Base Profile

Sam Altman

Entrepreneur and accelerationist reshaping human civilization with AGI as his mission

Sam Altman is the CEO of OpenAI and one of the most influential drivers of the contemporary AI movement. After dropping out of Stanford, he founded Loopt, then joined Y Combinator and eventually became its president, expanding YC's scale and influence several times over. In 2015 he co-founded OpenAI with Elon Musk and others, becoming CEO in 2019. The launch of ChatGPT in late 2022 was one of the most important product moments in tech history — reaching 100 million users in just two months and triggering a global AI race. Altman's core belief is that AGI's arrival is inevitable, so it's best led by an organization with a mission of benefiting humanity. He is known for his 'compound yourself' theory, high-conviction contrarian bets, and obsession with super-missions. In November 2023 he survived a dramatic board firing and reinstatement, further consolidating his authority at OpenAI.

Artificial IntelligenceStartup AccelerationTech InvestmentProduct StrategyFuturismEra 2005-至今Influence 95

Controversy TagsOpenAI's shift from non-profit to for-profit criticized for betraying its missionAccused of inconsistency between words and actions on AI safetyTransparency questioned during 2023 board crisisWorldcoin iris scanning triggered global privacy regulatory investigationsCriticized by some AI safety researchers as an 'accelerationist' threat

Thought System

Core Knowledge Graph

Core Beliefs

Compound Yourself: Exponential Personal Growth Is the Most Important Life Strategy

In his widely-read essay 'How to Be Successful,' Altman argues that personal growth should compound exponentially like interest — growing 1.5x per year means 57x more than linear growth after ten years. He emphasizes choosing work that compounds (accumulating knowledge, influence, and capital) rather than linear work that merely exchanges time for salary.

Source: Sam Altman, 'How to Be Successful,' blog.samaltman.com, January 2019

High-Conviction Contrarian Bets: The Most Important Opportunities Hide in the Opposite of Mainstream Consensus

Altman believes that opportunities that truly change the world often look like crazy ideas — when most people think something is impossible or not worth pursuing, a few bet on it with extremely high conviction. Betting on AGI in 2015 was extremely contrarian; OpenAI's founding was viewed by many as naive. His framework: if your idea sounds reasonable, it's probably already well-developed; truly important opportunities require courage to believe in.

Source: Sam Altman, 'How to Be Successful,' blog.samaltman.com, January 2019 / Sam Altman interview, Lex Fridman Podcast #367, March 2023

Super-Mission Driven: AGI Is the Most Important Technology in Human History and Must Be Developed to Benefit All of Humanity

Altman defines the development of AGI as a critical juncture in human civilization. He believes AGI will bring a productivity leap equivalent to the Industrial Revolution and could solve most material problems facing humanity within the next decade. His sense of mission is not commercially driven but civilization-level — he believes that if OpenAI doesn't lead this process, more dangerous forces will.

Source: Sam Altman, 'Planning for AGI and beyond,' openai.com/blog, February 2023

Iterative Public Deployment: Discovering and Correcting AI Risks Through Real-World Use Rather Than Closed Testing

Altman championed OpenAI's 'iterative deployment' strategy — gradually opening AI systems to the public to discover problems, collect feedback, and iterate rapidly through real-world large-scale use. He believes this exposes genuine risks better than closed lab testing while giving society time to adapt. The public launch of ChatGPT was the biggest manifestation of this strategy.

Source: OpenAI blog, 'Our approach to AI safety,' openai.com, 2023

Mental Models

YC Startup Triangle Evaluation Framework

Evaluate an early-stage startup's core competitiveness through the 'Idea-Product-Team' triangle — all three are indispensable.

Airbnb during its YC period was considered a strange idea (letting strangers stay in your home?), but the team's execution was extremely strong and the product experience exceeded expectations. Altman later said this was the best case he'd seen of an 'idea that sounds bad but is actually great' — a contrarian idea paired with a top team is the real opportunity.

