| Do you remember your third birthday party? How about learning to walk? Tristan Yates has heard from people claiming to recall being born. She's skeptical, but intrigued. Yates is a postdoctoral researcher at Columbia University in New York devoted to exploring the frontiers of people's earliest memories. Last week, she published a paper in Science showing babies can, in fact, record experiences in the region of the brain known as the hippocampus. The life of a memory has several stages: encoding data in the brain, storage and retrieval, for instance. Typically, humans can't recall things that happened in their first few years of life, a phenomenon known as "infantile amnesia." Some have speculated that this is because a baby's hippocampus is simply too immature to encode events. Yates' study demonstrated that, at least by a baby's first birthday, their brains certainly can pull that off. To collect the data, Yates and her fellow researchers had to think like a baby. Small kids have short attention spans. Try telling a 10-month-old to sit in a functional MRI machine for an hour and complete various tasks. So the team created an IMAX-like experience for the children (aged 4 to 24 months) using mirrors and screens. Frequent breaks occurred for milk and diaper changes. Between bouts of "Sesame Street" and "Baby Einstein," they snuck in their research, showing images of faces, scenery and objects against a psychedelic background. Later, the babies saw those pictures again — alongside a new, similar image. Generally, the babies who showed a preference for looking at the familiar image over the new one had experienced higher levels of activity in their hippocampus the first time they had seen the picture. "We interpret this as the hippocampus forming the memory during that initial viewing," Yates says. Such research has broader implications. Babies are outstanding learners, requiring far less data than current AI models to, say, master languages. Gaining a better understanding of how exactly their brains are processing inputs could help adults develop more human-like AI tools, Yates says. Beyond that, the research can inform our understanding of how early life experiences — good or bad — shape a person's development for the rest of their life. Then there's the question of what exactly happens to all those early memories. Do they simply fade away? Or is it possible that, deep down, we really do remember being an infant? In rodent studies, researchers have shown the latter may be the case. Typically, mice can't remember how to find an escape hole in a maze that they had discovered as babies. That is, until scientists reactivate hippocampal neurons that had originally fired when the baby mouse escaped the maze earlier in life. While you can't exactly do the same thing with humans, Yates is nonetheless inspired to come up with new ways of diving into the earliest impressions stamped in people's brains, like the first time someone tasted chocolate ice cream. "To some extent, the memories might be there," Yates says. "They may just be hidden away." –Tim Loh | 
No comments:
Post a Comment