The man who helped build the software that runs most of the world’s computers thinks the people who write that software will still have jobs a hundred years from now. That’s not a hedge. That’s a deliberate prediction from someone who has spent more time thinking about the intersection of technology and human labor than almost anyone alive.
Bill Gates has been adding to a short list of careers he believes AI won’t replace, and the additions keep coming. He started with three, then publicly offered a fourth one in a setting that didn’t exactly scream policy forum: a late-night TV couch. The fourth one made the audience laugh. But the three that came before it, and the broader picture Gates painted of where human work is actually heading, are worth taking seriously.
The question he’s really answering isn’t “which jobs are safe?” It’s something harder: in a world where intelligence itself becomes free and abundant, what does human work even mean?
The Tonight Show Moment That Wasn’t a Joke
Gates made his position clear during an interview with Jimmy Fallon on NBC’s The Tonight Show in February 2025. He was there to promote his memoir, Source Code: My Beginnings, but the conversation kept pulling toward AI. According to CNBC’s reporting on the appearance, Fallon asked him directly whether humans would still be needed. Gates responded, “Not for most things,” a line that landed in a studio full of nervous laughter.
He didn’t walk it back. Instead, he made a distinction. “There will be some things we reserve for ourselves,” he said. “But in terms of making things and moving things and growing food, over time those will be basically solved problems.”
Then came the baseball line. Gates noted that people probably don’t want to see machines playing baseball, and that some activities would remain human by human choice, not necessity. It got a laugh. But embedded in that joke was the fourth career on his list: professional athletes. Gates added athletes as a fourth example during the Tonight Show appearance, noting that “we won’t want to watch computers play baseball.”
The distinction is worth holding onto. Gates wasn’t saying athletes are safe because AI can’t replicate physical performance. Robotics is advancing fast enough that the physical challenge of hitting a 95-mph fastball may not be permanently beyond automation. He was saying people won’t want to watch machines compete. Sport isn’t just about peak performance. It’s about human drama, human failure, human stakes. That’s a different kind of job security, and arguably a more durable one.
The Original Three: Why These Specific Careers
Before the baseball moment, Gates had already named three professions he believed would remain indispensable. According to The Express Tribune, coders, biologists, and energy specialists made his initial list of roles that won’t be replaced by generative AI. His reasoning for each is distinct enough that it’s worth unpacking them separately.
Coders. This is the one that surprises people the most, given how much AI has already disrupted software development. In an interview with France Inter, as reported by Windows Central, Gates made the claim that programming will remain a 100% human profession, even 100 years from now. His argument isn’t that AI can’t write code. It clearly can, and does. He argues that humans are essential for identifying and correcting errors, refining algorithms, and bolstering AI development, and that AI-powered tools cannot replicate human creativity and judgment in that process. The deeper logic is that the more AI there is in the world, the more humans are needed to build, oversee, and interrogate it. Coders don’t get replaced by AI. They get redirected by it.
Biologists. Gates argues that while AI is great at diagnosing diseases and analyzing DNA, it lacks the creativity required for biological research and scientific discovery. Breakthroughs in biology rarely come from processing existing data more efficiently. They come from asking questions nobody thought to ask before, from connecting ideas across wildly different disciplines, from the kind of intuitive leap that still seems to be a fundamentally human capacity. AI can accelerate the work. It can surface patterns in genomic datasets that a human team would take years to find. But the question of what those patterns mean, and what to do with them, still requires a biologist in the room.
Energy professionals. Gates said energy professionals would continue to be in high demand, noting that energy experts are essential in addressing the unpredictable needs of a changing global climate. The energy sector is shaped not just by physics and engineering but by geopolitics, regulation, climate pressure, and sudden crises. Gates claims that while AI can help with analysis and efficiency, human judgment is paramount when it comes to decision-making, especially in energy and especially in crisis management. An algorithm can optimize a grid. It can’t navigate the political fallout of a nuclear plant going offline, or make the call when a hurricane is eight hours from a Gulf Coast refinery.
The Broader Vision: “Free Intelligence”
The list of safe jobs only makes full sense when you understand the larger picture Gates has been building. He described a near future where expertise, currently “rare,” pointing to “a great doctor” or “a great teacher,” will with AI over the next decade become free and commonplace. He called this era “free intelligence,” and the implications are significant. The result will be rapid advances in AI-powered technologies that touch nearly every aspect of life, from improved medicines and diagnoses to widely available AI tutors and virtual assistants.
On the surface, that sounds optimistic. And Gates frames it that way. But “free intelligence” is also a way of saying that scarcity, the scarcity that makes human expertise economically valuable, is going away. Right now, being a talented diagnostician or a skilled tutor commands a salary partly because there aren’t enough of them. When AI can deliver that same quality of output to anyone with a phone, the economic argument for paying a human to do it gets a lot harder to make.