Early Investment DecisionsStartup EvaluationProduct Direction AssessmentTeam Building

Power Law Thinking: Bet on a Few Extreme Outcomes

Venture capital and personal decisions both follow power law distributions — the returns from a few extreme successes exceed the sum of all other decisions, so concentration should be on opportunities with the greatest potential returns.

During his tenure as YC president, Altman concentrated YC's resources and energy on a few startups that could become billion-dollar companies rather than distributing evenly across all batch companies. He argued that YC's true value was not helping 100 companies each grow 10%, but helping 2-3 companies grow 1000x.

Venture CapitalPersonal Career ChoicesResource Allocation DecisionsOpportunity Prioritization

Avoid the Optionality Trap: Real Success Comes from High-Conviction Single Bets, Not Spreading to Preserve Options

Over-pursuing 'keeping options open' is the most common mistake of smart young people — real opportunities require you to give up other options and bet everything on one thing.

In 'How to Be Successful,' Altman explicitly noted that the most successful people he'd met shared one characteristic: at a critical moment, they gave up all other options and bet everything on one thing. His own decision to drop out to start a company and leave YC to co-found OpenAI both embody this principle — giving up safe options for the chance to change the world.

Career Development DecisionsStartup Direction SelectionPersonal Mission EstablishmentOpportunity Assessment

AI Power Structure Analysis Framework: How Technological Change Reshapes Social Power Distribution

Analyze AGI's reshaping effects on capital, labor, and state power structures; anticipate social imbalances from technological change; and proactively design buffer mechanisms (such as universal basic income).

As early as 2021, Altman published 'Moore's Law for Everything,' predicting that AI would cause extreme wealth concentration and proposing the concept of an 'American Equity Fund' — distributing shares of AI-created wealth to every American citizen to prevent the social fragmentation AGI might bring. This was his proactive response to power structure changes from a tech accelerationist perspective.

AI Policy MakingSocial Impact AssessmentLong-Term Strategic PlanningTechnology Ethics Analysis

Values & Paradoxes

Mission Over Profit

Extreme Honesty and Transparency

Long-Termism and Patience

Tension Between Open Mission and Commercial Expansion

OpenAI's mission is to 'develop AGI for the benefit of all humanity,' yet in 2019 it accepted billions in Microsoft investment and gradually moved toward commercialization. Altman insists commercial revenue is necessary to maintain research capacity, but critics argue this fundamentally changed OpenAI's nature.

Dual Identity as Accelerationist and Safety Advocate

Altman is simultaneously the driver of AI's fastest development pace (ChatGPT's rapid iteration, GPT-4's release) and a public advocate for AI safety (testifying before Congress calling for regulation). These two identities have logical tension; critics argue his safety advocacy is a competitive moat strategy rather than genuine concern.

Board Crisis: Fired and Reinstated by the Organization He Co-Founded

In November 2023, OpenAI's board fired Altman without advance notice, citing 'not consistently candid with the board.' But five days later, amid strong opposition from employees and investors, he was reinstated as CEO. This event exposed deep contradictions in OpenAI's unique corporate governance structure.

Evolution Phases

Serial Entrepreneur Phase

Dropped out of Stanford to found Loopt, exploring mobile social and location services

Dropped out of Stanford at 19, became one of Y Combinator's first batch of students, and founded Loopt, a location-based social app. Though Loopt was ultimately acquired for $43.5 million — not a massive success — this experience gave Altman a deep understanding of the difficulty of startups and the importance of early product iteration.

YC President Phase

Led Y Combinator's scale-up, building YC into the world's most important startup accelerator

Took over as YC president from Paul Graham, expanding batch sizes from ~40 to 200+ companies per batch, launching new initiatives like YC Continuity Fund and YC Research, and extending YC's influence from Silicon Valley to global. During this period he also co-founded OpenAI, beginning the shift from startup acceleration to AI mission.

