June 1, 2026
Jay Burstein, BAAI Fellow
I. The Fear
As artificial intelligence adoption becomes more widespread, the fear that AI will replace jobs is intensifying throughout our public discourse right now. It is, frankly, one of the most dominant narratives that we’re hearing. But the data tells a very different story than what much of the media and many prominent voices are communicating about the future of work.
II. What The Data Actually Shows
David George, a General Partner at Andreessen Horowitz, explained in a recent interview what he found when looking at the data: “If you look at what management teams have said about AI in earnings calls and transcripts, AI as augmentation out-mentions AI as substitution on these earnings calls by a factor of 8 to 1.” This is contradictory to the dominant narrative of AI replacing jobs. In reality, companies are viewing AI as a productivity enhancement to help employees fulfill and take on more tasks.
One just needs to look at examples where AI could be causing a negative impact, such as project management. Interestingly, Bandan Singh highlighted through a Substack post that product management postings have increased by over 7% month-to-month in the last year.
Another area that data is showing us a “boom” because of AI is with new apps. Appfigures, a market intelligence provider, found that in the first quarter of 2026, app releases globally expanded by 60% year-over-year on the App Store and Google Play. As Sarah Perez of TechCrunch highlighted, “AI will make it easier for anyone to create apps, driving a rebirth of the App Store. The new app gold rush could be led by creators who have ideas but not the technical skills to design mobile software.” AI isn’t just creating new jobs; it’s leading to expanding opportunities that didn’t exist.
Notably, a Bureau of Labor Statistics report published in early 2026, illustrated software development roles projected to experience nearly 16% growth year-over-year through 2034, which is 5x over the national average of other industries. We’re already seeing this projection taking shape. From April 2025 to April 2026, the number of software developers employed increased.
If AI was actually eliminating jobs, we would see it in the data already. Software development and product managers, in theory, might be most negatively impacted by AI. As many have hoped, employment and opportunities are, in fact, expanding as AI gains adoption.
III. Why This Happens (The Jevons Paradox Explained)
William Stanley Jevons, a famous economist explained in The Coal Question how as coal got cheaper, consumption went up. New industries that were once inconceivable became a part of daily life. This pattern of ‘Jevons Paradox’ has repeated itself in other areas, such as agriculture and with spreadsheets. As tractors were invented, new areas of farming grew, leading to greater efficiency, effectiveness, and economic output. Spreadsheets did not eliminate accounting and finance jobs, but instead led to more time being shifted towards strategizing and using the spreadsheets to drive decision making. Less time was spent on administrative and mundane tasks.
It can be expected that as AI grows in demand, costs drop and new work will emerge. An example of this could be with a marketing team. Rather than spending considerable time on execution, AI will allow more time to focus on strategy, creative ideation, and managing/building relationships with clients.
As George articulated in his essay, “The frontier of useful work expands and I don’t see any reason why that’s going to be different in this case. The early AI data looks consistent with that pattern. We’re not yet seeing an employment impact over the last few years, ever since we’ve had these tools.” History shows a reasonable expectation that AI will follow this pattern.
IV. The Narrative is Shifting
Anthropic’s CEO, Dario Amodei, famously offered what many felt was a stark view of the future last year. At that time, he predicted AI would eliminate 50% of white-collar jobs at the entry-level and cause 10 to 20% unemployment. A year later, the narrative is shifting. Prominent voices like Amodei are now citing Jevons Paradox, explaining “If you automate 90 percent of the job, then everyone does the 10 percent of the job. And the 10 percent kind of expands to be 100 percent of what people do and kind of 10x’s their productivity.” The emerging evidence is forcing influential leaders like Amodei to change their tune.
V. What’s Next?
Investing in skill-building for an AI economy is becoming even more imperative. The question isn’t whether AI will destroy jobs, but if we will equip members of society to gain the skills needed to dominate today and in the future, with AI. The Trump administration’s AI Action Plan and initiatives launched by federal agencies like the Department of Labor is a noble first step. Now we need Congress to increase funding of these programs to strengthen our workforce and economy. American strength and power domestically and abroad hinges on ensuring our workers are set up to succeed, not left behind. We can accomplish that goal by building a truly AI-ready workforce.
.png)
