What Andrej Karpathy’s move reveals about where AI is heading

POV by Luke Johnson, CEO

If you’re interested in tech and AI news in any way, you can’t have missed the news that Andrej Karpathy has joined Anthropic.

If the name doesn’t ring a bell, at the very least you’ll have heard of “vibe coding.” Karpathy coined that term.

But that’s not all.

Andrej Karpathy has done some amazing things. He cofounded OpenAI and spent two years there. He then left and went to Tesla for five years. He did a ton of things on his own. He went back to OpenAI. Now he’s joining Anthropic.

Karpathy has been described as the most respected AI researcher and the most respected AI educator alive.

What’s interesting to me is that he’s not returning to explain AI. He’s returning to build it.

His position at Anthropic will be working on the AI startup’s pretraining team, which is responsible for large-scale testing of the Claude LLM family.

Karpathy explained his thinking himself in a post on X:

“I’ve joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.”

Anthropic is clearly the place where he wants to be right now.

My guess is that he’s very passionate about solving a specific problem and feels Anthropic is best suited to solve that problem at present. I would also guess he’s excited about the deep focus on R&D and the specific role and opportunity in front of him.

For me, that makes perfect sense.

Why Does More AI Research Make Sense Now?

One reason this news caught my attention is that it connects to something I’ve been thinking about for a while: how long it often takes new technologies to move from exciting innovation to meaningful economic impact.

When I look at AI today, I think we’re still early in that process. Costs remain high and broader benefits are still emerging.

For that to change, we need more research.

That’s why I pay attention when talented researchers choose to return to the frontier of AI development. Karpathy is, at heart, both a researcher and an educator. Both will be necessary if AI is going to play a truly useful role in the world.

Looking back through history, though, many of the technologies that eventually transformed society took much longer to deliver widespread economic impact than people expected at the time.

The steam engine ultimately changed the world, but it took decades before its effects were fully felt across the economy.

Electricity followed a similar path. Commercial power stations existed long before electricity became the dominant source of industrial power.

Personal computers are another example. The first microprocessor appeared years before PCs became commonplace in homes and businesses.

These technologies all seem inevitable. Living through them was probably a very different experience. The same may prove true for AI.

AI May Need Years to Impact the Economy

There is often a long period between invention and widespread productivity gains. The major AI players are clearly reassessing where the technology is going and how long it may take to mature.

The economic impact may ultimately be enormous. But in the meantime, its maturity will depend on talented AI specialists pushing the boundaries of what’s possible – which is why Karpathy’s move caught my attention.

The major tech groups are all leapfrogging each other month by month or week by week. My take on this is that the talent will likely do the same.

What we’re seeing – and will continue to see – is that people with real talent move around. They move toward interesting problems, strong teams, and opportunities to do meaningful work.

With this move, I would not read too much into what it says about any of the individual companies.

I’ve watched all the key players over the last few years, and I’m not sure this move alone means Anthropic is destined to become the dominant AI company.

Andrej has a history of moving around, and I wonder if Anthropic can keep him challenged and entertained for a material amount of time.

We’ve never seen technology grow and change at this pace before, and it’s amazing to sit back and be a part of it all.

But the bigger story for me is not which company is ahead this month.

I see AI redefining what is possible for the masses. I don’t always see it as a threat. I see it as an enabler for so many people who can put their creativity and innovation to work in ways that were never possible before.

That’s why I pay attention when people like Andrej Karpathy make moves like this. Not because of what it says about a particular company, but because of what it might tell us about where AI is heading next.

Why Katalyst Pays Attention to AI Research

At Katalyst, we look beyond product announcements and industry headlines to understand the developments shaping the future of technology. The breakthroughs being explored today may become the business tools, productivity gains, and growth opportunities for mid-market businesses tomorrow.

If you’d like to discuss any of these issues, schedule a call and ask for me.