Latest / Deep Tech Germany – AI, Robotics & Frontier Innovation by Startuprad.io™ / Avi Loeb: Dark Matter, Dark Energy & AI in Scientific Discovery
Transcript
- WEBVTT 1 00:00:00.320 --> 00:00:04.160 Where do you I is it okay here? Yep. Close to my 2 00:00:04.319 --> 00:00:07.925 We we're already getting sound, so Okay. That's fine. And the only 3 00:00:07.925 --> 00:00:11.705 question is for the camera, it may be too big. No? 4 00:00:12.005 --> 00:00:15.445 No. No. It will be fine. It's okay? Or on the tie, would you 5 00:00:15.445 --> 00:00:18.680 prefer? No. I think there is better. It's better. Have it on the tie. It 6 00:00:18.680 --> 00:00:22.520 will make noise. On the shirt. Yep. Okay. Sounds great. As 7 00:00:22.520 --> 00:00:26.025 long as you are happy, I'm happy. We 8 00:00:26.185 --> 00:00:29.305 stick exactly to the questions that I've sent you. Only, 9 00:00:29.785 --> 00:00:33.385 upfront a little introduction. Oh, okay. I think you can 10 00:00:33.385 --> 00:00:36.969 introduce yourself. Right? You you've done that. Yeah. If you want me, I 11 00:00:36.969 --> 00:00:40.730 can, okay. So 12 00:00:40.730 --> 00:00:44.565 in terms of the questions, I mean, you can choose whatever you like. 13 00:00:45.265 --> 00:00:48.004 If you feel that you want to ask another follow-up question. 14 00:00:49.880 --> 00:00:53.579 I need to draw at least a slight, slight, curtains. Yeah. 15 00:00:53.800 --> 00:00:57.500 Too much light. But then you don't see the river. 16 00:00:58.225 --> 00:01:01.665 Well Is it okay if I sit 17 00:01:01.665 --> 00:01:05.425 here? Unfortunately, I only 18 00:01:05.425 --> 00:01:09.180 see a negative picture of you. Just like the outline. 19 00:01:09.560 --> 00:01:13.080 My camera doesn't work like that. It's sorry? Can you move a little 20 00:01:13.080 --> 00:01:16.600 bit over here to your right? Or we can close this curtain. 21 00:01:16.600 --> 00:01:20.445 Yeah. We could try. We could try. We can try. 22 00:01:21.385 --> 00:01:23.705 Yeah. As long as you don't need the river, it should be 23 00:01:27.390 --> 00:01:31.229 Yeah. But then do you have enough light or we need to turn on some 24 00:01:31.229 --> 00:01:34.990 light for you? That that one would be great. Yeah. That should 25 00:01:34.990 --> 00:01:38.695 totally do it. Yeah. The the other thing we can do in the post production, 26 00:01:39.075 --> 00:01:42.915 so we have enough light. We have enough sun from you. 27 00:01:42.915 --> 00:01:46.590 Okay. And so we can You think it's okay? How 28 00:01:46.590 --> 00:01:50.430 about the chair? Is it, sufficiently high? I'll 29 00:01:50.430 --> 00:01:53.884 adjust my camera. Don't worry about that. You can lower it. 30 00:01:53.945 --> 00:01:57.784 Yeah. Okay. Does that look 31 00:01:57.784 --> 00:02:01.380 good? I think so. I'll get a little bit 32 00:02:01.380 --> 00:02:04.980 closer. Did you change the height of this or no? No, no. Because that can 33 00:02:04.980 --> 00:02:07.480 also, we can change that also. 34 00:02:09.625 --> 00:02:13.305 I lost just one of the screws here in my tripod when I 35 00:02:13.305 --> 00:02:16.880 was traveling with the train here, so I always have to make sure it doesn't 36 00:02:16.880 --> 00:02:20.560 fall off the other side. But the the 37 00:02:20.720 --> 00:02:24.375 is it stable? It's okay. At least so, yeah. Just make 38 00:02:24.375 --> 00:02:27.675 sure that it's straight, because it looks a little tilted. 39 00:02:28.935 --> 00:02:32.640 Yeah, exactly. Better? Now it's better, I think. 40 00:02:32.780 --> 00:02:36.000 Yeah. You have a good eye for that. Well, you're a physicist, right? 41 00:02:41.715 --> 00:02:44.145 Now we also do have a recording. 42 00:02:46.690 --> 00:02:49.110 Looks good? Looks amazing. 43 00:02:52.215 --> 00:02:56.055 Okay. And now I'll send my piece. Hello. In terms 44 00:02:56.055 --> 00:02:59.655 of the light, is it good this way or you want it close, in this 45 00:02:59.655 --> 00:03:03.460 direction? A little bit more too too. Yeah. You can tilt it. Yeah. 46 00:03:03.840 --> 00:03:07.520 It's up I don't know how it looks. I 47 00:03:07.520 --> 00:03:10.045 can actually turn it around. You see here? Ah. 48 00:03:11.465 --> 00:03:14.905 But then there is some light on top of me. Where is that coming 49 00:03:14.905 --> 00:03:18.600 from? I think from the curtains, from the view, between the 50 00:03:18.600 --> 00:03:20.780 curtains. Because that doesn't look good, actually. 51 00:03:22.680 --> 00:03:24.040 Okay. So if we do that 52 00:03:28.105 --> 00:03:29.805 Okay. Maybe now it's better. 