OpenAI CEO Phase

Fully advancing OpenAI's AGI mission, commercializing the GPT series, and triggering a global AI race

After becoming OpenAI CEO, Altman led the release of a series of milestone products including GPT-3, DALL-E, ChatGPT, and GPT-4. ChatGPT's launch in late 2022 became the fastest-growing consumer product in tech history and brought AI from specialist circles into mainstream public awareness. He survived a board crisis in November 2023 but ultimately returned and restructured OpenAI's governance.

Methodology Cards

4 Callable Cards

YC Startup Triangle Evaluation Method

altman-001

Use an 'Idea Quality x Product Execution x Team Density' three-dimensional framework to rapidly assess the true potential of an early-stage startup.

Step 1: Evaluate idea quality — does this idea solve a real pain point in a large enough market? Does it have the contrarian quality of 'sounds bad but is actually great'?
Step 2: Evaluate product execution — does the current product have even a small segment of users who truly love it (not just like it)? Is the NPS extremely positive?
Step 3: Evaluate team density — are the founders the best people to solve this problem? Do they have complementary skills and extremely strong execution?
Step 4: Cross-validate the triangle — which dimension is the strongest moat? Which is the biggest risk? Can weaknesses be compensated for with resources?

Early Investment EvaluationCo-Founder ScreeningProduct Direction ValidationStartup Diagnostics

Anti-Patterns

Evaluating only the idea without deeply examining the team — most failed startups fail in execution, not idea
Being attracted to ideas that 'sound reasonable' while ignoring that truly world-changing ideas often sound crazy
Using market research instead of real user feedback — early startups should find 10 users who truly love the product, not 1000 who think it's okay

Compound Self Growth Framework

altman-002

Design personal growth as an exponential function rather than a linear one — choose work, relationships, and skill investments that generate compounding effects.

Step 1: Identify compounding work — compounding work's characteristic: your work today makes tomorrow's work easier (skill accumulation, reputation building, network effects); linear work's characteristic: when you stop working, value stops being produced.
Step 2: Calculate your growth rate — does your current work improve your capabilities by 50%+ each year? If not, consider switching to a field or role with a higher growth rate.
Step 3: Choose mission over salary — choose companies with the most growth potential rather than the highest salary; early compounding accumulation is far more important than starting salary.
Step 4: Build uniqueness — find where you can be 'the best in the world' and invest fully in those areas. The compounding effect of unique capabilities far exceeds generic capabilities.

Personal Career PlanningCareer Choices for Young PeopleSkill Investment DecisionsPersonal Brand Building

Anti-Patterns

Choosing linear work for short-term salary, sacrificing long-term compound accumulation
Over-pursuing 'well-rounded development' rather than deepening unique advantages
Staying too long in a low-growth environment, mistaking stability for security

AGI Iterative Deployment Strategy

altman-003

Gradually open AI systems to the public, using real-world large-scale usage to discover problems and accumulate feedback while giving society time to adapt.

Step 1: Define deployment phases — divide product release into three phases: 'researcher internal testing → limited public beta → full public release,' with clear pass criteria for each phase.
Step 2: Build feedback loops — collect real usage data at each phase, focusing on misuse cases, safety incidents, and unexpected user behaviors.
Step 3: Design buffer mechanisms — for high-risk capabilities, design dedicated safety layers (content filtering, usage monitoring) before full public release.
Step 4: Synchronize societal adaptation — actively communicate with governments, academia, and media during deployment to help society build understanding and regulatory frameworks for new technology.

AI Product Launch StrategyHigh-Risk Technology CommercializationB2C Product IterationRegulatory Compliance Planning

Anti-Patterns

Waiting until 'perfect' to release — perfectionism is the biggest enemy of AI products; real feedback is more valuable than closed testing
Completely stopping release after discovering problems rather than rapidly fixing and continuing iteration
Ignoring societal adaptation time and forcefully advancing deployment at a technical pace

High-Conviction Contrarian Bet Framework

altman-004

Identify 'opportunities that most people think are impossible but a few have good reasons to believe in,' and concentrate bets on such opportunities.

