Researchers question Microsoft Copilot and ChatGPT smarts as AI champs do well with "memorization rather than true reasoning abilities

[Windows Central]

A new study by MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) indicates LLMs tend to perform better at tasks they're familiar with than new territory.

Researchers question Microsoft Copilot and ChatGPT smarts as AI champs do well with "memorization rather than true reasoning abilities : Read more

 

[fjtorres5591]

And water is wet.

Too many people who should know better are letting themselves be caught up in AI histeria over what is little more than marketing hype.

LLMs are little more than advanced database query software.
Big database and very sophisticated query sistem but there is no intelligence, artificial or otherwise, involved. It's just a modern version of the Mechanical Turk, applied to Vannebar Bush's memex's concept.

Very useful and valuable ($$$) to developers and users alike but worthless to ivory tower academics interested in fanciful dreams of software minds. No different than the UFO-obsessed or cryptid hunters.

Let it go, folks.
Focus on what the software does, not what it isn't and never will be.
True AI, if ever developed, will not come from this technology.