Computer science graduates in New York outnumbered entry-level tech postings for the first time since 2010

Flat overall employment with a shrinking intake postpones a hiring cost to a later year. A firm that skips two graduate cohorts still needs mid-level engineers in three or four years, and it will buy them off a competitor's payroll at whatever the market asks. The Center for an Urban Future counted postings; the report does not price that later bill ^[1].

The report does not present the 49% as a measure of model capability. Entry-level tech hiring was already declining when ChatGPT was released in November 2022 ^[15][9]. "Rising interest rates beginning in March 2022 cooled tech investment and pushed companies to prioritize profitability over growth," the report said, and: "As margins tightened, the AI arms race redirected capital away from head count and toward data centers and compute infrastructure" ^[14]. It also names inflation, tariffs and geopolitical instability, plus the high cost of hiring in New York, which together it calls a "perfect storm for early-career tech jobs" ^[17].

The exposure data does not line up with the size of the tech drop either. Across occupations, entry-level postings in the jobs with the highest generative AI exposure fell 29% since 2022, while jobs with minimal exposure grew 21% ^[16]. That is a spread of 50 percentage points ^[1]. New York's entry-level tech postings fell 49% over roughly the same window ^[1], 20 points steeper than the most exposed group ^[2]. The two series do not sit on the same base, but the gap is wider than exposure alone accounts for.

Postings and hires are different measures, and the timing fits a rate cycle that started in March 2022 better than a chatbot released that November ^[14][9]. The report says the same about the rate cycle ^[14]. The graduate side moved regardless of cause, and Jonathan Bowles, the think tank's executive director, described the result in Business Insider's account. "The problem isn't that people were getting laid off left and right," he said. "The problem is that they're not getting into the job market in the first place at the lower rungs." ^[7]

CFU's recommendation is that early-career workers get their first experience somewhere other than the firms that stopped hiring them. "Put them to work on civic tech, working for government, for nonprofits, for small businesses," Bowles said. "There are clear opportunities to get people the experience that would help them in the job market, and that's what we recommend." ^[12] That puts the first year or two of training on public, philanthropic and small-business budgets.

Whether AI hiring closes the gap is a forecast. Demand for AI-related skills in entry-level tech postings has grown by more than 8,000% since 2022, measured from a base that predates ChatGPT ^[8][9]. Bowles said "New York is seeing significant growth in AI companies and real growth in companies across tech that are hiring for people with AI skills," and that "As that accelerates in the years ahead, I think there is potential for it to outweigh the kind of reductions that are happening" ^[11]. As of the report, AI job creation has not offset the decline in entry-level tech jobs ^[10].