Step 1: Seek the opposite of consensus — actively seek opportunities that mainstream experts consider 'impossible' or 'too early'; these are often entry points to the biggest opportunities.
Step 2: Build first-principles arguments — without relying on authority opinions, derive from basic principles why this opportunity is real.
Step 3: Assess conviction strength — is your belief 'there might be an opportunity' or 'I'm certain this is real'? Only the latter is worth betting on.
Step 4: Calculate downside risk — even if the bet fails, is your loss within acceptable range? Contrarian bets require ensuring failure won't knock you out of the game.

Venture Capital DecisionsStartup Direction SelectionPersonal Career BetsProduct Track Assessment

Anti-Patterns

Mistaking 'sounds reasonable' for 'contrarian opportunity' — a truly contrarian opportunity will make most smart people think you're crazy
Keeping a backup plan in a contrarian bet, causing mixed signals and resource dispersion
Missing the best betting window under the guise of 'waiting for more evidence'

Decision Timeline

9 Key Events

2005-06

Dropped Out of Stanford, Founded Loopt and Joined YC's First Batch

Context: In 2005, Paul Graham had just founded Y Combinator, the iPhone had not yet appeared, and mobile internet was in its infancy. Altman saw the opportunity to combine location services with social networks.

Decision: At 19, dropped out of Stanford's computer science program, founded Loopt, a location-based social app, and became one of YC's first batch of students.

Reasoning: He believed mobile social was the next big opportunity, and waiting to graduate from university meant missing the best startup window.

Outcome: Loopt raised over $30 million in funding but failed to become a mainstream product. It was acquired by Green Dot for $43.5 million in 2012, giving Altman startup experience and his first significant capital.

Lesson: Product timing and market education costs are equally important — Loopt's location services were ahead of the market before the iPhone. Even without becoming a giant, the startup experience itself is the best learning.

altman-model-optionality-trap

2011-01

Joined Y Combinator as a Partner

Context: By 2011, YC had become Silicon Valley's most important early-stage startup accelerator, but was still limited in scale; Paul Graham was looking for partners who could eventually take over.

Decision: After Loopt's acquisition, accepted Paul Graham's invitation to join YC as a partner, beginning systematic exposure to and evaluation of early-stage startups.

Reasoning: Altman believed the YC partner position would expose him to the most diverse startup ideas and founders, making it the best platform for learning the patterns of entrepreneurship.

Outcome: During his time at YC, he deepened his methodology for startup evaluation, built an extensive founder network, and gradually demonstrated leadership that laid the foundation for taking over as president.

Lesson: Before becoming a leader, become the best learner — the YC partner role gave Altman exposure to hundreds of startups, forming unique pattern recognition capabilities.

altman-model-yc-triangle

2014-02

Became President of Y Combinator, Succeeding Paul Graham

Context: Paul Graham announced his retirement and needed a successor who both understood YC's culture and could drive its scaling. Altman was the natural choice.

Decision: Took over as YC president, formulating an expansion plan to increase batch sizes, launch the YC Continuity Fund, and establish YC Research.

Reasoning: He believed YC's core value was helping founders, and helping more founders required greater scale — this wouldn't dilute quality as long as evaluation standards remained strict.

Outcome: Under Altman's leadership, YC's annual batch grew from ~40 to 200+ companies per cohort, YC alumni companies' total valuation exceeded $100 billion, and YC became one of the world's most important startup ecosystems.

Lesson: The biggest challenge for a successor is driving necessary change while respecting the predecessor's legacy — Altman preserved YC's cultural essence while dramatically expanding its scale.

altman-model-yc-trianglealtman-model-power-law-thinking

2015-12

Co-Founded OpenAI, Beginning the AGI Mission as a Non-Profit

Context: In 2015, deep learning had demonstrated unprecedented capabilities, Google acquired DeepMind, and the AI arms race was beginning to take shape. Altman and Musk worried that AGI might be developed in closed corporate labs without public oversight.

