Artificial intelligence is reshaping how software is built, but it is not diminishing the value of science, technology, engineering, and mathematics (STEM) education. Instead, the rapid rise of AI is increasing the importance of strong technical foundations, according to Google DeepMind CEO Demis Hassabis.
The CEO believes that future software engineers will need to combine traditional computer science knowledge with the ability to effectively harness increasingly capable AI systems.
This follows a growing debate over whether advances in generative AI and “vibe coding” will reduce demand for programmers. While AI can already generate large amounts of software code and automate many programming tasks, leading AI executives have been emphasizing that human expertise in software architecture, algorithms, systems design and engineering principles will become even more valuable as developers move into higher-level roles.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Speaking at a London business conference in an interview published on Wednesday, Hassabis said students should continue pursuing STEM subjects, particularly computer science, because AI is fundamentally changing programming rather than eliminating it.
“You absolutely needed to lean into STEM and computer science,” Hassabis said. “It’s just a higher-level programming language is the way you can think about what programming is going to become.”
Hassabis, who co-founded DeepMind in 2010 before Google acquired the AI research company in 2014, said programming has consistently evolved toward greater abstraction. Developers once wrote machine code before transitioning to languages such as C and later Python. AI, he argued, represents the next stage of that evolution, where natural language such as English increasingly becomes the interface for instructing computers.
Even so, he stressed that developers will still need to understand the underlying principles that determine whether software is reliable, scalable, and secure.
“You’re still going to need to know about architecting things and best software engineering practices,” he said. “Those people who understand the deep technical, they’ll be able to use these tools 10 times more effectively than people who don’t have that technical knowledge.”
His remarks suggest that AI will shift the role of software engineers away from writing every line of code manually toward designing systems, validating AI-generated outputs, optimizing performance and managing increasingly autonomous AI agents. In that environment, engineers with a strong grasp of computer science fundamentals are likely to enjoy a significant productivity advantage over users who rely solely on AI-generated code.
AI coding assistants are increasingly becoming force multipliers that automate repetitive programming while leaving humans responsible for higher-level decisions such as system architecture, security, debugging, optimization, and product design.
Beyond engineering, Hassabis argued that AI’s growing influence will also increase demand for expertise outside traditional technical disciplines.
“I also believe that the time is now for the humanities like philosophy, economics. I think we really need them in the world we’re about to enter,” he said.
As AI systems become embedded across healthcare, finance, education, scientific research and government, ethical questions surrounding safety, governance, labor markets, privacy and economic distribution are becoming more important. Hassabis suggested that addressing those issues will require closer collaboration between computer scientists and experts in fields such as philosophy, economics, and public policy.
His comments align with a growing consensus among prominent AI leaders that while generative AI will automate many coding tasks, deep technical education remains essential.
Geoffrey Hinton, widely regarded as one of the pioneers of modern artificial intelligence, said in a December interview with Business Insider that computer science degrees would remain valuable even as AI becomes capable of writing routine software.
“Obviously, just being a competent mid-level programmer is not going to be a career for much longer, because AI can do that,” Hinton said.
However, he noted that the value of a computer science degree extends well beyond coding itself, making it likely to remain relevant for many years.
Affirm CEO Max Levchin has expressed a similar view, arguing that strong computer science fundamentals enable engineers to distinguish high-quality software from poor-quality AI-generated code.
“There’s a matter of taste and elegance in programming,” Levchin said on a podcast earlier this year. “That’s certainly important to me as a programmer, and without having a solid foundation in computer science, I wouldn’t be able to have that conversation.”
The debate comes as AI coding tools have rapidly advanced over the past year. Models from companies including OpenAI, Google DeepMind, Anthropic and a growing number of Chinese AI firms are increasingly capable of generating production-ready code, debugging software, writing documentation and completing complex programming tasks with limited human intervention.
That progress has fueled concerns that entry-level programming jobs could become more difficult to obtain as companies automate routine software development. At the same time, demand is rising for engineers capable of supervising AI systems, integrating large language models into enterprise software, designing AI-native applications and managing complex computing infrastructure.