Donald Knuth: Ant Colonies and Human Cognition | AI Podcast Clips
00:00:00.000 | - You did mention that you thought that understanding
00:00:05.000 | of the way ant colonies are able to perform
00:00:08.240 | incredibly organized tasks might well be the key
00:00:11.160 | to understanding human cognition.
00:00:13.520 | So these fundamentally distributed systems.
00:00:17.080 | So what do you think is the difference
00:00:18.720 | between the way Don Knuth would sort a list
00:00:22.920 | and an ant colony would sort a list
00:00:24.840 | or perform an algorithm?
00:00:27.880 | - Sorting a list isn't the same as cognition though,
00:00:30.920 | but I know what you're getting at.
00:00:33.380 | Well, the advantage of ant colony,
00:00:37.080 | at least we can see what they're doing.
00:00:39.600 | We know which ant has talked to which other ant
00:00:42.640 | and it's much harder with the brains
00:00:47.640 | to know to what extent neurons are passing signal.
00:00:52.680 | So I'm just saying that ant colony might be,
00:00:56.760 | if they have the secret of cognition,
00:00:59.040 | think of an ant colony as a cognitive single being
00:01:04.920 | rather than as a colony of lots of different ants.
00:01:07.400 | I mean, just like the cells of our brain
00:01:10.320 | and the microbiome and all that is interacting entities,
00:01:15.320 | but somehow I consider myself to be a single person.
00:01:23.520 | Well, an ant colony, you can say might be cognitive
00:01:27.640 | somehow and--
00:01:30.560 | - It sounds strange.
00:01:32.040 | - Yeah, I mean, okay, I smash a certain ant
00:01:37.040 | and the organism's saying, "Hmm, that stung.
00:01:40.120 | "What was that?"
00:01:41.920 | But if we're going to crack the secret of cognition,
00:01:45.680 | it might be that we could do so by psyching out
00:01:51.160 | how ants do it because we have a better chance to measure
00:01:54.720 | the communicating by pheromones
00:01:56.320 | and by touching each other and sight,
00:01:58.840 | but not by much more subtle phenomenon
00:02:02.680 | like electric currents going through.
00:02:05.320 | - But even a simpler version of that,
00:02:06.880 | what are your thoughts of maybe Conway's Game of Life?
00:02:10.400 | - Okay, so Conway's Game of Life is able to simulate
00:02:15.040 | any computable process and any deterministic process
00:02:20.880 | is--
00:02:21.720 | - I like how you went there.
00:02:23.080 | I mean, that's not its most powerful thing, I would say.
00:02:28.080 | I mean, it can simulate it, but the magic is that
00:02:33.480 | the individual units are distributed and extremely simple.
00:02:38.480 | - Yes, we understand exactly what the primitives are.
00:02:41.960 | - The primitives, just like with the ant colony,
00:02:43.800 | even simpler though.
00:02:44.640 | - But still, it doesn't say that I understand,
00:02:48.620 | I understand life.
00:02:51.560 | I mean, I understand.
00:02:52.720 | It gives me a better insight into what does it mean
00:02:59.920 | to have a deterministic universe?
00:03:03.000 | What does it mean to have free choice, for example?
00:03:10.000 | - Do you think God plays dice?
00:03:12.960 | - Yes, I don't see any reason why God should be forbidden
00:03:16.640 | from using the most efficient ways to,
00:03:19.640 | I mean, we know that dice are extremely important
00:03:27.980 | in efficient algorithms.
00:03:29.800 | There are things that couldn't be done well
00:03:32.100 | without randomness, and so I don't see any reason
00:03:34.340 | why God should be prohibited from--
00:03:38.180 | - When the algorithm requires it,
00:03:40.540 | you don't see why the physics should constrain it.
00:03:46.000 | - Yeah.
00:03:46.840 | (audience applauding)
00:03:50.000 | (audience cheering)
00:03:53.000 | (audience cheering)
00:03:56.000 | (audience cheering)
00:03:59.000 | (audience cheering)
00:04:02.000 | [BLANK_AUDIO]