AI has displaced juniors and devalued Stanford diplomas - ForkLog: cryptocurrencies, AI, singularity, future

# AI has displaced juniors and devalued Stanford diplomas

Artificial intelligence has devalued the diplomas of programmers from Stanford: the “gold standard” status has shifted to “bronze,” reports the Los Angeles Times.

Students at one of the top universities in the US are shocked by the lack of job offers. Journalists note the rapid pace of change: when current graduates enrolled, ChatGPT did not yet exist, and today neural networks write code better than most people.

“Graduates of the Stanford Computer Science Department are struggling to find entry-level jobs at the most well-known tech companies,” — noted Associate Professor of Bioengineering at Stanford, Yan Liphardt.

Rapidly developing generative artificial intelligence capabilities in programming have made experienced developers more productive, but they have also worsened employment prospects for beginners.

“There is a gloomy atmosphere on campus. People looking for jobs are very tense,” — said a recent graduate of the Computer Science Department.

The crisis has affected educational institutions across California—from Berkeley to the University of Southern California. Holders of less prestigious diplomas find themselves in an even more vulnerable position.

Aylul Akgul shared that last year she received a degree in computer science from Loyola Marymount University. The girl did not receive a single offer, so she returned home to Turkey to gain experience at a startup. Even after returning to the US with real-world experience, it took her four months to find a job.

“The job market for programmers is becoming very saturated,” — she noted.

Technological leap and the labor market

After launching ChatGPT in 2022, it could code in 30 seconds per session. Modern AI agents can code for hours, quickly perform basic tasks with fewer errors.

Startups like OpenAI and Anthropic continue to expand their staff, but this does not offset hiring reductions in other sectors. According to Stanford, employment among beginner developers aged 22 to 25 has decreased by about 20% from the peak figures at the end of 2022.

AI competition is felt beyond IT. Customer support and accounting staff are hired 13% less compared to professions where automation is more difficult, for example, (caring for the sick).

In Los Angeles, about 200,000 jobs are at risk. 40% of tasks performed by call center employees, editors, and personal finance experts can already be delegated to algorithms.

Pride in automation

Many tech startups openly declare layoffs in the context of AI implementation.

Dario Amodei, CEO of Anthropic, said that Claude wrote 70-90% of the code for some of the company’s products. He predicts that in five years, almost half of entry-level “white-collar” jobs will disappear.

Nenad Medvidovic, a computer science professor at the University of Southern California, stated that if previously hiring managers needed ten programmers, today they need two highly qualified engineers and one AI agent.

“We no longer need juniors. Now artificial intelligence programs better than graduates of prestigious universities,” — said Amir Avadalla, CEO of the AI startup Vectara.

Programmers are still needed

Artificial intelligence is not yet ready to fully replace developers. It performs well on structured, repetitive tasks, while human work has shifted toward oversight.

Modern models are powerful but unstable. They can solve complex mathematical equations but also make errors in basic logic.

According to one study, neural networks slowed down experienced specialists by 19%, as they spend more time checking code and fixing bugs.

John David N. Dionisio, a computer science professor at LMU, recommends students shift their focus in training. A key skill becomes the ability to manage artificial intelligence and thoroughly verify its work.

Recall that in August, Coinbase laid off programmers who refused to use artificial intelligence in their work.