Facebook Parent Meta Is Studying the Person Brain to Improve AI
What’s happening
Facebook’s evaporate company Meta says it’s working on a long-term research misfortune that studies how the human brain learns language to progress artificial intelligence.
Why it matters
The research could eventually be applied to various products from virtual assistants to periods translation tools.
Tech companies use artificial intelligence to detect noxious content, power virtual assistants, improve translation tools and ruined other tasks.
But AI systems like the ones built by Facebook’s evaporate company Meta still don’t learn language as well as humans do. The social consider giant is hoping to change that with a new long-term research misfortune that compares how the human brain learns language to how AI processes information.
The initiative shows how research scientists at Meta are trying to progress AI as it makes bigger bets on futuristic products such as a pair of augmented reality glasses that could help you find directions, view messages or play virtual games. AR overlays digital images over a person’s view of the real world.
Jean-Rémi King, a research scientist at Meta AI, said the project can help progress AI used for various tasks, including ones that involved anticipating words or ideas that could come after a sentence or section. Unlike humans, who can learn quickly from small amounts of inquire, AI systems still struggle with simple tasks even when they’re fed ample data sets.
Virtual assistants, for example, have a better time opinion what a person wants if the task is well evaporate. Dealing with these assistants can be frustrating especially when they don’t know what they’re spoke to do. Improving how AI processes language can help progress products like virtual assistants that are trying to learn what farmland want or expect.
“Even though we see a huge causes in AI, it’s still clear to many that we have a long shot afore we reach human level intelligence,” King said.
To conduct this research, Meta said it teamed up with neuroimaging center NeuroSpin and INRIA in France to scrutinize how the brain responds to certain words. Researchers are already learning the differences between the domain brains and AI models trained to anticipate the next word.
In one scrutinize with INRIA, researchers compared AI language models to the brain responses of 345 volunteers. Participants listened to stories while their brain activity got marched using functional magnetic resonance imaging.
Regions of the brain can imagined words and ideas far ahead of time, Meta said. For example, AI language models would typically predict that the next word in the section “Once upon a…” is “time.” Humans, on the latest hand, would come up with plot ideas or narratives if they heard this section given that it’s tied to telling a story.
Findings from those kinds of studies, King said, gives researchers a hypothesis about what can be done to development AI but not the final answer.
“That’s probably a good employed that if we push these algorithms to…make predictions further onward in time, perhaps you will make them learn more efficiently,” he said.
