Scientists have made a breakthrough in AI research by developing a pocket-sized AI brain, thanks to the help of monkey neurons. This innovative model, which mimics a part of the brain's visual system, has been compressed to a tiny fraction of its original size, all while maintaining its performance. The team, led by Ben Cowley, an assistant professor at Cold Spring Harbor Laboratory, has created an AI model that is so small it could be sent in a tweet or an email. This compact model not only works more like a living brain but also provides insights into how diseases like Alzheimer's might occur. By understanding the inner workings of the human brain, scientists can develop more powerful and human-like artificial intelligence. The study, published in the journal Nature, is part of an effort to understand the human visual system, which takes in bits of light and transforms them into recognizable objects. The team used data from macaque monkeys to train the model and then applied statistical techniques to compress it. This allowed them to glimpse the specialized nature of the artificial neurons, which respond to specific shapes and patterns. The findings have implications for artificial intelligence, suggesting that AI systems could be smaller and simpler yet still perform better than current models. For example, self-driving cars might be able to run on less powerful computers while correctly distinguishing a pedestrian from an airborne plastic bag. However, AI systems still struggle with tasks that require recognizing a friend's face in any setting and from many angles. This may be because current AI models are based on an understanding of the human brain from the 20th century, and scientists may need to update the foundations of artificial networks to improve their performance.