53 00:03:31.465 --> 00:03:35.245 Yeah. Now now it's better. Definitely. I'm okay. Excellent. 54 00:03:37.819 --> 00:03:39.680 Okay. Let me do one clap. 55 00:03:41.500 --> 00:03:45.094 That's the sign where I can synchronize it. Hello 56 00:03:45.094 --> 00:03:48.474 and welcome everybody to another episode of Stalowave. Io. 57 00:03:48.694 --> 00:03:52.270 Today, we are attending Curis 2024, and 58 00:03:52.270 --> 00:03:55.970 we just got an amazing interview guest, professor 59 00:03:56.110 --> 00:03:59.775 Alfi Loeb from Harvard University. Very welcome 60 00:04:00.155 --> 00:04:03.695 to Startuprate.l. Can you briefly introduce 61 00:04:03.755 --> 00:04:07.590 yourself? It's a great pleasure to join you. My name 62 00:04:07.590 --> 00:04:11.190 is Avi Loeb. Fundamentally, I'm a curious farm 63 00:04:11.190 --> 00:04:14.790 boy. I was born on a farm and was connected to 64 00:04:14.790 --> 00:04:18.524 nature since a young age and became a scientist 65 00:04:18.524 --> 00:04:21.824 so that I can answer questions about nature without 66 00:04:22.285 --> 00:04:25.905 needing to listen to the adults in the room telling me the answers, 67 00:04:26.400 --> 00:04:29.760 telling me their opinions. And I 68 00:04:29.760 --> 00:04:33.300 was chair of the astronomy department at Harvard for 9 years, 69 00:04:33.440 --> 00:04:36.854 the longest serving chair. I also served as the 70 00:04:36.854 --> 00:04:40.615 founding director of Harvard's Black Hole Initiative, and I'm 71 00:04:40.615 --> 00:04:44.215 the current director of Harvard's Institute For Theory and 72 00:04:44.215 --> 00:04:47.550 Computation. I worked primarily on 73 00:04:47.690 --> 00:04:51.290 theoretical astrophysics, studies of black 74 00:04:51.290 --> 00:04:54.905 holes, the first stars in the universe, and most recently 75 00:04:55.445 --> 00:04:59.285 the search for extraterrestrial intelligence because we don't find it 76 00:04:59.285 --> 00:05:02.669 here on earth. That's already 77 00:05:02.669 --> 00:05:06.430 pretty exciting. Since you are a 78 00:05:06.430 --> 00:05:09.824 scientist, what do you believe are the most exciting 79 00:05:09.965 --> 00:05:13.724 future avenues for scientific research? Well, 80 00:05:13.724 --> 00:05:17.540 there are some fundamental questions that we don't have the answer to. 81 00:05:18.020 --> 00:05:21.860 I call these the known unknowns. In other 82 00:05:21.860 --> 00:05:25.635 words, these are things we know that we don't know. 83 00:05:26.355 --> 00:05:29.875 For example, what most of the matter in the universe is, 84 00:05:29.875 --> 00:05:33.395 85% of the matter is of a substance that 85 00:05:33.395 --> 00:05:36.830 we never identified, never detected. And 86 00:05:36.830 --> 00:05:40.450 so we are still in the dark, and it's called dark matter. 87 00:05:41.854 --> 00:05:45.615 There is also the substance that makes up the 88 00:05:45.615 --> 00:05:49.320 vacuum. You might think the vacuum is empty, but 89 00:05:49.320 --> 00:05:53.080 in fact not so. It turns out that there is some energy 90 00:05:53.080 --> 00:05:56.785 per unit volume and that is called dark 91 00:05:56.785 --> 00:06:00.465 energy. And we know that it exists because the universe is not just 92 00:06:00.465 --> 00:06:04.005 expanding, but, its expansion is accelerating 93 00:06:04.625 --> 00:06:08.110 over time. And, that means that there is a 94 00:06:08.110 --> 00:06:11.570 repulsive gravity pushing the universe apart. 95 00:06:12.190 --> 00:06:15.985 And the question is, what is the origin of this dark 96 00:06:15.985 --> 00:06:19.205 energy? Why is it there? Why isn't the vacuum 97 00:06:19.745 --> 00:06:21.685 completely empty, for example? 98 00:06:23.390 --> 00:06:27.150 And we also know that, if we go back in time, there 99 00:06:27.150 --> 00:06:30.290 was a beginning to the universe, the big bang, 100 00:06:31.315 --> 00:06:35.075 and we don't know what happened before the big bang. These 101 00:06:35.075 --> 00:06:38.790 are fundamental questions that we don't have answers to. We also 102 00:06:38.790 --> 00:06:41.370 don't know what happens inside the black hole. 103 00:06:42.470 --> 00:06:45.990 Where does the matter that falls into a black hole, where does it 104 00:06:45.990 --> 00:06:49.735 collect? Does it collect at the center or 105 00:06:49.735 --> 00:06:53.095 does it go somewhere else? Like water going down the 106 00:06:53.095 --> 00:06:55.195 sewer of our homes. 107 00:06:57.080 --> 00:07:00.919 So these are just some fundamental questions about the universe, but there 108 00:07:00.919 --> 00:07:03.580 are many more that I can mention. 