Gates isn’t being nihilistic here. He has consistently framed AI as something that will solve enormous problems, including the global shortage of healthcare workers and educational inequality. The concern isn’t cruelty. It’s math.
The careers that remain valuable in this framework are the ones that require something AI cannot yet replicate at any price: genuine creative judgment, ethical decision-making under uncertainty, and the human meaning that spectators and patients and collaborators bring to certain kinds of work. You can browse jobs that AI is already reshaping to see exactly which roles are most exposed right now.
What Other Tech Leaders Think
Gates is not the only voice in this conversation, and not all of them agree with him. NVIDIA CEO Jensen Huang has predicted that coding might be dead as a career because of AI’s prevalence, and has advised younger generations to pursue biology, manufacturing, education, or farming instead. That’s almost the exact inverse of Gates: where Gates says coders are safe and biology is safe for different reasons, Huang treats coding as the most vulnerable profession and biology as the refuge.
Anthropic CEO Dario Amodei has claimed that AI has a high propensity to cut entry-level white-collar jobs by 50%, leaving many early-career workers without a foothold. That’s a much darker short-term view than Gates holds. Gates acknowledged in his France Inter interview that fear of AI is reasonable and that he shares some of it, but his overall position is more optimistic on the net outcome.
These disagreements matter because they reflect genuine uncertainty. Nobody, including the people who built these systems, knows exactly how fast the capabilities will grow or where the floor on human irreplaceability sits. What they’re all doing is pattern-matching from what they know toward futures they can only partially model.
The Scale of What’s Coming
The World Economic Forum’s Future of Jobs Report 2025 found that job disruption will affect 22% of all jobs by 2030, with 170 million new roles set to be created and 92 million displaced, resulting in a net increase of 78 million jobs globally. That net positive number is real, but it papers over something important. The 92 million people displaced and the 170 million jobs created are not the same people in the same places doing the same kinds of work. The skills gap is already the most significant barrier to business transformation, with nearly 40% of skills required on the job set to change. Technology skills in AI, big data, and cybersecurity are expected to see rapid demand growth, but human skills such as creative thinking, resilience, and flexibility will remain critical.
The energy sector Gates flagged as safe is, interestingly, one of the areas the WEF projects will see the most new roles created, driven by the green transition. Renewable energy engineers, grid specialists, and climate-adaptation professionals are among the fastest-growing job categories globally. Gates’ instinct that energy expertise is both too complex and too consequential for full automation aligns with where the data is pointing.
The biology piece is consistent too. Healthcare, care economy roles, and scientific research roles are all listed as growth areas in the WEF data, precisely because they require the kind of judgment, creativity, and human relationship that resists full automation.
What This Actually Comes Down To
The framing of “which jobs are safe” is understandable. It’s the question most people are actually asking when they read about Bill Gates AI jobs predictions. But it’s also a slightly misleading frame. The more accurate question is: which aspects of a job require genuine human judgment, creativity, or meaning-making that AI cannot substitute for? Because those aspects will persist. The job titles around them may change dramatically.
Gates’ four categories, coders, biologists, energy specialists, and professional athletes, share a structural feature that doesn’t get named explicitly in most coverage of his remarks. In each case, what’s irreplaceable isn’t a skill set per se. It’s either the capacity to supervise and interrogate AI itself (coders), the creative leap that generates genuinely new knowledge (biologists), the judgment required in high-stakes, politically complex, real-world systems (energy specialists), or the human meaning that the audience brings to the performance (athletes). Those are not job descriptions. They’re descriptions of what humans bring to work that AI doesn’t bring.
That framing is more useful than a list, and more honest about the uncertainty involved. Gates himself has said his predictions may not be accurate, that the industrial revolution and the rise of the internet reshaped work in ways nobody fully anticipated at the time, and that AI will do the same.
Read More: Meta just changed how Facebook works — here’s what the new ‘AI Mode’ is all about
The Honest Answer Nobody Has Yet
The list of four careers is a starting point for thinking about value in the AI era, not a final answer. What it captures well is that value in the coming decades will concentrate where human judgment, creativity, and meaning are genuinely irreplaceable, not just where they’re currently cheaper than the alternative.
For most white-collar jobs that aren’t on that list, the honest answer isn’t “you’re safe” or “you’re not.” It’s that the parts of your work that are repetitive, predictable, and data-processable are going away, and whether what remains is enough to constitute a full-time job is a question that nobody, including Gates, has a clean answer to yet. The industrial revolution didn’t end work. It ended certain kinds of work, and created new ones that nobody had names for. The same is almost certainly true here. The uncomfortable part isn’t the change itself. It’s that the timeline for when it arrives at your specific desk is genuinely unknown.
AI Disclaimer: This article was created with the assistance of AI tools and reviewed by a human editor.