How AI Inspired by Genomes Could Revolutionize Intelligent Systems

**Scientists at Cold Spring Harbor Laboratory (CSHL) have crafted an innovative AI algorithm inspired by the genome's efficient data compression.** Researchers Anthony Zador and Alexei Koulakov explored why complex behaviors can arise despite our genome's limited capacity to store information. They theorized that these constraints might promote intelligence and swift learning. To test this, the team, including Divyansha Lachi and Sergey Shuvaev, developed a computer algorithm that mimics the genome's data compression. When evaluated against existing AI networks, the novel algorithm achieved comparable results in tasks like image recognition and video games, suggesting it possesses an inherent understanding of these tasks. However, while impressive, this new AI does not yet reach the brain's information capacity; the brain's architecture accommodates vastly more data. Yet, the algorithm’s high compression could be a significant tech advancement, potentially enabling complex models to run efficiently on compact devices such as smartphones. This development marks a crucial step in AI evolution, emulating the efficiency honed over billions of years of natural evolution.