In an interview with Stanford's Graduate School of Business in October 2026, OpenAI's chief executive, Sam Altman, had blunt advice for people starting their careers. "My chronic disappointment with MBA students is that they always want to go do the thing that was cool five years ago," he said, according to Fortune. "I think this is a great time to go do the new thing."
The remarks spread quickly, and "OpenAI Sam Altman young workers advice" became one of the fastest rising technology searches on Google Trends this week. The interest is understandable. Young people want to know whether the jobs they trained for will still exist, and the head of the company most associated with AI is an obvious person to ask.
But a chief executive's advice is one input. This piece puts it next to what the research actually shows about AI and entry-level work, then looks at the other side of the same question: what it means for the businesses deciding whether to hire juniors at all.
What Altman actually said
Fortune's report has three main points.
- AI will take over a lot of tasks. "It's totally true that huge numbers of tasks that people do today will be done better, cheaper, and faster by AI," Altman said. He added that the shift will come more slowly than many people expect.
- The opportunity is in new things. He argued that many companies can be built on top of the technology OpenAI has released, including ones that would have been hard to build a few years ago.
- Adaptability matters more than credentials. He described the people best placed for this period as those with drive, ideas and a new way of looking at things, who can adapt while the ground shifts.
Some context helps. Altman left Stanford in 2005 to co-found a start-up, then ran the start-up accelerator Y Combinator from 2014 to 2019, so "go build something" is advice he has given in one form or another for two decades. And it lands at a time when Americans are already starting businesses at record rates: Fortune notes that 5.7 million applications to start a new business were filed in 2025, the highest annual total in the two decades the Census Bureau has tracked.
What the data says about AI and entry-level jobs
The research is more mixed than either the optimists or the doom-sayers suggest, and the two most cited studies point in different directions.
Young workers in AI-exposed jobs are being hired less. Economists at Stanford's Digital Economy Lab have tracked millions of US workers through ADP payroll data. Their latest update, revised in August 2026, finds that employment of 22- to 25-year-olds in the occupations most exposed to AI is 19% below where it would be had it kept pace with less-exposed peers, up from 13% a year earlier. Experienced workers in the same occupations show no such gap. Two details matter. The gap comes mainly from fewer hires, not more layoffs: companies are not firing young staff, they are taking on fewer of them. And it is concentrated in jobs where AI replaces tasks outright; where AI mainly helps people do their work, employment is flat or rising.
But there is no spike in graduate unemployment yet. A September 2026 NBER working paper by Robert Fairlie and Jane Wu looked at recent college graduates in the summer of 2026 and found that their unemployment did not rise compared with previous summers, or compared with older graduates and young workers without degrees. The New York Fed puts the unemployment rate for recent graduates at about 5.6% and their underemployment at 42% in the second quarter of 2026: high, but not a collapse.
Both can be true at once. A narrowing of hiring in particular occupations can happen without showing up in the headline unemployment rate, because graduates move into other work. The Stanford authors are careful to call their results early warning signs, "canaries in the coal mine", rather than proof that AI caused the gap.
What this means for young people
Read together, Altman's advice and the data suggest something more specific than "start a company".
The tasks at risk are the routine ones. The jobs where young workers are losing ground are those where AI can do the work, not merely help with it. If a role is mostly drafting standard documents, answering routine queries or writing boilerplate code, it is the role that is shrinking, whoever fills it.
Using AI well is becoming the entry ticket. Where AI helps people work rather than replacing them, employment is holding up. Being the junior who knows how to check, correct and direct an AI tool is a better position than being the junior competing with it.
Starting something is one option, not the only one. Altman's audience was Stanford business students. For most graduates, the practical version of his advice is to choose roles where judgment, responsibility and contact with customers are the core of the job, because those are the parts AI does least well.
What it means for employers
The Stanford finding is really a finding about employers. The gap for young workers is created by companies deciding not to hire them. Before your business makes the same decision, three things are worth weighing.
A team with no juniors has no seniors in five years. Entry-level roles are where people learn the business. If AI absorbs that work and nobody is hired to learn it, the experienced staff of 2031 do not exist. Some companies will find that out the expensive way.
Automating a task is not the same as automating a job. Most junior roles are a bundle of routine tasks and judgment calls. Taking the routine part away can leave a better job, focused on the judgment, rather than no job at all. The research points the same way: where AI complements people, employment is stable or growing.
Train people to work with the system you are building. If you are bringing AI into your operations, the people who will run it day to day need to understand what it does well, where it fails and when to override it. Juniors who learn that from the start are an asset, not a cost the system replaced. Our guide on whether your business really needs AI covers how to pick the work worth automating in the first place.
Altman's advice was aimed at people choosing a career. The data says the bigger decision belongs to the companies choosing whom to hire, and that the businesses which keep training people alongside their AI systems will be the ones with a workforce left to lead them.