New study: Computer profs nix claim that chatbots think as we do
In a world where tech pundits are claiming that AI will soon think like people, the University of Amsterdam and the Santa Fe Institute offer some thoughts at ScienceDaily about why it won’t. Their focus is on chatbots like GPT-4:
Analogical reasoning is the ability to draw a comparison between two different things based on their similarities in certain aspects. It is one of the most common methods by which human beings try to understand the world and make decisions. An example of analogical reasoning: cup is to coffee as soup is to (the answer being: bowl)
“Why GPT can’t think like us,” February 21, 2025
A study by language and AI experts Martha Lewis (Institute for Logic, Language and Computation at the University of Amsterdam) and Melanie Mitchell (Santa Fe Institute) tested whether GPT models are as well able as capable of making analogies as humans:
Humans maintained high performance on most modified versions of the problems, but GPT models, while performing well on standard analogy problems, struggled with variations. ‘This suggests that AI models often reason less flexibly than humans and their reasoning is less about true abstract understanding and more about pattern matching’, explains Lewis.
In digit matrices, GPT models showed a significant drop in performance when the position of the missing number changed. Humans had no difficulty with this. In story analogies, GPT-4 tended to select the first given answer as correct more often, whereas humans were not influenced by answer order. Additionally, GPT-4 struggled more than humans when key elements of a story were reworded, suggesting a reliance on surface-level similarities rather than deeper causal reasoning.
On simpler analogy tasks, GPT models showed a decline in performance decline when tested on modified versions, while humans remained consistent. However, for more complex analogical reasoning tasks, both humans and AI struggled. “Why GPT can’t think like us,”
We are warned: “AI can be a powerful tool, but it is not yet a replacement for human thinking and reasoning.” The paper is open access.
The critical question is not whether the chatbots are “yet a replacement” for human thinking but whether they ever can be. The industry may not be ready for an honest discussion of that yet.
