Learning to code still important in AI era: Google’s head of research Yossi Matias

Matias’ remarks may come as a surprise at a time when most companies are looking to leverage generative AI tools to write code. The shift in the industry has sparked concerns of job loss among software engineers and reportedly discouraged college students pursuing traditional computer science degrees.

Global research firm Gartner, in a recently published note, said that over 80 per cent of software engineers need to pick up new skills such as natural-language prompt engineering and retrieval-augmented generation (RAG) in order to secure their jobs amid the generative AI boom.

Google’s Sundar Pichai, in the tech giant’s third quarter 2024 earnings call, said that “more than a quarter of all new code at Google is generated by AI, then reviewed and accepted by engineers.”

However, AI coding tools still have a few chinks that need to be worked out. Studies have found that there has been a decline in the quality of code generated using AI models.

A survey of 800 software developers using GitHub Copilot found that none of them saw any productivity gains. The respondents also reported experiencing a 41 per cent increase in bugs within pull requests while using the AI coding assistant to write code.

Hence, Matias may not be entirely wrong in urging people to “master the basic things” such as the fundamentals of coding.

Commenting on how AI is shaping the nature of work, the Google vice-president reportedly told the news outlet that AI is designed to help out with junior-level tasks “which may or may not have some impact on things.”

He also spoke about how AI is poised to revolutionise every field including biology, chemistry, and medicine.

“We got some testimonials from people that say, this might’ve saved my life because I got some information that got me to see a doctor,” Matias said on AI image classification tools designed by Google for healthcare purposes.