Nov 8, 2018 · 35m · y-combinator
Sam Altman on Choosing Projects, Creating Value, and Finding Purpose · Y Combinator
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this Y Combinator podcast episode, host Craig Cannon interviews OpenAI leader Sam Altman on core principles derived from his essays, covering cognitive load management, decision-making under uncertainty, institutional design, and the long-term impact of AGI. Altman shares actionable frameworks for strategic project selection, re-evaluating career risk, and aligning personal focus with ambitious, high-value pursuits.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The partners as informed peer 2.4 Guest teaching 2.5 Guest disagreement 2.1 The partners pushing back 1.1
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 26:34 The deferred life plan takedown
Sam strongly ridicules founders who claim they will run a crypto hedge fund for a few years just to fund a rocket startup, calling it inauthentic and virtually guaranteed to fail.
Hardest push from the partners ▶ 13:22 Questioning Modafinil-induced productivityCraig challenges Sam's passing remark about achieving 20 productive hours via Modafinil, questioning whether that state represents Sam's optimal intellectual output.
Biggest teaching moment ▶ 22:12 Reframing the state of human innovationSam reframes the host's premise regarding the alleged death of deep technological innovation by stating that creating AGI will exceed the total sum of all innovation in human history.
The partners hold their own ▶ 34:18 Debunking squat-to-cycling fitness mythsCraig steps into an expert stance on physical conditioning, dismantling naive internet claims that high squat numbers translate directly into cycling endurance without understanding power-to-weight ratios.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| Minimizing Cognitive Load and Stepping Back | 2 | 2 | 1 | 1 | Craig quotes Sam Altman's essay on minimizing cognitive load and asks how to practically step back. Sam explains his process of downtime reflection and his annual New Year's Eve review ritual. | |
| Metrics, Exploration, and Taking Time Off | 2 | 3 | 2 | 1 | Craig asks about standardizing metrics across various life projects. Sam rejects rigid universal rubrics, emphasizing an 'explore mode' where one accepts widespread failure while hunting for outlier successes, citing his gap year after selling his startup. | |
| Decisiveness and Brutal Honesty in Project Selection | 2 | 3 | 2 | 1 | Craig asks how to be decisive and avoid merely following trendy tech sectors. Sam stresses the necessity of brutal self-honesty to distinguish genuine traction from wishful thinking, referencing lessons learned from poker and angel investing. | |
| Physical Foundations and Maximizing Peak Focus | 2 | 2 | 2 | 2 | Craig discusses physical foundations for productivity and pushes back gently when Sam jokes about 20-hour productive days on Modafinil. Sam dismisses claims from people asserting they can be genuinely productive for 16 hours a day. | |
| Scientific Research Philosophy and Managing OpenAI | 3 | 3 | 2 | 1 | Craig brings up the Richard Hamming lecture and asks about managing disparate research paths. Sam explains OpenAI's broad North Star mandate and discusses how most conventional advice errs toward excessive risk aversion. | |
| Productivity Systems and Avoiding Excessive Complexity | 3 | 2 | 3 | 1 | Craig reviews Sam's three-pillar productivity framework, and Sam takes aim at over-engineered productivity systems, arguing that complex systems rarely correlate with high real-world impact. Craig adds a reference to an accountability email model. | |
| Innovation Frontiers and New Funding Mechanisms | 2 | 3 | 2 | 1 | Craig asks about the stagnation of innovation and future funding mechanisms. Sam pushes back on techno-pessimism, arguing that achieving AGI would outweigh all historical innovation, while defending the internet frontier as the natural locus of talent. | |
| Managing Interruption and Openness to Unplanned Opportunities | 3 | 3 | 4 | 1 | Sam delivers an extended rant against the 'deferred life plan,' criticizing founders who start crypto funds under the pretext of funding future aerospace projects. Craig validates and expands on the point with his own examples of developer side projects. | |
| Human Purpose and the Impact of AGI | 2 | 3 | 2 | 1 | Craig quotes Sam's essay 'The Merge' regarding human intelligence and self-worth. Sam reflects on how the arrival of AGI will diminish the societal premium placed on biological intelligence, shifting purpose toward human relationships and value creation. | |
| Personal Fitness Routine and Heavy Lifting | 3 | 1 | 1 | 1 | Craig and Sam bond over their shared shift toward heavy strength training over cardio. Craig demonstrates his own knowledge by dissecting power-to-weight misconceptions common in online fitness communities. |
Statements from this episode (20)
Altman: Daily cognitive output is a fixed budget easily drained by trivia
“I've at least found for myself, I have like a like fixed budget of cognitive output per day, and I can spend that on whatever, but if I let it go on unimportant stuff, then I never have time to get to the
The really important stuff.”
