WEBVTT - Can We Upload Our Brains to a Computer? 0:00:00.200 --> 0:00:02.400 Hey, please take a second and leave us a review 0:00:02.480 --> 0:00:06.279 on Apple Podcasts, Spotify, or wherever you listen to the podcast. 0:00:06.960 --> 0:00:11.039 Thanks a lot. Hey, welcome to sign Stuff and production 0:00:11.200 --> 0:00:14.440 of iHeartRadio. I'm Hoar hitch hamp and today we're asking 0:00:14.440 --> 0:00:18.079 the question can you upload your brain to a computer? 0:00:18.880 --> 0:00:22.200 Is it actually possible to digitize your mind and who 0:00:22.280 --> 0:00:25.200 you are and have it run on a robot or 0:00:25.239 --> 0:00:27.159 a hard drive. Would it be a way for you 0:00:27.240 --> 0:00:30.960 to lift forever or maybe a way to copy yourself. 0:00:31.320 --> 0:00:34.600 We'll talk to three experts about this, a mathematician who 0:00:34.760 --> 0:00:39.800 focuses on emerging technologies, a neuroscientist who studies brain connections, 0:00:40.080 --> 0:00:43.640 and a philosopher who's grappled with the question should we 0:00:44.080 --> 0:00:47.519 upload our brains? So don't pull the plug and get 0:00:47.560 --> 0:00:51.360 ready to live inside our electronic mind as we answer 0:00:51.440 --> 0:00:54.959 the question can you upload your brain to a computer? 0:01:04.680 --> 0:01:04.880 Hey? 0:01:04.880 --> 0:01:09.320 Everyone, So today we're talking to three people, a mathematician, 0:01:09.760 --> 0:01:13.720 a neuroscientist, and a philosopher. And they're not going to 0:01:13.720 --> 0:01:16.240 walk into a bar, but if each spend a lot 0:01:16.240 --> 0:01:19.880 of time thinking about this idea of uploading your brain. 0:01:20.400 --> 0:01:24.600 We'll start with the mathematician, Professor Ole Hackstrom. Here's what 0:01:24.640 --> 0:01:29.560 he does. Well, thank you doctor Hagstrom for joining us. 0:01:29.800 --> 0:01:32.520 Thank you glad to be here. So most of my 0:01:32.600 --> 0:01:36.520 every year has been as a mathematician, but since about 0:01:36.560 --> 0:01:41.520 twenty ten I've gradually become more interested in questions about 0:01:41.680 --> 0:01:44.080 emerging technologists and global risk. 0:01:44.319 --> 0:01:45.560 Can you tell us where you work. 0:01:45.720 --> 0:01:49.160 I'm at Chalmers University of Technology in Sweden. 0:01:49.480 --> 0:01:52.680 Okay, we're going to start with the concept of mind uploading. 0:01:52.800 --> 0:01:54.480 Can you tell us what it is for those of 0:01:54.520 --> 0:01:56.120 us that I have never heard of it before. 0:01:56.200 --> 0:01:59.240 So the idea of mind uploading, which is like a 0:01:59.320 --> 0:02:04.160 hypothetic future technology, is that if you make a simulation 0:02:04.600 --> 0:02:09.120 of the human brain, and you make it refined enough 0:02:09.680 --> 0:02:12.880 and accurate enough, it will be as good as the 0:02:12.919 --> 0:02:16.280 real thing. And the hope is that if it's good enough, 0:02:16.560 --> 0:02:21.280 we can use this technology to migrate over to machines. 0:02:21.560 --> 0:02:25.160 Our flash and blood bodies have all kinds of defects 0:02:25.280 --> 0:02:28.760 after one hundred years, so so we tend to disintegrate 0:02:28.800 --> 0:02:32.880 and die. And at least some people hope that uploading 0:02:32.919 --> 0:02:36.720 technology could need to much longer and perhaps in effect 0:02:36.919 --> 0:02:38.200 unbounded life. 0:02:38.400 --> 0:02:41.960 Wow, do you know anything about the origins of this idea. 0:02:42.080 --> 0:02:44.640 When it started, when people started thinking about it. 0:02:44.639 --> 0:02:50.400 It started in science fiction. Twentieth century science fiction writer 0:02:50.560 --> 0:02:54.760 Stanislav lam certainly wrote about it, and another person who 0:02:54.840 --> 0:02:59.520 helped popularize this concept was a philosopher, Derek Parfitt, who 0:02:59.639 --> 0:03:04.320 thought very hard about a similar concept, which is called teleportation. 