TL;DR
A Harvard Business School and INSEAD working paper finds that non-AI startups are 25% smaller, employ 13% more engineers, and have about 15% lower shares of entry-level employees and managers than their non-AI peers. Their hiring skews high-end, elite-educated, Silicon Valley-based and male, suggesting AI is concentrating rather than democratizing opportunity.
AI startups hire fewer entry-level employees than their peers working paper From Harvard Business School and INSEAD, first hand information Business Insider. Firms are very serious about leaner, flatter, and top-level technical talent.
Researchers Rembrand Koning and Hyunjin Kim examined Y Combinator startups from 2020 to 2024, along with a broader set of US venture-backed startups. They identify AI-driven startups in two directions: using AI internally to make employees more productive, and incorporating it into products so customers can automate work that once required human teams.
The numbers are stark, with 25% fewer domestic AI-based startups, 13% more engineers employed, and about 15% less of both entry-level employees and managers. The share of senior employees is 20% higher and valuations are comparable to their non-AI peers, implying greater value creation per employee.
The workers these firms employ are skewed in a certain way. “These workers are mostly male and graduates of elite institutions concentrated in Silicon Valley,” the authors write.
It undercuts the hopeful reading of an AI boom in which teenagers use AI to shoot above their level, and vibe coding lowers the technical bar. The paper suggests that opportunity is instead concentrated among those who are already secure.
The authors’ deeper concern is exacerbated inequality, warning that if AI accelerates learning for those who use it, “differential adoption rates could translate into widening performance gaps.” This applies to employees in firms and entrepreneurs who found them.
The bottom step is cracking
The findings reflect what is already being seen in the labor market AI is killing the summer experience and unemployment of graduates is increasing. Only 7% of new hires at large tech companies are recent graduates.
Big Tech is in the business of turning payroll into computing Meta and Microsoft cut 23,000 roles AI spending hit records. Meanwhile, the above demand is so hot that AWS puts $1 billion to forward-deployed AI engineers.
He himself became a recruiter AI-on-AI arms race. Machines for new graduates now sit on both sides of the table.
The result of the study is troubling for anyone who peddles artificial intelligence as a democratizing force. Technology can flatten hierarchies within companies, but it can also strengthen the way up to enter companies.