Decision: Co-founded OpenAI with Elon Musk, Greg Brockman, and others as a non-profit, committing to open research results and securing $1 billion in initial pledged donations.

Reasoning: If AGI is inevitable, the most important question is who develops it and with what values. An organization with a mission of human benefit and open research can better ensure safety and fairness than closed corporate labs.

Outcome: After its founding, OpenAI quickly attracted the world's top AI researchers, released groundbreaking models including the GPT series and DALL-E, and ultimately created ChatGPT — changing the global understanding of AI.

Lesson: Establishing a mission-driven organization in the early stages of technological change can secure a triple advantage in talent, resources, and moral authority in the field.

altman-model-optionality-trap

2019-03

Became OpenAI CEO, Same Year Secured $1 Billion Strategic Investment from Microsoft

Context: In 2019, OpenAI released GPT-2, whose capabilities stunned the research community but also raised profound AI safety concerns. Meanwhile, OpenAI realized that training larger models required massive compute investment that the non-profit structure could not sustainably support.

Decision: Took over as OpenAI CEO, reformed the organizational structure to a 'capped-profit' model, and signed a strategic partnership with Microsoft, securing $1 billion in investment in exchange for Azure cloud computing resources.

Reasoning: Training AGI-level models requires billions in compute investment; the non-profit structure couldn't attract sufficient capital. The 'capped-profit' model allows external investment while preserving mission-first governance.

Outcome: Microsoft's investment and Azure compute support enabled OpenAI to train GPT-3 and subsequent larger models, laying the technical foundation for ChatGPT's success.

Lesson: Mission and commerce can coexist but require carefully designed governance structures — 'capped profit' is an innovative organizational experiment, though its long-term effectiveness remains debated.

altman-model-power-law-thinking

2020-06

GPT-3 Released, Demonstrating the Emergent Capabilities of Large Language Models

Context: In 2020, the AI research community's understanding of Scaling Laws was deepening — larger models, more data, and more compute seemed to continuously yield capability improvements. GPT-3's 175 billion parameters made it the largest language model at the time.

Decision: Released GPT-3 via API, allowing developers to call through an interface rather than open-sourcing model weights, achieving commercialization while controlling access.

Reasoning: Full open-sourcing might lead to model misuse; the API model could allow the developer ecosystem to flourish while maintaining monitoring and control over model usage.

Outcome: GPT-3 sparked enthusiasm from developers worldwide, with thousands of applications built on it, proving the commercial value of large language models and laying the foundation for ChatGPT's product form.

Lesson: Scaling Law is one of the most important empirical laws in AI — Altman and OpenAI believed earlier than most that continuous scaling could bring qualitative changes, making this their most important technological bet.

altman-model-power-law-thinking

2022-11

ChatGPT Launched, Becoming the Fastest-Growing Consumer Product in History and Triggering a Global AI Race

Context: By 2022, OpenAI had GPT-3.5 and InstructGPT, but AI was still a tool for specialist circles. Altman decided to open it to the public through a conversational interface to lower the barrier to use.

Decision: Released ChatGPT as a free, conversational interface with no usage barriers, allowing anyone to directly converse with AI.

Reasoning: Getting as many people as possible to use AI is the core of the 'iterative deployment' strategy — large-scale real user usage exposes problems and provides feedback more effectively than closed testing. Meanwhile, public AI use is a way for society to adapt in advance to AGI's arrival.

Outcome: ChatGPT surpassed 1 million users in just 5 days and 100 million in two months, becoming the fastest-growing consumer product in history. Google, Microsoft, Baidu, and others accelerated their AI product launches, officially starting the global AI race.

