Although I agree that absolutes and superlatives are not a good idea when talking about the possibilities of the future, I have very strong doubts that we will one day be able to build a strong AI. At least with traditional digital computers (who knows what we would be able to do with quantum computers though?). I've studied computer science and artificial intelligence myself and although its possibilities are wide and almost unfathomable, I never saw true consciousness as really achievable.
That being said, up until a few weeks ago, I was convinced I wouldn't have lived to see a computer program beat up a human at Go. Yet there again, I saw that as very highly improbable when others described it as impossible : we can't speculate with any accuracy on something we don't fully understand.
One thing though, the author does specify that it will not be possible with digital computers we use now, based on the principle of Turing machines. He does not say it is impossible no matter how we try, just that we can't in the way we build computers now.
Although I agree that absolutes and superlatives are not a good idea when talking about the possibilities of the future, I have very strong doubts that we will one day be able to build a strong AI. At least with traditional digital computers (who knows what we would be able to do with quantum computers though?). I've studied computer science and artificial intelligence myself and although its possibilities are wide and almost unfathomable, I never saw true consciousness as really achievable.
That being said, up until a few weeks ago, I was convinced I wouldn't have lived to see a computer program beat up a human at Go. Yet there again, I saw that as very highly improbable when others described it as impossible : we can't speculate with any accuracy on something we don't fully understand.
[This comment was removed]