Altman: Annual December 31 review ritual reduces daily planning stress
“One specific thing I do, which has sort of become this annual ritual I really like, is December 31st every year, I take a break from vacation, and I, like, sit and I write down what went well the last year, what didn't. I look at, like, how I did on my to-do l…”
Altman: Explore widely and cheaply, then ruthlessly cut to top projects
“The fundamental pattern that has always worked for me is take time, explore a lot of things, try a lot of things, try to have like a beginner's mind about what will work and what won't work. But trust your intuitions, pursue a lot of things as cheaply and quic…”
Altman: Taking a gap year between jobs was a top career decision
“It's an incredibly privileged thing to be able to do this, but if you are in the position between jobs where you can take a year off, I highly recommend it. I think it was like one of the two or three best career things that I ever did.”
Altman: Project selection requires brutal honesty about what is truly working
“And this is where I think it takes brutal honesty with yourself about which projects are actually working and which you would like to convince yourself are working.”
Altman: Poker and angel investing taught him the most about business
“The things that I think have taught me a lot about, like, Business are poker and angel investing. So I recommend both of those.”
Altman: Poker winnings funded his college living expenses
“It did not fund my startup, but it funded my, like, living expenses as a college student.”
Altman doubts people who claim they only need four hours of sleep
“These people who claim they only need to sleep four hours a night, it's true. I suspect it's not. I suspect it's not.”
Altman: Skimping on sleep is a terrible trade for creative breakthroughs
“If you're trying to like win on really big creative ideas much better off, like cut something, have less of them, whatever. But like for me skimping on sleep, which I have tried at other points in my life is a terrible trade.”
Altman: OpenAI is one of the most interesting research labs in decades
“And I do think we've done one of the more interesting research labs of the last couple of decades. The world isn't set up to support this anymore, and I, and it's a very interesting question about why.”
Altman: Research strategy should back intuition, then double down on what works
“Like, try lots of things, let people follow their intuitions, and then when something starts working, put a lot more resources behind it. That's worked for us really well.”
Altman: Humans Focus on Short-Term Catastrophes Over Long-Term Chronic Risk
“We're very good at thinking about
Short-term catastrophic risk, and we're very bad about thinking about long-term chronic risk.”
Altman: Vast Majority of People Try to Discourage High Ambition
“Basically all people, almost like, 98% of people in the world will try to pull you back and say, it seems a little bit too crazy, a little bit too out there, a little bit too ambitious.”
Altman: People obsessed with complex productivity systems rarely move the world forward
“I think this whole, like, productivity porn thing where you have these, like, systems and, like, you know, like, these multivariable charts and three-dimensional whatever's Like, those people never seem to be the people who, like, really move the world forward…”
Altman: Building AGI at OpenAI outweighs all past human innovation combined
“If this one thing works, like, if we do actually build AGI at OpenAI, that is more important than all of the innovation sum total in humanity so far. So, clearly not has all the innovation happened.”
Altman: Taking random meetings pays off despite 90% being wasted time
“One is you take a lot of random meetings, and you accept that 90% of the time you will waste your time and be frustrated with yourself for wasting your time for that one meeting that makes it all worth it net.”
Altman: Cutting off inbound from outside your network is disastrous
“I think if you cut it off entirely, it's disastrous. Like, if you don't have any, like, flow of, like, People outside your network with new ideas or just a lot of talent that is under appreciated. It's really bad.”
Altman: The 'deferred life plan' almost never works
“One of the many problems with the deferred life plan is everyone can kind of tell that that's what you're on. And you're not that serious about, you're not that committed to either. And it just never works. I mean, it must work sometimes, but I've seen it fail…”