According to researchers, robots with artificial intelligence cannot match human brains in terms of learning new information.
Tech executives are pitching up a future with computer-generated music, rapid disease diagnoses, and self-driving automobiles.
But for the time being, experts claim that an extra step in learning made by our trusty grey matter makes us smarter.When they examined how humans learn to that of AI models, they discovered that humans first consolidate newly acquired information, abilities, or memories before fine-tuning neural connections to minimize mistakes during recall.
However, “backpropagation“—a method of learning by obstructing errors until the system finds the solution—is how Terminator-style machines acquire knowledge.
It implies that in order for AI to digest data, it must do so hundreds or thousands of times.
However, intelligent people like Albert Einstein are able to rapidly connect concepts.
Dr. Yuhang Song, the author of the Oxford University study, stated: “Learning in the brain is superior in many crucial aspects.”
For instance, backpropagation has severe interference from both new and old information and needs more exposures to learn than the brain.
“When trained on new tasks, performance on previous tasks is largely destroyed.
“Our biological systems have an advantage in that they frequently need to adapt quickly to changing environments.”
A letter warning of humanity’s “risk of extinction” by AI was signed by hundreds of professionals.