109 00:07:05.574 --> 00:07:08.995 And I think the most fundamental in my mind is 110 00:07:09.414 --> 00:07:13.034 where did we come from? What was there before the Big Bang? 111 00:07:13.220 --> 00:07:16.980 And in particular, I'm curious to know whether the 112 00:07:16.980 --> 00:07:20.760 Big Bang was created artificially by a scientist. 113 00:07:21.995 --> 00:07:25.755 Because if we ever understand how to unify quantum mechanics and 114 00:07:25.755 --> 00:07:29.455 gravity, we might have quantum gravity engineers 115 00:07:30.030 --> 00:07:33.790 that would have the recipe for creating a baby 116 00:07:33.790 --> 00:07:37.630 universe in the laboratory. That are 117 00:07:37.630 --> 00:07:40.665 all very interesting avenues for future research. 118 00:07:41.925 --> 00:07:45.764 What emerging technologies do you think will have the 119 00:07:45.764 --> 00:07:49.490 greatest impact on scientific advancement in the next few years? 120 00:07:49.490 --> 00:07:52.310 Talking about the instruments you can use for research. 121 00:07:53.250 --> 00:07:56.645 Well, in terms of instruments, of course, the most 122 00:07:56.645 --> 00:08:00.085 important is computers, if you call it 123 00:08:00.085 --> 00:08:03.865 instruments. I mean, that improves our cognitive 124 00:08:03.925 --> 00:08:07.580 abilities, and artificial intelligence, machine learning 125 00:08:08.039 --> 00:08:11.560 are entering into the scientific study of large data 126 00:08:11.560 --> 00:08:12.955 sets and 127 00:08:15.814 --> 00:08:19.574 also going through data that the 128 00:08:19.574 --> 00:08:22.960 human brain is unable to 129 00:08:22.960 --> 00:08:25.860 process and that already starts 130 00:08:26.560 --> 00:08:30.014 entering physics and biology. 131 00:08:30.715 --> 00:08:34.335 But I see a great future ahead for artificial 132 00:08:34.395 --> 00:08:38.100 intelligence machine learning. There is an interesting 133 00:08:38.100 --> 00:08:41.780 question of whether those computer 134 00:08:41.780 --> 00:08:43.559 programs will be able to identify 135 00:08:45.305 --> 00:08:48.745 trends in data and figure out new laws of 136 00:08:48.745 --> 00:08:52.425 nature that we have never figured out before. So will they 137 00:08:52.425 --> 00:08:56.120 behave like scientists? And if they do, will 138 00:08:56.120 --> 00:08:59.880 I be out of a job? For the near 139 00:08:59.880 --> 00:09:02.459 future, I don't think so. I still think that 140 00:09:03.735 --> 00:09:07.355 at least many people complain that AI systems are hallucinating. 141 00:09:08.295 --> 00:09:11.890 I wouldn't hold that against them because I have a lot of summer 142 00:09:11.890 --> 00:09:15.730 interns that are hallucinating, but, they are 143 00:09:15.730 --> 00:09:19.430 not at the level where they can innovate, the way scientists, 144 00:09:19.975 --> 00:09:23.274 human scientists do, but perhaps in the future they will. 145 00:09:23.654 --> 00:09:27.334 And, I'm open minded about that. I think that will be the biggest breakthrough 146 00:09:27.334 --> 00:09:29.980 when science will be done, 147 00:09:30.920 --> 00:09:34.760 with, the aid of AI. And then the question is who to give the 148 00:09:34.760 --> 00:09:38.495 Nobel prize? It's already a good question right now 149 00:09:38.495 --> 00:09:41.855 because there 150 00:09:41.855 --> 00:09:45.310 was a 151 00:09:45.310 --> 00:09:48.370 major problem in biology that was resolved 152 00:09:49.310 --> 00:09:53.095 by AI. And, you know, 153 00:09:53.095 --> 00:09:56.535 the question is who should be rewarded for 154 00:09:56.535 --> 00:09:57.035 that. 155 00:10:02.199 --> 00:10:05.180 Now, maybe you remember, but it was 156 00:10:06.199 --> 00:10:09.425 what's the name of this British AI company that 157 00:10:09.565 --> 00:10:10.065 solved 158 00:10:13.805 --> 00:10:15.345 the structure of 159 00:10:18.699 --> 00:10:22.379 proteins. The protein folding problem, yeah. That 160 00:10:22.379 --> 00:10:25.575 was it. I mean, I can say it if you want. 161 00:10:27.015 --> 00:10:29.255 Protein folding, it was solved by, 162 00:10:30.855 --> 00:10:33.435 a company, an AI company in England. 163 00:10:34.680 --> 00:10:36.460 But I forgot the name of the company. 164 00:10:39.000 --> 00:10:39.980 Protein Folding. 