More brainlike computers could change AI for the better

Researchers are developing neuromorphic computing systems inspired by animal brains for more efficient and adaptable AI. These systems mimic brain features like spiking neurons and combined memory-processing. Companies such as Intel and IBM are scaling up neuromorphic chips, with Intel's Hala Point containing 1.15 billion electronic neurons. Liquid AI and Numenta are creating adaptable networks based on worm brains and human neocortex structure, respectively. These approaches aim to reduce energy use and improve AI capabilities, but their practical effectiveness requires further testing.