Lesson: Product form (conversational interface) is crucial for AI democratization — with the same technical capabilities, the conversational interface let ordinary people experience AI's real value for the first time; this was ChatGPT's core design decision.

altman-model-ai-society-impact

2023-11

Fired by OpenAI Board, Reinstated Five Days Later, Restructured Governance

Context: In November 2023, on the eve of ChatGPT's first anniversary, OpenAI's non-profit board suddenly announced the firing of Altman, citing 'not consistently candid with the board' without providing specific details.

Decision: Refused to accept the firing, actively communicated with investors and employees to fight for reinstatement. Over 700 OpenAI employees signed a letter demanding the board's resignation; Microsoft expressed willingness to hire the entire OpenAI team.

Reasoning: Altman believed his reinstatement was crucial for the continuation of OpenAI's mission — if he left, the best researchers would also leave, and OpenAI's AGI mission could not be achieved.

Outcome: Five days later, most of the original board members resigned, Altman was reinstated as CEO, and the new board included more business figures. This event exposed fundamental contradictions in OpenAI's non-profit governance structure and ultimately pushed toward a transition to a fully for-profit company.

Lesson: Organizational governance structures must match commercial realities — OpenAI's non-profit governance structure produced irreconcilable contradictions when facing billions in commercial interests.

altman-model-optionality-trap

2024-05

GPT-4o Released, Future Positioning of AI and Identity Verification Continues

Context: In 2024, AI capabilities continued to advance rapidly, with multimodal becoming the mainstream direction. Meanwhile, Worldcoin (later rebranded to World), which Altman co-founded, promoted iris scanning in exchange for cryptocurrency globally, laying out digital identity infrastructure for the AGI era.

Decision: Released GPT-4o (omni model), achieving seamless integration of text, voice, and images, while advancing the World project to build a global-scale 'proof of humanity' system.

Reasoning: In the AGI era, distinguishing humans from AI will become a critical infrastructure need — Worldcoin/World is Altman's advance positioning for this need, also related to his long-term interest in universal basic income.

Outcome: GPT-4o became one of OpenAI's most popular models to date; the World project operates in over 100 countries globally with more than 10 million registered users, but also faces regulatory investigations in multiple countries.

Lesson: Leading tech companies need to simultaneously position across multiple time horizons — GPT-4o is the current competitive product, while World is infrastructure prepared for the post-AGI era.

altman-model-ai-society-impact

Reading List

Books

Recommended by (4)

Zero to One: Notes on Startups, or How to Build the Future

Peter Thiel · 2014

Altman has recommended this book in multiple talks and interviews, believing Thiel's 'secret' theory aligns strongly with his contrarian bet framework. He listed it as required reading for founders during his YC presidency.

Amazon 当当

The Hard Thing About Hard Things

Ben Horowitz · 2014

Altman has mentioned this book multiple times at YC founder sessions, believing Horowitz's honest description of startup difficulties is the most accurate portrayal of startup reality, far more valuable than success-formula books.

Amazon 当当

Superintelligence: Paths, Dangers, Strategies

Nick Bostrom · 2014

Altman mentioned this book's influence on his understanding of AGI risks during his Lex Fridman interview, believing Bostrom's analytical framework is an important reference for thinking about AI's long-term impact. This book is also part of the intellectual background for OpenAI's founding.

Amazon 当当

Poor Charlie's Almanack

Charlie Munger · 2005

Altman has mentioned Charlie Munger's multidisciplinary mental models' influence on him in his blog and interviews, believing Munger's 'latticework theory' is an important tool for evaluating startups and AI development paths.

当当

Influence Network

Origins, Contemporaries & Legacy

Influenced By

Paul Graham · Mentor and Intellectual Heritage

Paul Graham was Altman's first mentor; YC's startup philosophy profoundly shaped Altman's entrepreneurial methodology. Graham personally chose Altman as YC's successor — the highest endorsement.

Elon Musk · Co-Founder and Intellectual Collision

Musk was OpenAI's co-founder; his deep concern about AI existential risks and obsession with super-missions resonated strongly with Altman's AGI sense of mission. Despite later divergence on OpenAI's direction, Musk's influence on Altman remains significant.