0:03:04.840 --> 0:03:09.880 Like you're faxing yourself to another planet exactly, all right, 0:03:10.000 --> 0:03:14.000 that's the basic idea. Do somehow simulate our brains in 0:03:14.080 --> 0:03:19.359 a computer to escape our soft, perishable bodies. The next 0:03:19.400 --> 0:03:23.880 question is, of course, how would this work? What are 0:03:23.960 --> 0:03:27.240 the possible technologies that people have thought about doing this? 0:03:27.480 --> 0:03:28.200 Mind uploading? 0:03:28.600 --> 0:03:32.560 I should emphasize that this is not something we can 0:03:32.760 --> 0:03:35.960 expect in the next decade or so. I think that 0:03:36.040 --> 0:03:39.120 this is probably at least two or three decades away, 0:03:39.160 --> 0:03:44.080 and maybe even further. There are several steps involved in uploading. First, 0:03:44.080 --> 0:03:47.800 you need to scan the brain in extreme detail, and 0:03:48.040 --> 0:03:51.080 the main concept there is that somehow you freeze the 0:03:51.160 --> 0:03:54.400 brain and you cut it up in thin, thin, thin slices, 0:03:54.480 --> 0:03:55.600 and then you scan. 0:03:55.520 --> 0:03:57.560 Each slice whoa. 0:03:57.640 --> 0:04:01.120 And then you have to somehow organize all the information 0:04:01.240 --> 0:04:04.160 you have gotten out of this and use that on 0:04:04.200 --> 0:04:07.520 the machine. That is then meant to run your mind. 0:04:08.440 --> 0:04:12.360 That would be an example of so called destructive uploading, 0:04:13.920 --> 0:04:16.600 which requires that your brain is destroyed. 0:04:16.800 --> 0:04:18.600 Yeah, it seems you really have to commit to. 0:04:18.560 --> 0:04:23.000 It absolutely, and many people might hesitate about that. 0:04:24.160 --> 0:04:28.160 So that's option one. Freeze your brain, slice it really thin, 0:04:28.640 --> 0:04:32.839 scan each slice, and then reconstruct the three D structure 0:04:33.120 --> 0:04:36.680 of your brain, all the neurons and how they're connected 0:04:36.720 --> 0:04:39.159 to each other. Of course the problem with that is 0:04:39.160 --> 0:04:42.840 that you have to freeze your brain and slice it. 0:04:43.320 --> 0:04:46.200 Fortunately there's option number two. 0:04:47.600 --> 0:04:51.479 But the more ambitious concept if you're worry about having 0:04:51.480 --> 0:04:55.800 your brain destroyed, is the idea of non destructive uploading. 0:04:56.360 --> 0:04:59.520 So that requires figuring out how to do this scanning 0:04:59.600 --> 0:05:02.359 thing with out destroying the brain. That would be to 0:05:02.839 --> 0:05:06.800 have nano robots somehow, maybe through the blood stream, enter 0:05:06.880 --> 0:05:11.239 the brain and photograph and otherwise record what is going 0:05:11.279 --> 0:05:12.599 on there in the brain. 0:05:12.680 --> 0:05:15.719 WHOA, I thought you were going to say something like fMRI, 0:05:16.400 --> 0:05:20.479 like some kind of magnetic scanner somehow reads yourselves. Is 0:05:20.520 --> 0:05:21.479 that a possibility? 0:05:21.760 --> 0:05:25.679 I don't know. If that were doable, then that sounds 0:05:25.920 --> 0:05:27.080 less invasive. 0:05:28.480 --> 0:05:31.080 So then idia would be nano robots who'd go into 0:05:31.120 --> 0:05:33.880 your blood stream. It would get into your brain. And 0:05:33.920 --> 0:05:36.960 then how do they record all your neurons. 0:05:37.160 --> 0:05:40.320 That's something that has to be figured out, but some 0:05:40.440 --> 0:05:45.680 sort of electromagnetic recording device, and then maybe they can 0:05:46.000 --> 0:05:49.400 transfer this information the radio link or. 0:05:49.360 --> 0:05:54.839 Something that's uption number two nanobots. We'll talk about other 0:05:54.880 --> 0:05:57.960 ways to scan or capture who you are later in 0:05:58.000 --> 0:06:00.640 the program, but for now, that's pretty good. 0:06:01.080 --> 0:06:01.400 Okay. 0:06:01.520 --> 0:06:06.400 Once you scan your brain, what's thanks? Okay. And then 0:06:06.400 --> 0:06:09.560 once you scan your brain, you presumably have a picture 0:06:10.040 --> 0:06:12.760 of how all your neurons are connected to each other. 