165 00:10:43.175 --> 00:10:46.875 Here it says deep mind? Yes, deep mind. Exactly. Deep mind. 166 00:10:47.095 --> 00:10:49.675 So I can just say it and then you can. Yeah. 167 00:10:51.500 --> 00:10:54.880 For example, there is already a precedent of the DeepMind 168 00:10:55.420 --> 00:10:59.195 company being able to solve the 169 00:10:59.195 --> 00:11:00.655 protein folding challenge 170 00:11:02.955 --> 00:11:06.529 that was really very difficult to solve for 171 00:11:06.529 --> 00:11:10.209 human scientists, but this AI code 172 00:11:10.209 --> 00:11:13.764 that they developed went through so many configurations that it 173 00:11:13.764 --> 00:11:16.745 was able to figure out how 174 00:11:18.324 --> 00:11:21.700 proteins fold when they have very big 175 00:11:21.700 --> 00:11:25.240 molecules. So that's just one example where a major 176 00:11:25.780 --> 00:11:28.920 challenge in science was resolved by AI already. 177 00:11:31.134 --> 00:11:34.975 And, we're talking about research, research instruments, 178 00:11:34.975 --> 00:11:38.814 and of course everybody here is interested in startups who are listening to 179 00:11:38.814 --> 00:11:42.590 this. Where would you see the role in startups and how they 180 00:11:42.590 --> 00:11:46.430 can contribute? You've already hinted at AI. Do do you 181 00:11:46.430 --> 00:11:50.115 see a big potential there? Yeah. The most important 182 00:11:50.115 --> 00:11:53.175 thing about the innovation is that one should, 183 00:11:53.795 --> 00:11:57.550 take risks because very often the most important 184 00:11:57.550 --> 00:12:01.310 breakthroughs are not recognized ahead of time. Otherwise they would 185 00:12:01.310 --> 00:12:05.009 be accomplished quickly. And so you need to explore 186 00:12:06.735 --> 00:12:10.415 territories that are unknown and very often you 187 00:12:10.415 --> 00:12:13.775 fail. But you do that because there is a chance that you will be 188 00:12:13.775 --> 00:12:17.420 successful. And if you do you are successful, it would be a 189 00:12:17.420 --> 00:12:21.260 major advance. And so having startups is 190 00:12:21.260 --> 00:12:24.080 one way of exploring the land of the unknown, 191 00:12:25.175 --> 00:12:28.235 where, people with innovative ideas 192 00:12:28.775 --> 00:12:32.155 try to materialize them and make them real and, 193 00:12:33.410 --> 00:12:37.170 very often they fail. But it's part of the process 194 00:12:37.170 --> 00:12:40.915 of arriving at important breakthroughs. And that's 195 00:12:40.915 --> 00:12:44.675 why startups are so important. The same is true about 196 00:12:44.675 --> 00:12:48.295 the science. I do think that funding should be given, a small 197 00:12:48.670 --> 00:12:52.510 fraction of funding should be given to risky propositions that 198 00:12:52.510 --> 00:12:56.270 are not proven and, nevertheless, could 199 00:12:56.270 --> 00:12:59.425 change the landscape of scientific discovery. 200 00:13:00.205 --> 00:13:03.825 So startups play this role in the commercial sector where they 201 00:13:04.420 --> 00:13:08.260 explore opportunities and I'm 202 00:13:08.260 --> 00:13:11.960 very much in the view of letting many flowers bloom 203 00:13:12.415 --> 00:13:15.774 because you never know which would be the most beautiful flower ever 204 00:13:15.774 --> 00:13:19.535 seen. Mhmm. You kind of, 205 00:13:19.935 --> 00:13:22.980 hit already next 2 questions that I had on my list. 206 00:13:24.240 --> 00:13:28.020 We talked about startups and working together in scientific 207 00:13:28.160 --> 00:13:31.945 research. What are the main challenges that startups face when 208 00:13:31.945 --> 00:13:34.365 trying to contribute to scientific research? 209 00:13:35.785 --> 00:13:38.980 Surprisingly, there is a lot of conservatism 210 00:13:39.440 --> 00:13:42.020 in science, even though it's supposed to be 211 00:13:43.280 --> 00:13:46.020 open minded endeavor with blue sky 212 00:13:47.755 --> 00:13:50.975 research in the sense that it's not supposed to be applied. 213 00:13:51.515 --> 00:13:55.160 But what happens is that there is another 214 00:13:56.019 --> 00:13:59.620 factor that enters the calculation, and that's the human 215 00:13:59.620 --> 00:14:02.920 ego, which is very prominent in academia. 