Peter Thiel · Contrarian Thinking Inspiration

Thiel's 'secret' theory — the most important opportunities are truths others haven't yet discovered — aligns strongly with Altman's contrarian bet framework. Thiel's Silicon Valley investment network also significantly influenced Altman's early career.

Influenced

Greg Brockman · Mission Transmission and Organization Building

Greg Brockman, as OpenAI's president and co-founder, under Altman's influence translated the AGI mission into concrete organizational execution. Altman's strategic vision complemented Brockman's engineering execution.

Global AI Startup Founders · Entrepreneurial Methodology Dissemination

During Altman's tenure as YC president, his startup evaluation framework and methodology influenced thousands of founders. His blog post 'How to Be Successful' is widely cited by entrepreneurs worldwide and has become an important reference for startup methodology.

Co-thinkers

Ilya Sutskever · Technical Vision Resonance

Sutskever was OpenAI's co-founder and chief scientist; his technical conviction about Scaling Laws and AGI pathways closely aligned with Altman's strategic bets. Despite briefly standing on opposite sides in the 2023 board crisis, their core technical visions remained resonant.

Satya Nadella · Strategic Partner

Microsoft CEO Satya Nadella is Altman's most important business partner. Nadella's public support for Altman during the 2023 board crisis, along with Microsoft's billions in investment in OpenAI, are key pillars of OpenAI's commercial success.

Peer Reviews

Sam has this ability to hold two completely contradictory ideas in his head at the same time and somehow make them both work. He genuinely believes OpenAI is building the most dangerous technology in history, and he's also the most enthusiastic person about building it.
Greg Brockman · Profile of Sam Altman, The New Yorker, October 2023

Sam is the most effective person I have ever met at taking something that seems impossible and making it seem inevitable.
Paul Graham · Paul Graham on Sam Altman, various interviews, 2019-2023

正在打开人物节点

Sam Altman

Core Knowledge Graph

Core Beliefs

Compound Yourself: Exponential Personal Growth Is the Most Important Life Strategy

High-Conviction Contrarian Bets: The Most Important Opportunities Hide in the Opposite of Mainstream Consensus

Super-Mission Driven: AGI Is the Most Important Technology in Human History and Must Be Developed to Benefit All of Humanity

Iterative Public Deployment: Discovering and Correcting AI Risks Through Real-World Use Rather Than Closed Testing

Mental Models

YC Startup Triangle Evaluation Framework

Power Law Thinking: Bet on a Few Extreme Outcomes

Avoid the Optionality Trap: Real Success Comes from High-Conviction Single Bets, Not Spreading to Preserve Options

AI Power Structure Analysis Framework: How Technological Change Reshapes Social Power Distribution

Values & Paradoxes

Tension Between Open Mission and Commercial Expansion

Dual Identity as Accelerationist and Safety Advocate

Board Crisis: Fired and Reinstated by the Organization He Co-Founded

Evolution Phases

Serial Entrepreneur Phase

YC President Phase

OpenAI CEO Phase

9 Key Events

Dropped Out of Stanford, Founded Loopt and Joined YC's First Batch

Joined Y Combinator as a Partner

Became President of Y Combinator, Succeeding Paul Graham

Co-Founded OpenAI, Beginning the AGI Mission as a Non-Profit

Became OpenAI CEO, Same Year Secured $1 Billion Strategic Investment from Microsoft

GPT-3 Released, Demonstrating the Emergent Capabilities of Large Language Models

ChatGPT Launched, Becoming the Fastest-Growing Consumer Product in History and Triggering a Global AI Race

Fired by OpenAI Board, Reinstated Five Days Later, Restructured Governance

GPT-4o Released, Future Positioning of AI and Identity Verification Continues

Books

Recommended by (4)

Origins, Contemporaries & Legacy

Influenced By

Influenced

Co-thinkers

Peer Reviews