0:06:12.920 --> 0:06:16.560 Right, a kind of three D picture. But the key 0:06:16.600 --> 0:06:20.560 feature is that it's going to contain all the information 0:06:20.960 --> 0:06:24.839 about connectedness between synapses and so on and so forth, 0:06:25.160 --> 0:06:29.240 enough to reconstruct the whole brain state. And then yeah, 0:06:29.320 --> 0:06:32.359 the brain processes, the thinking and so on started on 0:06:32.440 --> 0:06:32.960 the computer. 0:06:33.320 --> 0:06:35.640 Fascinating. Okay. So, now, once you have a bottel of 0:06:35.680 --> 0:06:38.880 your brain, then you would create a new brain, or 0:06:38.920 --> 0:06:40.880 you would stimulate one in the computer. 0:06:41.360 --> 0:06:44.560 The idea would be to run this on a computer, 0:06:45.360 --> 0:06:49.400 and if this simulation is faithful enough, some people hope 0:06:49.440 --> 0:06:52.800 that this would actually be as good as having a 0:06:53.080 --> 0:06:56.640 new brain. Wow, in a sense, even better. If you 0:06:57.000 --> 0:07:01.640 are the kind of person who valuesness, maybe getting your 0:07:01.720 --> 0:07:06.320 robot body as a replacemant for your present and highly imperfect, 0:07:06.480 --> 0:07:11.520 presumably flesh and blood body. Or alternatively, if you want 0:07:11.560 --> 0:07:16.040 to live in maybe other environments than the physical environment 0:07:16.080 --> 0:07:19.400 that you maybe not maybe you could migrate to virtual 0:07:19.600 --> 0:07:24.280 worlds on some really big computer that can host maybe 0:07:24.640 --> 0:07:28.360 emulations of you and your loved ones and fans and 0:07:28.600 --> 0:07:31.640 everybody who wants to join, could maybe live in a 0:07:32.200 --> 0:07:36.200 society on this machine. It's the kind of technology that 0:07:36.360 --> 0:07:39.760 we don't know if it's doable, but we also don't 0:07:39.800 --> 0:07:44.200 have like physical limits that show that it's impossible, so 0:07:44.360 --> 0:07:45.840 it may be worth thinking about. 0:07:47.040 --> 0:07:50.360 All right, So that's uploading your brain and simulating it 0:07:50.480 --> 0:07:53.080 in a computer, and you can imagine there are several 0:07:53.200 --> 0:07:56.240 reasons you might want to do this. Maybe you're near 0:07:56.280 --> 0:07:58.200 the end of your life and this could be a 0:07:58.200 --> 0:08:02.160 way for you to live longer, maybe even forever. But 0:08:02.200 --> 0:08:05.520 according to doctor Hagstrom, there are other things you could 0:08:05.600 --> 0:08:09.720 do with this technology, starting with cloning yourself. 0:08:11.080 --> 0:08:13.920 In a sense, a more radical thing you might do 0:08:14.520 --> 0:08:19.120 is that once you have this emulation running on the machine, 0:08:19.200 --> 0:08:22.000 it's an easy thing to copy it, so you can 0:08:22.080 --> 0:08:26.040 have several emulations of yourself running on the machine. 0:08:26.080 --> 0:08:29.520 WHOA, You could clone yourself basically. 0:08:29.400 --> 0:08:32.440 Yes, And it's a much more precise thing than just 0:08:32.640 --> 0:08:36.560 genetic cloning, because if you would, through genetics create clowns 0:08:36.559 --> 0:08:40.200 on yourself, well, that would be persons who would share 0:08:40.440 --> 0:08:44.560 your g you know, but not your upbringing and education 0:08:45.000 --> 0:08:47.920 and so on. So they might be very different people, 0:08:48.280 --> 0:08:53.400 just like identical twins typically somewhat different people. But in 0:08:53.440 --> 0:08:57.960 this case you get a pair of ultra identical twins, 0:08:58.080 --> 0:09:01.880 much much more identical than the identical twins we have now. 0:09:02.840 --> 0:09:07.080 And each one of these simulations would feel like they argue, 0:09:07.360 --> 0:09:11.520 So this would give rise to maybe even legal questions 0:09:11.559 --> 0:09:16.960 about who has ownership to your possessions and if you're married, 0:09:17.720 --> 0:09:18.920 who gets yourselves? 