A new AI company called NeoCognition just raised $40 million to create AI that learns the way humans do. Instead of training AI on massive datasets, they want to build AI agents that can become experts in any field by learning as they go.
This could change how we think about AI completely. Today’s AI systems like ChatGPT need enormous amounts of data and computing power to learn anything. NeoCognition thinks there’s a better way – making AI that can learn new skills quickly, just like a person would.
Learning Without the Data Dump
The company was started by a researcher from Ohio State University who spent years studying how human brains actually learn. Their approach is totally different from big tech companies that throw millions of examples at their AI systems.
Instead, NeoCognition wants to create AI agents that can watch, experiment, and figure things out on their own. Think of it like the difference between cramming for a test versus actually understanding the subject.
The $40 million seed round is huge for such an early-stage company, showing investors believe this approach could work. The money will help them build their first AI agents and test whether machines can really learn like humans.
What’s Next
NeoCognition plans to start with specific industries where quick learning matters most. If they’re right, we could see AI that doesn’t need years of training data to become useful. Instead of building separate AI systems for every task, one agent could learn to be a doctor, then learn to be a lawyer, then master cooking – all by learning the way humans do.