Meta chief AI officer Alexandr Wang has urged teenagers to immerse themselves in AI-assisted coding, arguing that early familiarity with the technology could give them an advantage in the future job market.
“If you are, like, 13 years old, you should spend all of your time vibe-coding. That's how you should live your life,” Wang said on the TBPN podcast.
Vibe-coding refers to building software by describing tasks in natural language and using artificial intelligence tools to generate and refine code. The approach allows users to create applications and websites without writing every line of code themselves.
Wang compared the current moment to the early years of personal computing, when people who learnt to work with computers gained an advantage as the technology spread.
“It's almost like when personal computers first came about, or just computing in general, the people who spent the most time with them, or who grew up with it, had this immense advantage in the future economy,” he said.
He also pointed to the advantage of spending thousands of hours experimenting with AI tools. “If you just happen to spend like 10,000 hours playing with the tools and figure out how to use them better than other people, that's a huge advantage,” Wang said.
Debate over learning to code in the AI era
Wang’s remarks come amid debate over whether AI coding tools will reduce the need for conventional programming skills or make those skills more useful across professions.
Andrew Ng, founder of DeepLearning.AI and co-founder of Google Brain, has said that AI-generated code is not a reason to stop learning programming.
“As coding becomes easier, more people should code, not fewer,” Ng wrote in a March 2025 post. He has argued that people who understand coding can use AI tools more effectively, including in fields outside software engineering.
Research has also highlighted the limitations of AI-assisted development. A July 2025 study by the research organisation Model Evaluation and Threat Research (METR), which tested 16 experienced open-source developers on 246 tasks, found that participants took 19% longer to complete tasks when allowed to use AI coding tools. The researchers cautioned against generalising the result beyond the study’s setting.
The findings do not establish that AI tools make all coding slower, but they underline the importance of checking and correcting AI-generated output rather than relying on it without review.
Wang co-founded Scale AI in 2016 after leaving the Massachusetts Institute of Technology. He joined Meta as chief AI officer following the company’s $14.3 billion investment in Scale AI in 2025.