0:09:22.320 --> 0:09:25.120 Yeah, I'm not sure my spouse can handle more than 0:09:25.400 --> 0:09:28.400 one of me. She already gets frustrated enough with one 0:09:28.400 --> 0:09:28.720 of me. 0:09:30.160 --> 0:09:33.720 So it's quite a far out scenario, and it's not 0:09:34.080 --> 0:09:38.480 easy to figure out how we would organize a society 0:09:38.640 --> 0:09:42.880 where we have easy access to this uploading technology. 0:09:43.679 --> 0:09:47.280 Yeah, once you're in the cloud, you could copy yourself, 0:09:47.640 --> 0:09:50.679 and your copies could live in different robots or in 0:09:50.800 --> 0:09:56.400 different virtual worlds, although this opens up some logistical problems. 0:09:57.840 --> 0:10:01.160 And I give one other example of a legal issue 0:10:01.200 --> 0:10:02.160 that might arise. 0:10:02.400 --> 0:10:04.120 Yeah, please, namely if. 0:10:04.040 --> 0:10:07.800 At this future time we still value the idea of 0:10:07.920 --> 0:10:13.120 democracy and every person has one vote, and an argument 0:10:13.160 --> 0:10:16.960 should be made that emulated minds should have the same 0:10:17.320 --> 0:10:21.800 rights as democratic citizens as biological minds, because they are 0:10:22.160 --> 0:10:25.559 as capable as us, and they have the same canceled 0:10:25.640 --> 0:10:28.800 feelings and so forth, so they should really have the 0:10:28.840 --> 0:10:32.160 whole package of human rights. But if there's an election 0:10:32.320 --> 0:10:36.440 coming up and you really root for this candidate rather 0:10:36.480 --> 0:10:39.360 than the other candidate, what is going to prevent you 0:10:39.480 --> 0:10:44.679 then from making a thousand emulated copies on you and 0:10:44.840 --> 0:10:47.880 the collective of you will get one thousand votes in 0:10:47.920 --> 0:10:53.240 this election. How would you construct a system of democracy 0:10:53.400 --> 0:10:57.480 which can function in a world with emulated minds without 0:10:57.600 --> 0:10:59.920 having these manipulations of the system. 0:11:00.200 --> 0:11:03.240 Oh my goodness. Yeah, that would probably blow the minds 0:11:03.240 --> 0:11:06.800 of our current politicians, who can't even handle voting by mail. 0:11:07.679 --> 0:11:10.640 Right, we can do it and Sweden, but I hear 0:11:10.720 --> 0:11:12.920 you have some problems in the United States. 0:11:14.360 --> 0:11:18.400 Yes, democracy will get kind of sticky once people start 0:11:18.480 --> 0:11:22.000 uploading and copying their brains. I mean, just think about 0:11:22.000 --> 0:11:24.360 how many votes you have to count. 0:11:25.200 --> 0:11:28.000 When coppying is cheap. I don't know what will happen 0:11:28.080 --> 0:11:31.560 to the world population. I can imagine that we might 0:11:31.720 --> 0:11:36.079 quickly go from the billions to the trillions or quadrillions 0:11:36.120 --> 0:11:36.720 of people. 0:11:37.000 --> 0:11:39.680 Do you think it'll be cheaper to run a person 0:11:39.720 --> 0:11:42.280 on a computer than to have them be alive. 0:11:42.600 --> 0:11:45.440 Yeah. One reason for thinking so is that we have 0:11:45.640 --> 0:11:50.800 this very rapid development in computer efficiency, which doesn't seem 0:11:50.880 --> 0:11:52.640 to be about to start. 0:11:52.960 --> 0:11:55.400 Yeah. Although on the plus that you wouldn't need healthcare 0:11:55.600 --> 0:12:01.840 only tech support. Hey, today's IT technicians could be tomorrow's doctors. 0:12:02.240 --> 0:12:07.520 Another interesting application of uploading your brain is space travel. 0:12:11.600 --> 0:12:15.920 There is really strong reason to replace these bodies that 0:12:15.960 --> 0:12:19.400 we now have that are extremely ill suited for space 0:12:19.440 --> 0:12:25.240 travel by robotic bodies, and uploading technology could achieve that 0:12:25.960 --> 0:12:29.679 could be a case sep in conquering the galaxy and 0:12:29.760 --> 0:12:30.679 other galaxies. 