If you're an engineer who thinks AI is just hype you can ignore, Andrew Ng has some bad news for you. The Stanford professor and AI pioneer is drawing a pretty stark line between engineers who've embraced AI tools and those who haven't, and the gap is widening fast.
The New Hierarchy of Engineering Talent
Speaking on the "20VC" podcast Monday, Ng laid out what essentially amounts to a new caste system in software engineering, sorted by AI adoption. At the very top? Seasoned engineers with 10 to 20 years of experience who are actively leveraging AI in their work.
"The most productive engineers I know, they're not fresh college grads. They are people of 10, 20 years of experience or whatever, and really on top of AI," Ng explained.
These folks have the domain knowledge, the battle scars, and now the tools to multiply their output. They understand the problems deeply enough to know when AI is helping and when it's hallucinating nonsense.
Fresh Talent With AI Chops Can't Be Hired Fast Enough
Just below the experienced AI-savvy engineers are new graduates who learned AI through online communities and modern coursework. This group, Ng says, is being snatched up as quickly as companies can find them.
"We can't find enough of them," he said, pointing to massive industry demand for people who understand both software fundamentals and AI capabilities.
But then things get uncomfortable. Ng singled out a category he sees struggling: experienced developers who simply haven't updated their skills. These are people with years in the industry who decided to sit out the AI revolution.
"I just don't hire people like that anymore. Those people may get into trouble at some point," Ng said bluntly.
And at the bottom of his hierarchy? New computer science graduates who somehow made it through their degree without AI training. Ng compared this to graduating without knowing about cloud computing, which is to say it's nearly unthinkable in 2025.
"Imagine graduating a CS undergrad that has never heard of cloud computing," he remarked.
The Workforce Is Changing Fast, and Not Everyone's Happy About It
Airbnb Inc. (ABNB) CEO Brian Chesky raised a different concern: if companies lean too heavily on AI and skip hiring junior employees, where does the next generation of leaders come from? Without entry-level roles, there's no mentorship pipeline and no place for people to build foundational skills.
That's not a theoretical worry. Companies like Amazon.com Inc. (AMZN), Meta Platforms Inc. (META), and Salesforce Inc. (CRM) have already slashed entry-level positions. A Randstad study found that postings for junior roles dropped 29% since January 2024.
Meanwhile, former Tesla Inc. (TSLA) AI director Andrej Karpathy poured cold water on the idea that artificial general intelligence is right around the corner. He said AGI is still years away, noting that current AI models lack reliability, true understanding, and advanced reasoning capabilities.
Anthropic CEO Dario Amodei made waves earlier this year by predicting that AI could handle nearly all coding tasks within a year, though he acknowledged that human guidance would remain essential. OpenAI CEO Sam Altman has been encouraging everyone to master AI tools, and Y Combinator reported heavy AI adoption among its startup cohort.
The productivity gains are real. JPMorgan Chase & Co. (JPM) reported a 20% productivity boost for engineers using AI coding assistants, which is a massive jump in an industry where efficiency improvements are usually measured in single-digit percentages.
So where does this leave everyone? If you're an engineer with experience and you're already using AI tools, congratulations—you're in the catbird seat. If you're a new grad who took AI courses seriously, companies are fighting over you. But if you're anywhere in between and you've been ignoring the AI wave, it might be time to start catching up. Because according to Ng, the industry isn't slowing down to wait.