Comprehensive coverage

Scientists discover amazing similarities between baby brains and artificial intelligence

Recent research shows that babies use their time of helplessness to develop cognitive models, similar to the early training of artificial intelligence (AI), challenging old theories about the immaturity of the infant brain and may even lead to advances in AI technology

Recent research shows that babies use their time of helplessness to develop cognitive models, similar to the early training of artificial intelligence (AI), challenging old theories about the immaturity of the infant brain and may even lead to advances in AI technology.

The modern data on the brain do not support the classical explanation for the helplessness of babies. New research suggests that baby brains are not as immature as previously thought; Instead, they use the postnatal period of helplessness to develop basic models similar to those that drive generative AI.

The study, led by a neuroscientist from Trinity College Dublin and just published in the journal Trends in Cognitive Sciences, finds for the first time that the classic explanation for infant helplessness is not supported by modern brain data.

Compared to many animals, humans are helpless for a long time after birth. Many animals, such as horses and chickens, can walk the day they are born. This prolonged period of helplessness puts human infants at risk and places a heavy burden on parents, but surprisingly, it has survived the evolutionary pressure.

Insights from cross-gender research

"Since the 1960s, scientists have thought that the helplessness exhibited by human infants is due to the limitations of birth. It was assumed that in order for babies with big heads to be born, they had to be born early, causing their brains to be immature and a period of helplessness lasting up to one year. We wanted to understand why human babies are helpless for such a long period of time," explains Prof. Rhodri Cusack, professor of cognitive neuroscience and lead author of the article.

The research team included Prof. Cusack, who measures the development of the brain and consciousness of the babies through neurological imaging; Prof. Christine Cherbet from Auburn University, USA, who compares brain development between species; and Dr. Marc'Aurelio Ranzato, senior AI researcher at DeepMind.

"Our study compared brain development between animal species. It drew from a long-standing project, Translating Time, which links age-matched species across species to determine that the brains of humans are more mature than those of many other species at birth," says Prof. Cherbet.

The researchers used brain imaging and found that many systems in the human infant's brain are already functioning and processing the rich streams of information from the senses. This contradicts the long-held belief that many brain systems in infants are functionally immature.

Comparison to learning in AI models

The team compared human learning to state-of-the-art machine learning models, in which deep neural networks benefit from a powerless period of prior training.

In the past, AI models were trained directly on tasks for which they were required, for example an autonomous car was trained to recognize what it sees on the road. But today, models are first used to identify patterns in vast amounts of data, without performing any task of importance. The resulting basic model is then used to learn specific tasks. This has been found to lead to faster learning of new tasks and better performance.

Implications for future development of AI

"We propose that human infants similarly use their period of helplessness for early training, and the learning of strong foundational models, which go on to support later-life cognition with high performance and rapid generalization. This is very similar to the powerful learning models that have led to major breakthroughs in generative AI in recent years, such as OpenAI's ChatGPT or Google's Gemini," explains Prof. Cusack.

The researchers say that future research into how babies learn may well inspire the next generation of AI models.

“Although there have been major breakthroughs in AI, basic models consume enormous amounts of energy and require much more data than babies. Understanding how babies learn could inspire the next generation of AI models. The next steps in the research will be to directly compare the learning in brains and AI", he concluded.

for the scientific article

The study was supported by the European Research Council.

More of the topic in Hayadan:

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismat to prevent spam messages. Click here to learn how your response data is processed.