Autodesk's CEO says these skills are more important than coding in the AI era

• Autodesk's CEO said systems-level and interdisciplinary thinking is more important than coding.
• AI tools will increase code generation, reducing the need for traditional computer science roles, he said.
• He said education systems need to adapt to teach critical thinking and AI engagement for creativity.

Many tech executives still preach the importance of learning to code — but the CEO of software company Autodesk says there's another skill that's more important.

"If the coding models are going to be doing the code for you, what's more important is that you understand there's this whole notion of systems-level and interdisciplinary thinking," Andrew Anagnost told Business Insider.

As someone who followed a non-traditional educational path before earning a Ph.D. in aeronautical engineering and computer science, Anagnost is a big advocate for interdisciplinary thought and exploration.

While he understands the value of diving deeply into a topic and gaining expertise in it, he doesn't think going "incredibly deep into a narrow discipline" is what's most important in today's job market, unless the person has plans to be a research scientist.

Anagnost said that in a world where AI agents can perform specialized skills, understanding multiple disciplines and engaging in the "what and the how of how to create a product" will become increasingly important. He added that humans will need to take the role of "creative orchestrators" and manage the outcomes of AI systems.

A new era for computer scientists

Anagnost's comments come as AI tools like Codex and GitHub Copilot have grown in popularity and are increasingly handling coding tasks that were previously core to the work of software engineers.

Anagnost said that with the emergence of AI tools, there will be "more people generating code than ever before," and many of them won't have backgrounds in computer science. The CEO said it will take just "a little bit of effort" to generate code that can perform very specific tasks.

"There's no doubt as we move into the future, more people are going to be generating code in some way that runs computers in new and interesting ways," Anagnost said. "It's just going to be different people."

Education needs may shift

Anagnost explained that the true value of a computer science education won't lie in entry-level coding, and companies will likely reduce hiring for those roles. However, he said that computer scientists will still be needed for more advanced work, like deep modeling and algorithms.

Anagnost said that a software company typically employs four kinds of workers: a product manager, a product designer, an engineer, and a QA specialist who tests the product. The CEO said that in the future, those four roles can likely be reduced to two, with product designers working with a coding agent to test the software.

In those situations, Anagnost said there will need to be people with "total systems thinking" about how everything works together. Anagnost said that education systems will need adapt to that shift and teach students how to think critically and engage with AI tools to expand creativity.

"There'll probably be less people with traditional computer science degrees and software companies," Anagnost said. "But there'll probably be more people creating product than ever before."