216 00:14:03.815 --> 00:14:07.515 And, on the one hand you have scientists who 217 00:14:08.055 --> 00:14:11.435 are driven by their ego to get honors, awards, recognition, 218 00:14:12.020 --> 00:14:15.540 and are building echo chambers around them of students and 219 00:14:15.540 --> 00:14:19.240 postdocs. But at the same time they are 220 00:14:19.380 --> 00:14:22.975 pushing back against the new ideas because they 221 00:14:22.975 --> 00:14:26.415 got their stature, their position, their status, their 222 00:14:26.415 --> 00:14:30.175 recognition based on past knowledge. And so when 223 00:14:30.175 --> 00:14:33.899 they see new knowledge emerging, for example, when they 224 00:14:33.899 --> 00:14:37.440 see anomalies, deviations, outliers 225 00:14:38.625 --> 00:14:42.465 from what is known, they often dismiss it or shove it under the 226 00:14:42.465 --> 00:14:46.140 rug or say the uncertainties are too big. They 227 00:14:46.360 --> 00:14:50.040 basically raise enough dust in the air to claim that they 228 00:14:50.040 --> 00:14:53.640 don't see anything new. And they do it because they 229 00:14:53.640 --> 00:14:57.105 protect their territory of past knowledge. 230 00:14:57.885 --> 00:15:01.245 And unfortunately that has the negative 231 00:15:01.245 --> 00:15:04.779 effect of suppressing innovation because when young people see 232 00:15:04.779 --> 00:15:08.140 that, they worry about their careers, their job 233 00:15:08.140 --> 00:15:11.765 opportunities, and and as a result, they don't take risks. 234 00:15:11.765 --> 00:15:15.305 They don't pay attention to anomalies. They don't explore 235 00:15:16.005 --> 00:15:19.660 ideas that deviate from what the senior established 236 00:15:19.720 --> 00:15:23.180 people already, announced as valid 237 00:15:25.160 --> 00:15:28.825 areas of research. And it gets to a point where the 238 00:15:28.825 --> 00:15:32.585 theoretical physics community, for example, has been working 239 00:15:32.585 --> 00:15:36.265 for 50 years on, how to unify 240 00:15:36.265 --> 00:15:39.730 quantum mechanics and gravity in the form of string theory, which 241 00:15:39.730 --> 00:15:42.870 is a theory that is mathematically 242 00:15:43.330 --> 00:15:46.825 very sophisticated, but has not been demonstrated 243 00:15:47.365 --> 00:15:51.045 to be true by any experiment. And doing that for 244 00:15:51.045 --> 00:15:54.700 50 years means that the entire community is 245 00:15:54.700 --> 00:15:58.540 basically driven by societal forces. What 246 00:15:58.620 --> 00:16:02.155 by some leaders that say this is interesting for us to 247 00:16:02.235 --> 00:16:05.835 explore without any experimental verification. We know from the history of 248 00:16:05.835 --> 00:16:09.595 science that you always need evidence to guide you and you may be going 249 00:16:09.595 --> 00:16:13.440 in the wrong direction. Nobody would have forecasted quantum mechanics to be 250 00:16:13.440 --> 00:16:17.279 real, unless experiments would lead us in that direction. 251 00:16:17.279 --> 00:16:20.899 We still don't understand quantum mechanics. So my point is that 252 00:16:21.035 --> 00:16:24.795 we really need to be humble, in physics and let 253 00:16:24.795 --> 00:16:27.855 experiments guide us. And we can have a lot of ideas, 254 00:16:28.600 --> 00:16:31.500 but at the same time we should allow 255 00:16:32.440 --> 00:16:36.279 them to be ruled out by experiments and then move on. And if experiments 256 00:16:36.279 --> 00:16:39.825 are not available, there is, we should let 257 00:16:39.825 --> 00:16:43.505 other ideas, appear. And 258 00:16:43.505 --> 00:16:47.280 that's not the case in theoretical physics at the moment. There is only one 259 00:16:47.280 --> 00:16:50.980 mainstream even though it didn't get support from data. 260 00:16:51.600 --> 00:16:54.885 And so I see that as driven 261 00:16:55.105 --> 00:16:58.785 by intellectual gymnastics, by people trying to show off, just 262 00:16:58.785 --> 00:17:02.269 show that they are smart by societal forces, not 263 00:17:02.269 --> 00:17:06.109 the raw curiosity about nature that is supposed to lead 264 00:17:06.109 --> 00:17:09.569 science. And it's not just in theoretical 265 00:17:09.950 --> 00:17:13.345 physics, it's in many other areas where, 266 00:17:15.085 --> 00:17:17.724 progress was suppressed. I mean, even when, 267 00:17:18.524 --> 00:17:22.180 Alfred Wegener realized in 1910 268 00:17:22.480 --> 00:17:25.859 that the continents perhaps were all 269 00:17:25.920 --> 00:17:29.655 together based on the fact 270 00:17:29.655 --> 00:17:33.175 that they match like a jigsaw puzzle. His theory was 271 00:17:33.175 --> 00:17:36.855 ridiculed by experts for decades and he died 272 00:17:36.855 --> 00:17:40.680 in the thirties thinking that it will never be 273 00:17:40.680 --> 00:17:44.280 accepted. But then in the 19 fifties, evidence about 274 00:17:44.280 --> 00:17:47.755 rocks and about vegetation confirmed it. That's a 275 00:17:47.755 --> 00:17:51.595 classic example where the progress was delayed by 40 276 00:17:51.595 --> 00:17:55.215 years. And the same