AI agents wrong ~70% of time: Carnegie Mellon study

CMU is one of the most right-wing, corpo controlled schools in the US colonies (especially influenced by Palantir), so it is interesting they are admitting the "agent" AI concept is fake.

The writing in this article is ridiculously ignorant and credulous though, you can tell the writer doesn't understand "AI"/ML and is still trying to pretend AI is real. For example, only a complete idiot, specifically the complete idiot Thomas Claburn, would repeat the following propaganda after the numbers they cite.

Neubig said he expects agents will become more capable in time but added that even imperfect agents can be useful, at least in the context of coding agents – a partial code suggestion can be filled out and improved.