WEBVTT NOTE This file was generated by Descript 00:00:16.920 --> 00:00:21.610 So, how many of you by a show of hands remember during COVID going to a grocery 00:00:21.610 --> 00:00:23.740 store and seeing shelves like this? 00:00:24.920 --> 00:00:25.950 Well, it's not fun, right? 00:00:26.500 --> 00:00:30.209 Uh, I think a lot of the things that we relied on a daily basis were unavailable. 00:00:30.665 --> 00:00:35.535 Um, it was very difficult to just get by, uh, doing the most basic things. 00:00:35.735 --> 00:00:38.425 You know, these are consumer products, things that we use on a daily basis. 00:00:38.425 --> 00:00:41.465 But if you go, you know, the parts of the power plants needed, we're missing 00:00:41.465 --> 00:00:44.784 the parts that airplanes needed, we're missing the parts that, uh, you know, 00:00:44.785 --> 00:00:47.884 we use to run our cell phones and to build things, all of them are missing. 00:00:48.224 --> 00:00:51.425 Uh, and it was an incredibly difficult time for all of human civilization. 00:00:51.905 --> 00:00:54.665 Uh, recently, uh, if anybody was following the news, there 00:00:54.665 --> 00:00:55.675 was a dock worker's strike. 00:00:56.025 --> 00:00:59.305 And again, you know, there was this massive shock to the supply chain. 00:00:59.315 --> 00:01:00.635 Luckily it didn't last very long. 00:01:01.055 --> 00:01:05.405 Uh, but it was just a reminder to us of how dependent we are on this 00:01:05.405 --> 00:01:08.775 flow of goods and on the ability to build all the most basic things. 00:01:10.275 --> 00:01:15.004 Um, and this is all a reminder for us that the global manufacturing base, 00:01:15.034 --> 00:01:20.285 this giant machine that feeds us all of our daily needs, uh, is incredibly, 00:01:20.295 --> 00:01:25.294 incredibly, uh, fragile at any moment, you know, things could go wrong and 00:01:25.294 --> 00:01:26.644 we would be really, really suffering. 00:01:27.070 --> 00:01:31.490 Um, and, and, uh, I'm here to tell you a little bit about how AI, how robotics 00:01:31.490 --> 00:01:32.970 is going to change that landscape. 00:01:33.289 --> 00:01:34.780 What are the things that are happening right now? 00:01:34.940 --> 00:01:36.190 What are the things that we can do about it? 00:01:39.940 --> 00:01:45.849 Um, so first fact to know, uh, is today in America, American manufacturing 00:01:45.850 --> 00:01:46.780 is really, really suffering. 00:01:46.789 --> 00:01:48.439 And this is not just an American phenomenon. 00:01:48.439 --> 00:01:49.119 This is global. 00:01:49.530 --> 00:01:53.980 There's a massive demographic demographic problem that's leading factories to be 00:01:54.190 --> 00:01:57.669 extremely, extremely uncompetitive more and more of them are shutting down every 00:01:57.670 --> 00:02:02.400 single day in America, about half the factories have shut down since about 30 00:02:02.400 --> 00:02:09.070 years ago and the ones that do remain are sitting idle 65 percent of the time. 00:02:09.380 --> 00:02:14.060 So every factory out there, you know, they build giant facilities and forklifts 00:02:14.060 --> 00:02:16.790 and they have air conditioning and have trucks and they have all this stuff. 00:02:16.865 --> 00:02:21.685 Uh, and all of it ends up sitting idle two thirds of the production 00:02:21.685 --> 00:02:24.465 hours, uh, in, in, in, uh, in the day. 00:02:26.095 --> 00:02:30.785 Uh, and a lot of this is a result, a result of, uh, this, you know, really 00:02:30.795 --> 00:02:35.575 vicious, uh, cycle in the manufacturing process, which is, uh, where we are today. 00:02:35.775 --> 00:02:37.985 If you look at the stat here on the bottom, there's one 00:02:37.985 --> 00:02:42.485 and a half million unfilled manufacturing jobs today in America. 