is true in many other frontiers, 277 00:17:56.470 --> 00:18:00.250 where the correct ideas were simply ridiculed by 278 00:18:00.630 --> 00:18:03.450 experts that wanted to maintain their, 279 00:18:04.150 --> 00:18:07.745 intellectual turf. Mhmm. Talk about 280 00:18:07.804 --> 00:18:11.404 investments in research here. How important is investment in 281 00:18:11.404 --> 00:18:15.190 technology and science for the future of research? Technology, 282 00:18:16.210 --> 00:18:19.350 goes hand in hand with science because fundamental science, 283 00:18:20.290 --> 00:18:23.915 allows us to understand nature. And then once we understand 284 00:18:23.915 --> 00:18:26.255 nature we can use it for our 285 00:18:27.515 --> 00:18:31.295 purposes. And, the know how 286 00:18:31.860 --> 00:18:35.620 of how to use nature for what we want to accomplish comes 287 00:18:35.620 --> 00:18:39.160 from fundamental science and therefore technology always 288 00:18:39.300 --> 00:18:43.145 makes use of fundamental science and, that's 289 00:18:43.145 --> 00:18:46.445 why science and technology should be funded. 290 00:18:46.825 --> 00:18:50.670 Now technology is using what we know about science for 291 00:18:50.670 --> 00:18:53.250 the benefit of humanity. That's extremely important, 292 00:18:54.030 --> 00:18:57.470 because once, for example, we developed the 293 00:18:57.470 --> 00:19:00.895 computer after understanding how to make transistors, 294 00:19:01.835 --> 00:19:05.675 then, and that came from understanding quantum mechanics, 295 00:19:05.675 --> 00:19:09.520 you know, so everything is entangled, but the new technologies 296 00:19:09.660 --> 00:19:12.800 that were developed allow science to progress 297 00:19:13.580 --> 00:19:16.940 much more because those technologies can be used in 298 00:19:16.940 --> 00:19:19.585 sensors that are taking data, 299 00:19:20.845 --> 00:19:24.684 and they allow us to do computations, meaning processing 300 00:19:24.684 --> 00:19:28.500 of the data. So I think it's extremely important to support 301 00:19:28.500 --> 00:19:31.400 both fundamental science and technology 302 00:19:32.500 --> 00:19:36.235 because they progress hand in hand. The strange 303 00:19:36.235 --> 00:19:39.755 thing is that so much money was made in the 304 00:19:39.755 --> 00:19:42.335 commercial sector that you might find that innovation, 305 00:19:43.500 --> 00:19:47.340 actually happens more routinely in companies these days 306 00:19:47.340 --> 00:19:50.880 than in academia, or maybe they sort of go 307 00:19:51.500 --> 00:19:54.924 at similar levels in the case of 308 00:19:54.924 --> 00:19:58.225 quantum computers or in the case of AI, 309 00:19:59.485 --> 00:20:03.020 whereas it was supposed to be the responsibility 310 00:20:03.160 --> 00:20:06.220 of academia to really promote innovation 311 00:20:07.000 --> 00:20:10.375 much more than people who are aiming to make 312 00:20:10.375 --> 00:20:13.895 money. Talking about the 313 00:20:13.895 --> 00:20:17.550 impact on society, how do you envision new 314 00:20:17.550 --> 00:20:21.070 scientific advancements impacting everyday life and 315 00:20:21.070 --> 00:20:23.810 society as a whole? I think, 316 00:20:24.935 --> 00:20:28.535 science and technology is what makes us look like an 317 00:20:28.535 --> 00:20:31.355 intelligent civilization because if you read the news 318 00:20:33.090 --> 00:20:36.850 every day, you realize, you know, we are doing a lot of foolish 319 00:20:36.850 --> 00:20:38.710 things. We invest $2,400,000,000,000 320 00:20:40.850 --> 00:20:43.965 every year worldwide in military 321 00:20:44.825 --> 00:20:48.184 budgets, meaning in conflicts, in trying to kill each 322 00:20:48.184 --> 00:20:50.284 other. That makes very little sense. 323 00:20:51.789 --> 00:20:55.250 We engage routinely in zero sum games, 324 00:20:55.870 --> 00:20:57.330 conflicts over territories, 325 00:20:59.635 --> 00:21:03.475 whether, you know, these are just territories on the rock that was left 326 00:21:03.475 --> 00:21:07.150 over from the formation of the sun. It's a tiny real estate 327 00:21:07.150 --> 00:21:10.590 that we are focusing on. There is so much more real estate in 328 00:21:10.590 --> 00:21:13.010 interstellar space if we were to look up. 329 00:21:14.030 --> 00:21:17.684 And the intelligent thing to do is 330 00:21:17.684 --> 00:21:21.365 actually for us to work together, to collaborate, not to engage 331 00:21:21.365 --> 00:21:24.660 in conflicts because science is an infinite 332 00:21:24.960 --> 00:21:28.640 sum game. It's not a zero sum game. If we 333 00:21:28.640 --> 00:21:31.940 gain knowledge, everyone benefits from it. 334 00:21:32.845 --> 00:21:35.825 So what I argue for is science and technology 335 00:21:36.845 --> 00:21:40.660 are much better than politics, Because politics is 336 00:21:40.660 --> 00:21:43.799 very often a zero sum game, whereas science and technology 337 00:21:44.500 --> 00:21:46.200 are infinite sum games. 