Lex Fridman: Recipe for Progress in AI (Hard Work)
00:00:00.000 | I mentioned soup in terms of progress.
00:00:03.600 | There's been a little bit of tension,
00:00:05.920 | a little bit of love online in terms of deep learning.
00:00:08.640 | So I just wanted to say that the kind of criticism
00:00:13.080 | and skepticism about the limitations of deep learning
00:00:16.280 | are really healthy in moderation.
00:00:18.740 | Jeff Hinton, one of the three people
00:00:22.140 | to receive the Turing Award, as many people know,
00:00:25.440 | has said that the future depends on some graduate student
00:00:28.280 | who is deeply suspicious of everything I have said.
00:00:31.160 | So that suspicion, skepticism is essential,
00:00:35.200 | but in moderation, just a little bit.
00:00:37.520 | The more important thing is perseverance,
00:00:41.040 | which is what Jeffrey Hinton and the others have had
00:00:44.000 | through the winters of believing in neural nets
00:00:47.080 | and an open-mindedness for returning to the world
00:00:50.320 | of symbolic AI, of expert systems,
00:00:53.560 | of complexity and cellular automata,
00:00:56.280 | of old ideas in AI and bringing them back
00:00:58.840 | and see if there's ideas there.
00:01:00.160 | And of course, you have to have a little bit of crazy.
00:01:03.080 | Nobody ever achieves something brilliant
00:01:04.760 | without being a little bit of crazy.
00:01:06.520 | And the most important thing is a lot of hard work.
00:01:09.920 | It's not the cool thing these days,
00:01:13.640 | but hard work is everything.
00:01:15.320 | I like what JFK said about us going to the moon, us.
00:01:20.320 | I was born in the Soviet Union.
00:01:23.320 | See how I conveniently just said us?
00:01:26.020 | (audience laughing)
00:01:27.980 | Going to the moon is, we do these things
00:01:31.460 | not because they're easy, but because they're hard.
00:01:34.100 | And I think that artificial intelligence
00:01:37.200 | is one of the hardest and most exciting problems
00:01:39.740 | that are before us.
00:01:41.500 | So with that, I'd like to thank you
00:01:43.980 | and see if there's any questions.
00:01:46.180 | (audience applauding)
00:01:49.340 | (upbeat music)
00:01:51.920 | (upbeat music)
00:01:54.500 | (upbeat music)
00:01:57.080 | (upbeat music)
00:01:59.660 | (upbeat music)
00:02:02.240 | (upbeat music)
00:02:04.820 | [BLANK_AUDIO]