00:02:42.840 --> 00:02:44.870 Uh, if you expand that globally, it's even more. 00:02:45.070 --> 00:02:47.960 And if those factories that are sitting idle 60 percent of the time, 00:02:48.220 --> 00:02:51.090 we're actually operating at full capacity, there's another 5 million 00:02:51.589 --> 00:02:54.659 people that we need just to run the factories that we have right now. 00:02:55.080 --> 00:02:58.880 Um, and so, you know, these factories are, are, are, are incredibly 00:02:58.880 --> 00:03:01.309 desperate to have more people to help them build these products. 00:03:02.005 --> 00:03:05.015 Because they're so desperate and so shorthanded, the people that are 00:03:05.015 --> 00:03:07.924 working there actually have to pull, you know, multiple double duty. 00:03:08.105 --> 00:03:12.215 They have to do a lot of very dangerous tasks that are complex and painful. 00:03:12.454 --> 00:03:15.094 Uh, they're unhappy, and so worker turnover is very high. 00:03:15.480 --> 00:03:20.790 The typical factory in America has about 200 percent per year headcount turnover. 00:03:20.869 --> 00:03:24.489 That means for every job, they have to replace that, that position 00:03:24.500 --> 00:03:26.489 twice, uh, a year on average. 00:03:26.589 --> 00:03:28.079 There's many jobs more than that. 00:03:28.549 --> 00:03:31.390 Um, and so the end result of all of this, you know, challenge, 00:03:31.459 --> 00:03:34.540 uh, in the manufacturing process is that they end up making worse 00:03:34.549 --> 00:03:36.649 products at much higher prices. 00:03:37.090 --> 00:03:39.880 Uh, and this is really one of the core reasons for inflation. 00:03:40.200 --> 00:03:45.320 Um, where all the products that we need and use on a daily basis Just become 00:03:45.320 --> 00:03:47.299 worse and they become, uh, more expensive. 00:03:47.749 --> 00:03:50.239 Uh, and naturally the next result of that is that there's 00:03:50.239 --> 00:03:51.559 less demand for those products. 00:03:51.880 --> 00:03:54.589 And that leads to this vicious cycle where the factories are even less 00:03:54.589 --> 00:03:55.809 productive than they were before. 00:03:56.119 --> 00:03:58.109 Uh, and this cycle just gets worse and worse. 00:03:58.459 --> 00:04:00.239 Uh, so this is something that we really have to figure out 00:04:00.239 --> 00:04:01.660 ways to break out of today. 00:04:02.730 --> 00:04:05.119 Uh, so i'll talk to you a little bit about kind of where I come 00:04:05.119 --> 00:04:07.890 from and my background and why i'm so passionate about this topic. 00:04:08.329 --> 00:04:11.719 Um, you know, so I, I may not look like it, but I grew up in China and I 00:04:11.719 --> 00:04:15.984 lived in China during a time when of massive industrialization, uh, from 00:04:15.984 --> 00:04:19.355 when I was six years old until I was, uh, about twenty five, I lived there. 00:04:19.595 --> 00:04:23.734 And I got to see, uh, you know, when I, when I, I remember when I moved to China, 00:04:23.765 --> 00:04:26.794 for those of, those of you who speak Mandarin, one of the most common things 00:04:26.794 --> 00:04:30.174 you would hear when you go to a store and ask for something is they would say, mao, 00:04:30.574 --> 00:04:32.005 which means, sorry, we don't have it. 00:04:32.295 --> 00:04:35.505 Uh, and, uh, you know, this was, you know, just coming out of communism. 00:04:35.640 --> 00:04:39.980 Uh, and so, uh, you know, China had really not built a strong supply chain. 00:04:40.300 --> 00:04:45.520 And, and, and in the, in the 20 plus years that ensued, uh, we saw massive 00:04:45.520 --> 00:04:50.829 industrialization, we saw the, the, uh, capability to build all kinds of products. 00:04:51.050 --> 00:04:55.960 And that led to this, uh, advancement and, and moving up, uh, 400, 500 00:04:55.960 --> 00:04:57.080 million people out of poverty. 00:04:57.480 --> 00:05:01.150 Um, and, and so seeing that was really something that, that stuck with me. 00:05:01.400 --> 00:05:03.439 I myself, you know, love, love to tinker. 00:05:03.440 --> 00:05:04.260 I'm a builder. 00:05:04.329 --> 00:05:05.640 I love to figure out how things work. 00:05:06.400 --> 00:05:08.440 That led me to go and, uh, go to New York. 00:05:08.440 --> 00:05:11.170 I studied engineering, uh, and I learned how to actually 00:05:11.170 --> 00:05:12.140 build things with my hands. 