Humans tend to suffer from the confirmation bias, which is why I think we need to make a conscious effort to listen to well-informed contrarians. On the AI/DL front:
- Gary Marcus is in the spotlight, giving TV interviews and testifying in the Senate. He's a psychologist, by background. And his long-standing critique of DL is largely related to its reliability.
- Judea Pearl, a Turing award recipient, thinks that causal/counterfactual reasoning is needed, and is largely missing from the current approaches.
- Stuart J. Russell co-authored a popular textbook on AI. He thinks DL is limited by its representational power.
- Erik J. Larson authored "The Myth of AI". He's a computer scientist, who worked on AI applications. He argues that the current approaches work inductively and deductively, while humans think mostly abductively.
I'm wondering who else I might be overlooking?
[link] [comments]