0:12:30.760 --> 0:12:34.040 This could be one of the positives of uploading your brain. 0:12:34.080 --> 0:12:38.359 If we could colonize space with our emulations of brains. 0:12:38.240 --> 0:12:40.800 Yes, potentially the biggest positives. 0:12:41.520 --> 0:12:45.800 Yes, we could upload our brains and explore the cosmos. 0:12:46.080 --> 0:12:49.600 Your mind could live inside a spacecraft and be powered 0:12:49.600 --> 0:12:53.360 for thousands of years by a solar or nuclear battery, 0:12:53.960 --> 0:12:56.600 and you could travel and get to distant stars and 0:12:56.720 --> 0:12:59.840 explore the other side of the galaxy, or visit this 0:13:00.040 --> 0:13:02.839 super massive black hole at the center of it. Or 0:13:02.880 --> 0:13:07.040 you could even imagine traveling to the edge of the universe. 0:13:07.559 --> 0:13:10.200 All right, The last question I asked doctor Hagstrom was 0:13:10.800 --> 0:13:15.000 what would it mean if we could do this? What 0:13:15.000 --> 0:13:17.360 do you think it says that uploading your mind is possible? 0:13:17.360 --> 0:13:19.080 What does it say about us as humans? 0:13:19.440 --> 0:13:25.200 Perhaps if this is true, then it gives reason for 0:13:25.320 --> 0:13:28.400 us to be a bit more humble in realizing that 0:13:28.640 --> 0:13:34.960 our present kind of existence is not as uniquely tied 0:13:35.080 --> 0:13:38.520 to our present bodies as we thought it was. 0:13:38.920 --> 0:13:41.240 The idea that we can be copied. She tells that 0:13:41.320 --> 0:13:42.439 we're replaceable. 0:13:42.640 --> 0:13:46.560 I guess, which is certainly worrying thought. And I think 0:13:46.600 --> 0:13:51.320 that it's something we should ask ourselves very carefully before 0:13:51.320 --> 0:13:54.439 we launch this technology, because if we don't think about 0:13:54.480 --> 0:13:59.120 these questions carefully and arrive at the solution of how 0:13:59.160 --> 0:14:02.720 to design in the society with uploaded minds that makes 0:14:02.800 --> 0:14:06.320 us happy and satisfied, we might end up with something 0:14:06.400 --> 0:14:08.920 that we would then deeply regret. 0:14:10.720 --> 0:14:13.920 Okay, that's a pretty good overview of what the technology is, 0:14:14.440 --> 0:14:17.120 how it might work, and what we could use it for. 0:14:17.640 --> 0:14:21.800 The next question is can it actually be done. We'll 0:14:21.800 --> 0:14:24.600 dig into the nitty gritty details of scanning your brain 0:14:24.800 --> 0:14:27.600 with a neuroscientist, and then later we'll talk to an 0:14:27.600 --> 0:14:32.520 expert about whether an uploaded you would still be you 0:14:32.920 --> 0:14:46.800 stay with us, we'll be right back. Welcome back. Okay, 0:14:46.920 --> 0:14:49.840 we're talking about uploading your brain to a computer, and 0:14:49.880 --> 0:14:52.560 so far we've learned what that means and what you 0:14:52.600 --> 0:14:54.720 could do with it. Next, we're going to talk to 0:14:54.800 --> 0:14:59.880 Professor Dwayne Godwin, a neuroscientist at Wake Forest University, about 0:15:00.000 --> 0:15:03.760 whether he thinks any of this is actually possible. Doctor 0:15:03.800 --> 0:15:06.120 Godwin and I wrote a book together called Out of 0:15:06.160 --> 0:15:09.880 Your Mind, in which we explore this and other fascinating 0:15:09.960 --> 0:15:15.440 questions about the brain. Here's what he said, what's the 0:15:15.600 --> 0:15:19.000 current state of the art in terms of scanning your brain? 0:15:19.240 --> 0:15:22.400 What kind of information into what resolution can we do that? 0:15:22.920 --> 0:15:23.200 Right? 