[cs702]

Hinton qualified that statement by adding, "but I do think there's going to have to be quite a few conceptual breakthroughs." He explicitly conditioned his belief on future conceptual breakthroughs that have not been made yet. Also, his statement is not a prediction; it starts with "I do believe."

Here's the full quote, copied verbatim from the article:

> I do believe deep learning is going to be able to do everything, but I do think there’s going to have to be quite a few conceptual breakthroughs. For example, in 2017 Ashish Vaswani et al. introduced transformers, which derive really good vectors representing word meanings. It was a conceptual breakthrough. It’s now used in almost all the very best natural-language processing. We’re going to need a bunch more breakthroughs like that.

Please don't criticize him or the article without first reading it in full.

[bo1024]

Yeah, it's a bit tricky ... like saying "Turing Machines are going to be able to do everything, but we'll need quite a few conceptual breakthroughs." Hard to say what percent of progress we've made since 1936 toward artificial general intelligence (whatever that means).

[cs702]

In principle, yes, of course, because we use Turing machines for deep learning :-) but I think Hinton's point is much less broad than you imply: He believes "AGI" ultimately will be achieved with deep learning approaches, as opposed to other approaches, e.g., symbolic AI.

[bo1024]

Yes, but "quite a few conceptual breakthroughs" leaves a lot of room open for if it would be at all similar to deep learning today.

[cs702]

Agree :-)

Hinton's comment was meant as, and is, an educated guess, expressed as a belief -- much like expressing belief that P≠NP, say, or that string theory will prove to be a dead end in Physics.

FWIW, I happen to agree with Hinton, but only time will tell if we're right!