In a Y Combinator interview published in November 2024, Sam Altman, CEO of OpenAI, talks about starting technology companies, conviction, and the ideas in his essay on the age of intelligence. This post summarizes the video. The quotes are from auto-generated captions; the forecasts are Altman’s own opinions, not facts.
Key Takeaways #
- Best Time for Startups: Asked whether this is the best time ever to start a technology company, Altman answers “let’s at least say it’s the best time yet”, because each technological revolution lets you do more than before.
- Conviction Matters: He says most of the world still does not understand the value of “a fairly extreme level of conviction on one bet”, and adds that conviction is great only until you have data one way or the other.
- The Age of Abundance: He says abundant intelligence and abundant energy are the two key inputs to almost everything else, and that the outcomes in his essay are “maybe not that far away”.
- Learning from Mistakes: He says OpenAI “made plenty of mistakes along the way” and that conviction without adapting to results is not effective.
- AI’s Future: The interviewer reads his essay as saying superintelligence (ASI) is thousands of days away; Altman answers “that’s our hope, our guess, whatever”. It is a guess, not a forecast anyone has verified.
The Best Time To Start A Tech Company #
Asked whether this is the best time ever to start a technology company, Altman said: “let’s at least say it’s the best time yet, hopefully there’ll be even better times in the future”. His reasoning: with each successive major technological revolution you have been able to do more than you could before.
The Power of Conviction #
Altman said most of the world still does not understand the value of “a fairly extreme level of conviction on one bet”, and tied it to OpenAI’s early choice to pick one approach and concentrate on it rather than spread resources across many bets. He added that OpenAI said from the beginning it was going after AGI, at a time when saying so seemed “possibly crazy and borderline irresponsible”. He also set a limit: conviction is great “until the moment you have data one way or the other”, and people who hold on past that moment are not effective.
The Age of Abundance #
The interviewer lists what the essay says to look forward to: fixing the climate, establishing a space colony, discovering all of physics, near-limitless intelligence and abundant energy. Altman replied that he thinks all of those, and probably more, are “maybe not that far away”. These are his predictions. The interview does not show evidence for them or a date.
Learning From The Journey #
Altman said the startup journey needs iteration: there are no adults in the room who will tell you exactly what to do, so you iterate quickly and find your way. On OpenAI he said it had “made plenty of mistakes along the way” and “done a few things really right”. On his first startup, Loopt, he said it felt like time wasted, but that he does not regret it.
The Future of AI #
The interviewer says the essay claims superintelligence is thousands of days away. Altman called that “our hope, our guess”, and described a path where the rate of progress of the last three years continues for several more, with nine years being “like 3500 days or whatever”. He said a system at that point would be quite capable. That is a conditional guess, not a date: it assumes the rate of improvement continues.
Conclusion #
Three points are Altman’s own words in this interview: it is “the best time yet” to start a tech company; conviction pays only until the data arrives; and “thousands of days” is his guess for superintelligence, not a measured forecast.