0:15:23.240 --> 0:15:25.360 So, one of the ways that we can do it 0:15:25.400 --> 0:15:29.840 on the structural side is something called ultra high field MRI. 0:15:30.320 --> 0:15:34.520 There's this giant magnet that's alternating a magnetic field, and 0:15:34.560 --> 0:15:38.120 it's able to compose a functional image as well as 0:15:38.120 --> 0:15:40.760 the structural image of the brain. And it's able to 0:15:40.800 --> 0:15:43.840 do that at a pretty high resolutions. It can resolve 0:15:44.160 --> 0:15:48.720 the living brain at submillimeter detail, so less than a millimeter, 0:15:49.400 --> 0:15:53.000 but not at the level of synapses are on the 0:15:53.120 --> 0:15:57.960 level of microns and nanometers. That's the level at which 0:15:58.200 --> 0:16:00.800 synapses and synaptic protein are operating. 0:16:01.120 --> 0:16:03.120 I see you're not getting the details of like this 0:16:03.280 --> 0:16:06.120 neuron is sending a signal to that neuron. Right, that's 0:16:06.160 --> 0:16:06.640 too small. 0:16:06.800 --> 0:16:08.080 That's right, that's too small. 0:16:08.240 --> 0:16:10.720 Do you think that can get better over time? Like? 0:16:10.880 --> 0:16:14.040 Are people trying to make that micro level fMRI? 0:16:14.680 --> 0:16:18.240 Yeah, everything gets better all the time. I guess the 0:16:18.400 --> 0:16:22.400 question is is the current state capable of doing what 0:16:22.480 --> 0:16:26.480 you posed in your initial question? And I would say no, 0:16:26.720 --> 0:16:29.560 at the moment, it's incapable of doing that. But can 0:16:29.680 --> 0:16:33.120 you imagine that some day in the future there would 0:16:33.160 --> 0:16:36.800 be you know, sufficiently strong magnets or ways to capture 0:16:36.800 --> 0:16:39.880 that information. Yeah, sure, you could imagine that. It's just 0:16:39.920 --> 0:16:40.960 that it's not here yet. 0:16:42.120 --> 0:16:46.440 All right. That sounds promising. There's also the option of 0:16:46.600 --> 0:16:50.160 slicing your brain. Would that work to capture who you are. 0:16:51.240 --> 0:16:53.840 So there are methods where you can look at that 0:16:54.040 --> 0:17:00.120 level of detail. Unfortunately, those methods require that you die. 0:17:00.320 --> 0:17:03.160 We don't yet have wadys of looking at it without 0:17:03.360 --> 0:17:07.239 taking chunks of your brain out, essentially pickling them and 0:17:07.240 --> 0:17:11.119 then slicing them into very very thin sections of brain 0:17:11.359 --> 0:17:14.920 that we then aim an electron beam at in order 0:17:15.000 --> 0:17:18.120 to look at the proteins that are in those very 0:17:18.160 --> 0:17:22.919 fine level synapses. So yeah, it's in principle possible for 0:17:23.040 --> 0:17:26.600 you to capture that level of detail from a human brain. 0:17:26.920 --> 0:17:29.840 We do it from animal brains all the time. But 0:17:30.240 --> 0:17:34.600 what you're getting is the instant after you had actually died. 0:17:34.840 --> 0:17:38.320 So I don't think you want to be necessarily preserved 0:17:38.359 --> 0:17:39.840 in that state, right. 0:17:40.520 --> 0:17:43.600 Yeah, that would not be my preferred way to scan 0:17:43.720 --> 0:17:46.480 my brain. And the problem here is also that you'd 0:17:46.480 --> 0:17:51.879 be scanning a not so fresh brain. Yes, Like the 0:17:51.960 --> 0:17:54.520 process would have to be like I die and then 0:17:54.600 --> 0:17:57.719 my brain is sliced. It has to be sliced right 0:17:57.760 --> 0:17:59.280 away before it spoils. 0:17:59.320 --> 0:18:03.560 Basically, yeah, it would take many hours most likely before 0:18:03.680 --> 0:18:07.040 I could be placed into a state where my body 0:18:07.080 --> 0:18:09.880 could be preserved in order to be able to make 0:18:09.960 --> 0:18:13.600 those fine sections of tissue. And even then it would 0:18:13.640 --> 0:18:17.000 be after my brain had been deprived of oxygen for 0:18:17.040 --> 0:18:20.320 a while. Things might not be looking so good inside 0:18:20.400 --> 0:18:23.240 my old brain. So bringing me back might bring me 0:18:23.320 --> 0:18:25.880 back to the moment of my demise, and I might 0:18:25.880 --> 0:18:27.119 not be in such good shape. 