A study from Anthropic found that AI systems can already assist in 75% of the tasks previously performed by software programmers, and the entry-level hiring market is starting to reflect that shift in uncomfortable ways.
When Shailesh Mishra, director of program management at Adobe, turned to LinkedIn to find a software engineering job for his son after his campus offer was rescinded, the post was viewed 350,000 times and brought to light a conversation that had been quietly brewing for months in the tech industry.
He wrote on LinkedIn that a leading tech company has withdrawn the campus placement offer given to his son, a computer engineering graduate from Thapar College, as part of a wider cancellation, which will affect many placements. He asked his network for referrals.
The post gained popularity when an X user shared a screenshot with the caption: “If the son of an Adobe director is facing this, the job market is really cruel.” This point resonated with thousands of respondents that framing connections and credentials no longer guarantee results.
However, tech companies are reducing campus hiring commitments made during the post-pandemic hiring surge. Withdrawals of offers, once rare, are now becoming more frequent with newcomers taking accepted offers with some skepticism.
Commenters on the viral post described the pattern clearly: “The market is so ripe that even people with connections are posting ‘ready to work’ posts.”
The Anthropic study doesn’t conclude that programmers are being replaced, but its conclusion that AI could assist in three-quarters of their daily tasks explains why companies are hiring fewer junior engineers.
Stephen Mai, former Meta and Amazon engineer and co-founder of interview coaching platform Hello Interview, who told CNN that AI has impacted engineering recruiting “like a nuclear bomb” Why are traditional coding interviews increasingly struggling to measure real-world preparation?
Executives from OpenAI, Google and Anthropic have described a shift where engineers spend less time writing code and more time directing AI systems and making architectural decisions.
