Startups use AI coding tools to keep teams smaller

Technology startups are increasingly using AI coding tools to build products with smaller teams, while some executives say the shift is reducing demand for entry-level programmers.

Eric Lauer, who runs the online gift-giving platform Giftory, said he now looks for experienced developers who can use AI tools to increase their output rather than young coders just leaving college. He said workers need previous experience and knowledge of software workflows to act as "architects" who guide AI-generated code.

The tools discussed include Anthropic's Claude Code and OpenAI's Codex. The article says such tools can write, test and fix software from text prompts, changing programmers' roles from writing every line of code to managing projects and checking AI output.

Among developers at small startups surveyed by The Pragmatic Engineer, 75 percent said they used Claude Code. Jared Friedman, managing partner of Y Combinator, said a quarter of startups in Y Combinator's Winter 2025 batch were built on code that was 95 percent generated by AI.

Lauer said Giftory's roughly 30 employees receive a premium AI subscription costing about $200 a month, which he described as small compared with an average annual salary of $100,000 and cheap enough to make offshoring less competitive. Haitham Mengad, co-founder of Stems Labs, said his company was trying to do more with an already lean team. Lindsay Euller, vice president of customer success at Espresa, said her team's AI use was saving millions of dollars a year and could become part of any future decision on approving more staff.

The article also cites research suggesting pressure on junior jobs. A Stanford Digital Economy Lab study of payroll data from millions of US workers found that employment among 22- to 25-year-olds in occupations most exposed to AI, including software development, fell nearly 20 percent from a late 2022 peak. Harvard researchers analyzing resume and job posting data covering about 62 million US workers across 285,000 firms found that junior employment at companies adopting generative AI fell roughly nine percent compared with non-adopters within six quarters, while senior employment kept rising.

Ian Amit, chief executive of cybersecurity startup Gomboc AI, said many companies were interviewing candidates but not making hiring decisions. Amazon Web Services chief executive Matt Garman warned against replacing junior developers with AI, saying the industry could cut off its future leadership pipeline.

Computer science enrollment has begun to fall, according to the article. It says enrollment dropped six percent across the University of California system and declined at two-thirds of computing programs nationwide, according to the Computing Research Association. Lauer said startups still face a choice between adding more resources or more people, and the article says many in the tech sector are choosing more AI and fewer workers.