Luca Ambrogioni, Radboud University assistant professor, argues current AI still fails Eliezer Yudkowsky's 2016 Turing test prediction
Manheim says AGI requires superhuman capabilities to pass.
——0——
@davidmanheim Most things cannot be sorted on one axis.
LLMs are just their own thing. They are neither super or sub human. They just aren't hunan
@LucaAmb It seems like Yudkowsky was exactly right about part of his prediction: LLMs are effectively superhuman in a wide variety of areas, but we keep being able to tell they are not human in the remaining domains.
12:07 PM · May 24, 2026 · 134 Views
12:56 PM · May 24, 2026 · 125 Views