0:18:27.160 --> 0:18:30.040 Then you might be missing parts of the brain, like 0:18:30.119 --> 0:18:32.520 it might be a corrupted scan of your brain. 0:18:32.640 --> 0:18:35.280 Yeah, exactly, all right. 0:18:35.119 --> 0:18:38.119 So you can't do any of this right now. But 0:18:38.520 --> 0:18:42.720 as doctor Godwood mention, these are technical problems. You could 0:18:42.800 --> 0:18:45.560 imagine that in the future we could come up with 0:18:45.720 --> 0:18:50.160 better fMRI technology or figure right away to perfectly slice 0:18:50.200 --> 0:18:53.520 your brain without any of it going bad. The next 0:18:53.560 --> 0:18:56.000 question I had for him was would it be enough? 0:18:56.480 --> 0:19:00.000 Would a scan of your brain be sufficient to recreate 0:19:00.440 --> 0:19:04.040 who you are? Well, let's talk a little bit about 0:19:04.119 --> 0:19:07.439 the challenges here. Like, let's say I get an fMRI 0:19:07.880 --> 0:19:11.120 down to where I can detect neurons and how they're 0:19:11.119 --> 0:19:14.520 connected to each other, basically the snapshot of the web 0:19:14.640 --> 0:19:17.280 of the network in my brain of neurons. Is that 0:19:17.440 --> 0:19:20.320 enough you think to recreate me in a computer? 0:19:20.560 --> 0:19:23.280 Well, let's think about it this way. By an analogy. 0:19:23.520 --> 0:19:26.399 Let's say you wanted to know how your computer worked. 0:19:26.720 --> 0:19:28.320 Okay, what if. 0:19:28.160 --> 0:19:31.400 You could take the very best photograph and the very 0:19:31.440 --> 0:19:35.439 best scan of the structure of your computer right now, Okay, 0:19:35.640 --> 0:19:40.359 down to the level of individual connections and logic gates 0:19:40.680 --> 0:19:43.119 like the circuit. Yeah, like the circuit. Would you be 0:19:43.200 --> 0:19:47.080 able to tell whether your computer was on social media 0:19:47.440 --> 0:19:51.359 or running Microsoft Word versus some other thing that it 0:19:51.480 --> 0:19:53.720 might be doing. So that there's kind of a difference 0:19:53.800 --> 0:19:56.600 here between getting a good picture of the hardware of 0:19:56.600 --> 0:20:00.000 the brain and the ability to understand what the brain 0:20:00.240 --> 0:20:01.719 is doing at any given moment. 0:20:02.720 --> 0:20:05.320 What doctor Godwin is saying here is even if we 0:20:05.359 --> 0:20:08.840 could scan your brain down to the molecular level, that 0:20:08.880 --> 0:20:14.280 doesn't mean we understand how the brain works. Let's say 0:20:14.400 --> 0:20:18.080 I'm able to perfectly scan your brain and detect every 0:20:18.200 --> 0:20:21.520 neuron connection it's in abs Wait, I have all this 0:20:21.600 --> 0:20:24.440 information about my brain, and let's say it's pretty intact. 0:20:24.560 --> 0:20:26.000 Could I get a computer to run it? 0:20:26.440 --> 0:20:29.200 You know, it's always a mistake to say something can't 0:20:29.280 --> 0:20:31.399 be done. The best I can say right now is 0:20:31.440 --> 0:20:33.440 that we're not in a position to do that right now. 0:20:33.640 --> 0:20:37.720 But with some of the new computer architectures and chip architectures. 0:20:37.840 --> 0:20:40.639 You know, there are these things called neuromorphic chips. I 0:20:40.680 --> 0:20:44.240 think we could probably run simulations at some point in 0:20:44.280 --> 0:20:47.600 the future that will almost perfectly emulate a lot of 0:20:47.640 --> 0:20:48.840 the different states. 0:20:48.440 --> 0:20:49.000 Of your brain. 0:20:49.240 --> 0:20:52.679 And I think, you know, looking who knows ten one 0:20:52.720 --> 0:20:56.679 hundred years into the future, when AI is actually able 0:20:56.920 --> 0:20:59.840 to help us, not only mimic the brain, but all 0:21:00.080 --> 0:21:02.239 so help us and tell us the things that need 0:21:02.280 --> 0:21:05.840 to be done to improve resolution and computation, then yeah, 0:21:06.080 --> 0:21:08.240 I could see these things happening in the future. 