

Andrej Karpathy

> Andre Kaparthy stands out to me as a true embodiment of the hacker spirit; he emphasizes getting things done and building intuition through coding, which resonates deeply with my own way of approaching challenges. His method of "coding up the simplest possible thing" to grasp new concepts reflects how I often let my thoughts flow through code, essentially sketching ideas out in a structured form.

> Collaborating with brilliant minds like Andre is invigorating, and while he may have a focus on educational work now, I can't help but feel that his exceptional talent will draw him into an irresistible opportunity soon. It's all about being in the right mindset to explore and share practical insights that can spark real advancements in AI, and I admire his commitment to cultivating that hacker ethos.

Martial arts

> Martial arts, particularly judo and jiu-jitsu, have been incredibly influential in shaping my outlook and resilience. Training in these disciplines from a young age, and then seriously in my mid-30s, provided a structured and humbling experience. The constant grappling with opponents, pushing through pain, and the camaraderie with fellow practitioners instilled a commitment to "embracing the grind." This practice not only served as a motivating form of exercise but also offered a vital perspective on mortality and humility—each tap-out during a choke serving as a visceral reminder of one's vulnerabilities.

> My journey with martial arts underscores a personal philosophy of persistent learning and adaptability. Over the years, especially as I've aged, my approach has evolved from seeking immediate mastery of techniques to appreciating the full spectrum and nuance of these arts. Martial arts is not just a physical activity but a rich mental exercise, akin to solving intricate problems or playing chess, which aligns seamlessly with my professional mindset. Interacting with masters like Carlos Machado or observing the systematic depth of John Danaher’s teachings further deepened my fascination with the scientific and artistic aspects of human movement and leverage. These insights translate beyond the mats, fostering a balanced approach to both professional and personal challenges.

Advice for young people

> "The advice I give young people is that the path I took may not be best for everyone. Embrace the grind, understand as much as you can, and be prepared for opportunities to present themselves based on a broad set of tools and awareness of the changing world."

>

> "I believe in weaponizing curiosity to find value and significance in various areas. You can attain happiness by deeply exploring and understanding a particular subject to make an impact and become a craftsman of that thing."

Meaning of life

> Life doesn't need to be overly complicated; I'm just a biological being shaped by evolution, and I find meaning in the simple act of focusing on the present and taking “little tiny steps” towards my goals rather than getting lost in existential dilemmas or grand visions.

> The journey of creation, whether it's building iconic games or tackling the challenge of artificial general intelligence, is about “following locally wherever the gradient ascent takes you.” It’s a process of hard work, experimentation, and finding success through focused effort, not some grand blueprint.