338 00:21:48.585 --> 00:21:52.045 Coming close to the end. Advice for aspiring entrepreneurs. 339 00:21:52.585 --> 00:21:55.840 What advice would you give to aspiring entrepreneurs 340 00:21:56.220 --> 00:21:59.760 looking to enter the tech and science startup scene? 341 00:22:00.380 --> 00:22:03.955 My advice is follow your passion and do 342 00:22:03.955 --> 00:22:07.735 not do not listen to the adults in the room 343 00:22:08.195 --> 00:22:11.880 because the adults in the room first may not have 344 00:22:11.880 --> 00:22:15.020 good ideas about the future, and secondly, 345 00:22:15.720 --> 00:22:19.425 they may not understand how innovative your 346 00:22:19.425 --> 00:22:23.025 ideas are. And so take the future in 347 00:22:23.025 --> 00:22:26.730 your hands and shape it to be 348 00:22:26.730 --> 00:22:30.330 better than the past. And that is the approach of a 349 00:22:30.330 --> 00:22:33.070 kid that, is independent. 350 00:22:33.815 --> 00:22:37.255 And, you know, that's the approach that I like. Being 351 00:22:37.255 --> 00:22:41.034 62 years old, I'm still the kid that I was 352 00:22:42.539 --> 00:22:45.440 early on in my life, I realized 353 00:22:45.980 --> 00:22:49.775 that it's much better to do the right 354 00:22:49.775 --> 00:22:53.075 thing than to listen to people in the room telling you 355 00:22:53.455 --> 00:22:56.915 what is right to do. And I very much encourage 356 00:22:57.470 --> 00:23:01.070 young people to stay young in spirit and 357 00:23:01.070 --> 00:23:04.910 follow their passion and shape the world to be better because there is 358 00:23:04.910 --> 00:23:06.450 a lot of room for improvement. 359 00:23:08.845 --> 00:23:10.385 I would just need to 360 00:23:12.765 --> 00:23:16.285 start another video because there's a limit of 20 minutes per 361 00:23:16.285 --> 00:23:19.860 recording. Okay. We just go to the last question. Mhmm. 362 00:23:20.000 --> 00:23:23.620 Looking ahead, in your opinion, what will be the most significant 363 00:23:24.635 --> 00:23:28.475 scientific breakthrough in the next decade and how can 364 00:23:28.475 --> 00:23:31.375 startups position themselves to be part of it? 365 00:23:32.155 --> 00:23:35.900 In my mind, the most important development 366 00:23:35.900 --> 00:23:39.740 in the future would be artificial intelligence taking over our 367 00:23:39.740 --> 00:23:40.240 life. 368 00:23:44.605 --> 00:23:47.745 And perhaps augmenting human intelligence. 369 00:23:49.325 --> 00:23:52.845 If we have a way of connecting the human 370 00:23:52.845 --> 00:23:56.640 brain to a computer chip. I 371 00:23:56.640 --> 00:24:00.400 call it putting a chip on our shoulder. It's already being 372 00:24:00.400 --> 00:24:04.215 done, but in the future it might 373 00:24:04.215 --> 00:24:06.715 be even more common. 374 00:24:08.455 --> 00:24:11.435 And so the question is, what will it do? And 375 00:24:12.200 --> 00:24:15.820 artificial intelligence, at some point will have enough parameters 376 00:24:16.680 --> 00:24:20.220 to, have more capacity 377 00:24:20.945 --> 00:24:23.924 than the human brain. We have a 100,000,000,000 378 00:24:25.345 --> 00:24:29.110 neurons in our brain and about 379 00:24:29.170 --> 00:24:32.870 10,000 times more synapses that connect the neurons. 380 00:24:34.050 --> 00:24:37.755 And artificial intelligence systems as of 381 00:24:37.755 --> 00:24:41.515 now have somewhat less than that, but they grow in 382 00:24:41.515 --> 00:24:45.300 the number of neurons in the artificial neural networks. 383 00:24:45.920 --> 00:24:48.660 They grow exponentially with characteristic 384 00:24:49.280 --> 00:24:52.975 doubling time of less than a couple of years. And 385 00:24:53.275 --> 00:24:57.115 so within the next 10, 20 years, there 386 00:24:57.115 --> 00:25:00.870 is a good chance that AI systems will 387 00:25:01.090 --> 00:25:04.930 have more parameters than the human brain, than the number 388 00:25:04.930 --> 00:25:08.550 of synapses in the human brain. So at that point, 389 00:25:09.395 --> 00:25:12.695 phenomena that we attribute to humans such as, 390 00:25:13.155 --> 00:25:16.835 consciousness or free will, might 391 00:25:16.835 --> 00:25:20.630 be visible when we interact with those AI systems. 