“David Krakauer, director of the Santa Fe Institute, argues that Large Language Models are not truly intelligent because they operate by "doing more with more," leveraging vast knowledge to find shortcuts rather than reasoning from first principles.”