[acchow]

This argument has a name. I've seen it presented before and it is rather good. A gross simplification: a computer could never "see" the truth of Godel's Incompleteness Theorems, but we humans can; therefore, we cannot be computers.

[harshreality]

It's not a good argument at all. It's circular, or a tautology depending on how you define things. You're using different words, but what you start with is the same as what you end up with. You assert that humans are different from computers: computers can't "see" something that humans can. And you use that to "prove" that humans cannot be computers. Of course. If you start by asserting humans are computers plus some magic, then of course humans cannot be mere computers.

You could write a computer program to generate random propositions, and test those logically against known truths. After enough testing, the program could assert that propositions not disproven are true. Voila! You have a computer program that has "arrived" at truths outside of a logical system. Some of those truths will be wrong, and some will be right, much like humans and their "truths" that are not based on logical reasoning.

If you've never encountered a human claim to "see" the truth of something that turns out to be false, well, it happens... a lot.

[akkartik]

It's by Roger Penrose: https://en.wikipedia.org/wiki/The_Emperor's_New_Mind

[GigaMos]

Actually Penrose has the idea/argument from Lucas, which has been written in 1959 and published in 1961 : https://www.jstor.org/stable/3749270

[akkartik]

Interesting! http://www.iep.utm.edu/lp-argue

[acchow]

Ah, right. It's called the Lucas-Penrose argument.

[matt_kantor]

I'm not sure what you mean by "see", but there are computer verified proofs of Gödel's theorems: https://en.wikipedia.org/wiki/Gödel%27s_incompleteness_theor....