392 00:25:21.090 --> 00:25:24.789 The other thing is that, if you go to graveyards, 393 00:25:25.434 --> 00:25:29.274 what you see, the tradition, is putting a 394 00:25:29.274 --> 00:25:32.990 tombstone, above the grave of a 395 00:25:32.990 --> 00:25:36.370 person. And if you think about it, that's a very primitive approach. 396 00:25:38.110 --> 00:25:41.710 Basically, a stone on which you engrave in a 397 00:25:41.710 --> 00:25:45.395 few words what the person was about. And so the 398 00:25:45.395 --> 00:25:49.235 entire lifetime of a person is summarized in a few words. And there 399 00:25:49.235 --> 00:25:52.920 are 2 problems with that. 1 is that this message is static. The 400 00:25:52.920 --> 00:25:56.760 amount of information in it is not 401 00:25:56.760 --> 00:26:00.140 changing, and also that it's very little information. 402 00:26:01.465 --> 00:26:03.485 Now granted there were a 117,000,000,000 403 00:26:04.825 --> 00:26:08.125 people on earth in the last 10000000 years 404 00:26:08.640 --> 00:26:12.420 and information was not a priority. I mean, we know that because 405 00:26:12.960 --> 00:26:16.180 some traditions burn the bodies of dead people. 406 00:26:17.345 --> 00:26:21.044 And that is basically an act of destroying the genetic information 407 00:26:21.105 --> 00:26:24.625 about the person. So, for example, the new 408 00:26:24.625 --> 00:26:28.180 horizons spacecraft that went to visit, 409 00:26:28.880 --> 00:26:32.480 Pluto included a small box with the ashes 410 00:26:32.480 --> 00:26:35.785 of Clyde Tombaugh, the person who discovered 411 00:26:36.005 --> 00:26:39.545 Pluto. And if extraterrestrials were to find this box, 412 00:26:40.005 --> 00:26:42.985 they would really wonder about the barbaric 413 00:26:43.365 --> 00:26:47.030 act of our civilization in destroying any 414 00:26:47.030 --> 00:26:50.490 information about the discoverer that it wants to commemorate. 415 00:26:51.270 --> 00:26:54.715 It there would be much more information in a hair 416 00:26:55.015 --> 00:26:58.615 taken from Tan Bao's body because it will have the DNA information 417 00:26:58.615 --> 00:27:02.395 there than in the box of ashes. So that was not an intelligent 418 00:27:02.455 --> 00:27:06.170 thing to do, to put a box full of no information about the person 419 00:27:06.170 --> 00:27:09.790 you want to commemorate. But if you think about it, 420 00:27:09.930 --> 00:27:13.315 in the future, tombstones might be replaced 421 00:27:13.775 --> 00:27:17.535 by AI avatars, meaning, there will 422 00:27:17.535 --> 00:27:21.139 be an AI system training on all the 423 00:27:21.139 --> 00:27:24.840 digital information that we leave behind. Voice, 424 00:27:25.059 --> 00:27:28.905 text, videos, And once a person 425 00:27:28.905 --> 00:27:32.585 dies, this AI system could represent the 426 00:27:32.585 --> 00:27:36.320 person and imitate the person in interaction with 427 00:27:36.320 --> 00:27:40.160 living people. So in my mind, an AI avatar 428 00:27:40.160 --> 00:27:43.905 of this type would be much better than a tombstone to 429 00:27:43.905 --> 00:27:47.665 have over my grave. And in fact, there is a CEO of a 430 00:27:47.665 --> 00:27:51.490 startup that contacted me recently and 431 00:27:51.490 --> 00:27:55.270 asked if I'm allowing his new company 432 00:27:56.050 --> 00:27:59.765 to make a replica of me. And, of 433 00:27:59.765 --> 00:28:03.465 course, there are risks in allowing that, but I agreed 434 00:28:03.525 --> 00:28:07.070 because I see it as an AI avatar that 435 00:28:07.389 --> 00:28:11.149 if I'm asked in the future to participate in podcasts like 436 00:28:11.149 --> 00:28:14.909 this one, I could send my avatar to answer all the questions. 437 00:28:14.909 --> 00:28:18.755 If it's good enough, it will do a good job, And that would save 438 00:28:18.755 --> 00:28:22.595 me time. Especially if 439 00:28:22.595 --> 00:28:26.375 I die, this avatar would give a much better representation 440 00:28:27.029 --> 00:28:30.730 of my body of work than a tombstone. 441 00:28:32.710 --> 00:28:36.415 It was such a pleasure having you here. And the next time you send 442 00:28:36.415 --> 00:28:40.195 your avatar, I send my avatar, and then make a recording together. Okay? 443 00:28:41.295 --> 00:28:44.195 Thank you very much, professor Loeb. Thanks for having me. 444 00:28:46.640 --> 00:28:50.240 Tada. We're good. Okay. Excellent. Thank you very 445 00:28:50.240 --> 00:28:51.300 much. That went