0:21:08.480 --> 0:21:09.879 I thought you were going to say that it was 0:21:09.960 --> 0:21:13.280 not possible because we don't know all the processes that 0:21:13.320 --> 0:21:16.800 are happening at the synapse level and the neurons. We 0:21:16.880 --> 0:21:19.639 don't know yet how to create a simulation of the brain. 0:21:20.000 --> 0:21:22.359 Well, I think that's correct that we don't yet know 0:21:22.400 --> 0:21:24.280 how to do that, But I think your question was 0:21:24.320 --> 0:21:26.640 broader about will we ever be able to do that. 0:21:26.880 --> 0:21:30.160 I'm an optimist about these things. I think there could 0:21:30.160 --> 0:21:33.360 be at some point in the future a powerful enough computer, 0:21:33.560 --> 0:21:37.840 enough computational power, enough understanding of how these different elements 0:21:37.880 --> 0:21:40.360 within the brain, the different connections are made, and how 0:21:40.440 --> 0:21:42.760 information flows from one part of the brain to the 0:21:42.800 --> 0:21:45.560 other that, yeah, you could run some sort of simulation 0:21:45.680 --> 0:21:48.600 that is very much like you, but we're aways from that. 0:21:50.119 --> 0:21:52.919 What doctor Godwin is saying here is that right now 0:21:53.200 --> 0:21:56.960 we just don't yet fully understand how the brain works. 0:21:57.480 --> 0:22:00.440 But if doctor Godwin said, with more signs, we could 0:22:00.600 --> 0:22:03.919 know all of this in the future. Until then, I 0:22:03.960 --> 0:22:07.640 asked doctor Godwin about an alternate scenario, one in which 0:22:07.640 --> 0:22:10.760 you maybe don't have to scan all of your brain, 0:22:11.200 --> 0:22:17.760 and that is using AI or artificial intelligence. Okay, next question, 0:22:18.000 --> 0:22:20.359 do you think you could train an AI to be. 0:22:20.480 --> 0:22:23.160 You could do you train an AI to think like you? Well, 0:22:23.320 --> 0:22:26.800 I think that you could certainly train an AI to 0:22:26.880 --> 0:22:30.280 follow you around, follow your movements, listen to your voice, 0:22:30.600 --> 0:22:34.359 see what you see. Maybe you've written several books, you know, 0:22:34.400 --> 0:22:36.760 the AI could train on that and make an astment 0:22:36.800 --> 0:22:39.600 of what it would be like to have an AI 0:22:39.720 --> 0:22:41.760 Woge or an AI Dwyane. 0:22:42.160 --> 0:22:45.119 Okay, how about this sci fi scenario. What if I 0:22:45.320 --> 0:22:49.040 create an AI that is conscious, that is like thinking 0:22:49.520 --> 0:22:52.480 and is sentient, but it's sort of like a blank 0:22:52.560 --> 0:22:54.560 version of a human, And then I can go, Okay, 0:22:54.680 --> 0:22:56.959 I make a copy of this and train it on Whoorge, 0:22:57.160 --> 0:23:00.119 so that it's basically like a sentient being trained to 0:23:00.160 --> 0:23:01.160 be like Orge. 0:23:00.920 --> 0:23:03.280 Yeah, that's not a bad approach. And I wrote a 0:23:03.320 --> 0:23:05.920 blog post some years ago where I was saying, well, 0:23:05.960 --> 0:23:08.320 maybe this is a path because you know, there are 0:23:08.400 --> 0:23:11.840 probably some aspects of how the brain works that are 0:23:11.920 --> 0:23:16.320 common to almost everyone. So for example, walking walking is 0:23:16.359 --> 0:23:19.560 a thing that people with two legs all do, and 0:23:19.720 --> 0:23:24.199 that program is probably pretty consistent from one person to another. 0:23:24.440 --> 0:23:27.040 So you could see stringing together a number of those 0:23:27.040 --> 0:23:30.639 different things and then say, Okay, let's put that in 0:23:30.720 --> 0:23:35.320 an android brain or body, and then let's just focus 0:23:35.400 --> 0:23:38.360 on the thing that makes four hay or hay interesting. 0:23:38.520 --> 0:23:41.919 Meaning maybe you don't need to upload your whole brain, Like, 0:23:41.960 --> 0:23:44.399 maybe you don't need to scan every single synapse and 0:23:44.440 --> 0:23:47.080 you're run in your brain just the parts that really 0:23:47.080 --> 0:23:48.680