AI kills junior jobs, but creates a market for experts

The dream of a first position in tech is moving away for generation Z. In 2024, new recruits from the best engineering schools struggle to penetrate a market locked by AI, budgetary constraints and a new logic of productivity. The figures are without appeal the hires of young graduates in tech have dropped from Over 50 % Since 2019. The phenomenon is no longer anecdote but now structures the new rules of recruitment in the sector.

A generation sacrificed on the altar of efficiency

According to Signalfire, an investment fund installed in San Francisco, young graduates no longer represent 7 % of new hires in Big Tech And less than 6 % in startups. The share has dropped by 25 % in one year among large companies and 11 % among startups, but it is above all the comparison with the forefront which alerts with the hiring volume of junior profiles which was divided by two in five years.

The entry -level posts are rare, and when they are published, they are occupied by seniors. It’s the ” Experience paradox », To get your first job, you must have already proven yourself.

AI eliminates tasks, not talents

Contrary to popular belief, it is not AI that eliminates mass jobs, but These are companies that redefine what they expect from recruitment. By automating repetitive or standard tasks (writing, debugging, data analysis), AI allows companies Reduce the share of assignable work to junior profiles. Result, they concentrate their hires on roles with high added value.

Experienced engineers have become the nerve of war. The Big Tech thus increased their hiring of profiles by 27 % between 2 and 5 years of experience. Startups follow the same logic, with an increase of 14 % on this population.

Anthropic, the new ia talent school

In this elite race, some companies are doing well. Anthropic, Rival of Openaidisplays a 80 % retention rate Over two years, far ahead of Openai (67 %) and Meta (64 %). The company literally siphoning the talents of its competitors, with a report of 8 for 1 vis-à-vis Openai and 11 for 1 vis-à-vis Deepmind. Its recipe is based on a culture without titles, without micro-management, and centered on the autonomy of researchers.

The generation resources

Graduates are forced to reinvent their journey. Bootcamps, open-source projects, freelance missions or community contributions become new entry tickets. Those who know how to use AI have no more advantage unlike those who know how to correct AI errors. The debugging of the code generated by machines could become the new rare competence.

A pipeline in danger

The absence of a long -term strategy on the training and integration of juniors asks another question, Who will replace the seniors tomorrow? By refraining from training, companies weaken their own future. The talent pipeline empties. And short-termist logic could cost dearly in the medium term. It remains to be seen if these figures are found in Europe, a project of next subject. If you wish to share with us information about this: redaction@frenchweb.fr