First there were prompt engineers. Then multimodal model architects. Now, the most in-demand role inside leading AI companies is something few expected: neuroscientists.
A century ago, early ideas about neural networks—loosely inspired by how neurons fire in the brain—laid the foundation for today’s AI boom. But as large language models have become standard across the tech industry, companies are searching for fresh ways to push the frontier. The race to make models smarter, faster, and more efficient is now pulling talent directly from academic neuroscience labs.
The draw was strong enough to convince Aldo Battista, a career academic at New York University's Center for Neural Science, to leave his research position and join Meta as a research scientist this fall.
“Academia is fun because you can explore weird, innovative ideas that may not necessarily have an impact,” Battista said. “But at the end of the day, you want to have an impact on people.”
As a fellow at NYU, Battista studied how the brain makes subjective decisions — like choosing what to eat for dinner. At Meta, he now works on the neural networks powering social media recommendations. Surprisingly, the work doesn’t feel so different. The pace does. “You can measure if people are liking the algorithm more with this tiny modification you made, or not,” he said. “The feedback is immediate.”
Why Neuroscience Matters for the Next Era of AI
Beyond applications in ranking algorithms and wearables, neuroscience intersects with two of the biggest priorities in AI research today: energy efficiency and interpretability.
-
Energy efficiency:
The human brain operates on about 20 watts — yet performs an astonishing number of computations. Meanwhile, modern AI systems can achieve similar capabilities only by consuming exponentially more power. As companies build larger and more complex models, they’re desperately seeking ways to be more efficient. Understanding how the brain accomplishes so much with so little energy is becoming a valuable competitive advantage. -
Interpretability:
Neuroscience offers decades of methods for studying why biological systems make decisions. These same approaches can help researchers understand why AI models generate certain outputs — a crucial challenge as companies look to build systems that are more transparent, safe, and predictable.
In other words: the next leap in AI may come not from bigger datasets or faster GPUs, but from a deeper understanding of human cognition.
Why Researchers Are Leaving Academia
AI companies aren’t just offering cutting-edge problems — they’re offering budgets academia can’t match. With soaring salaries and nearly limitless resources, the private sector has become a powerful magnet for neuroscientists, especially as federal science funding shrinks.
“There’s both a push and a pull,” said Ray Perrault of SRI International. Many researchers want the stability and funding of industry work; others feel forced out as grant opportunities disappear.
In June, neuroscience outlet The Transmitter reported that approximately $323 million in NIH neuroscience grants had been cut — “a significant chunk” of the agency’s budget, according to a former director of the National Institute of Mental Health.
As funding dries up, tech companies are swooping in.
Apple, Google, Neuralink, and now Meta and OpenAI have been aggressively hiring neuroscientists — and the competition is only intensifying. Traditional computer science pipelines aren’t producing enough AI developers to keep pace with demand. As a result, companies are expanding their search.
“The trend of people leaving academia has always existed, just to a lesser extent,” said Matthew Law, now working on post-training at OpenAI after a stint at Stanford’s Institute for Human-Centered AI. “Now, AI companies are shifting away from recruiting traditional CS majors and toward the broader research base. There’s both supply, and the pool of AI devs is also getting exhausted.”
As AI companies battle for an edge in speed, capability, efficiency, and safety, they’re increasingly looking to the one model no machine has yet matched: the human brain.
And for neuroscientists long accustomed to fighting for grant dollars, that shift has opened a new frontier — one where their expertise may shape the next generation of artificial intelligence.