00:05:12.460 --> 00:05:13.640 Uh, I worked on that. 00:05:13.640 --> 00:05:16.479 I built a couple of companies and then I worked in venture capital, which is 00:05:16.480 --> 00:05:19.200 this kind of place where you get to luckily live a little bit in the future. 00:05:19.300 --> 00:05:21.300 Uh, being here in Silicon Valley, I think many of you 00:05:21.560 --> 00:05:22.830 get to experience that as well. 00:05:23.059 --> 00:05:25.909 And we see all these incredible technologies on a daily basis coming 00:05:25.909 --> 00:05:29.200 out of academia, coming out of large companies, people build all 00:05:29.200 --> 00:05:30.370 this stuff that looks futuristic. 00:05:30.400 --> 00:05:33.049 And as an, as a venture capitalist, I invested in. 00:05:33.505 --> 00:05:37.205 Companies that used, you know, robots to pick apples and robots that would 00:05:37.615 --> 00:05:40.985 sort garbage and robots that would do all these incredible tasks for us. 00:05:41.325 --> 00:05:43.755 Uh, and in my mind, I was like, this is a foregone conclusion. 00:05:43.775 --> 00:05:46.355 You know, these things are going to be everywhere, uh, any day now. 00:05:46.645 --> 00:05:49.044 Uh, and seeing all that technology was, was incredibly inspiring. 00:05:49.675 --> 00:05:54.135 Uh, the reality is I was very disappointed when, when many of those companies 00:05:54.135 --> 00:05:55.915 didn't actually make it to scale. 00:05:56.485 --> 00:06:00.695 Um, and then COVID hit, which was this, you know, massive shock to the system. 00:06:00.695 --> 00:06:03.515 It was a reminder to me, and I think many of us. 00:06:03.849 --> 00:06:06.040 That all of these products that we rely on on a daily basis, 00:06:06.300 --> 00:06:07.530 uh, were very hard to come by. 00:06:07.560 --> 00:06:11.630 And if we didn't drastically adopt more robots, you know, we would be, be stuck. 00:06:11.919 --> 00:06:12.840 And so I started Formic. 00:06:12.879 --> 00:06:15.599 We've now deployed in about a hundred something, uh, factories across the U 00:06:15.599 --> 00:06:19.570 S and I've seen and heard from a lot of them about why it is that robots are so 00:06:19.570 --> 00:06:21.570 hard to get into use in the real world. 00:06:23.130 --> 00:06:27.223 Uh, so as I think I've already made clear, you know, manufacturing is not 00:06:27.223 --> 00:06:30.390 really this kind of standalone thing that just happened somewhere else. 00:06:30.420 --> 00:06:31.389 And we get to buy products. 00:06:31.689 --> 00:06:33.860 It actually is upstream from all the things that we do. 00:06:34.190 --> 00:06:38.490 Whether it's, you know, the medicines that we, that we use to stay alive, 00:06:38.490 --> 00:06:41.490 whether it's the military, whether it's agriculture and food, whether 00:06:41.490 --> 00:06:44.610 it's cars and airplanes, you know, everything relies on having a strong 00:06:44.680 --> 00:06:46.090 and powerful manufacturing base. 00:06:46.160 --> 00:06:52.100 Uh, and so, uh, you know, despite that, uh, right for every single product, even 00:06:52.100 --> 00:06:55.629 these kind of very complex and advanced products like jets and solar panels 00:06:55.629 --> 00:07:00.370 and, uh, you know, electric cars, uh, you know, we, we continue to invest in 00:07:00.610 --> 00:07:03.809 what's seen as the, you know, the sexy parts of manufacturing all of these cool 00:07:03.810 --> 00:07:07.540 and incredible complex products, but actually what under what underlies that 00:07:07.550 --> 00:07:10.400 is we have all of these very foundational. 00:07:10.685 --> 00:07:12.865 Components that are needed to build those products. 00:07:13.115 --> 00:07:16.605 You can't build a solar panel without brackets and, you know, aluminum 00:07:16.605 --> 00:07:20.035 extrusions, you can't build a jet, uh, you know, without kind of turbine 00:07:20.035 --> 00:07:23.735 blades, you can't build a car, uh, without rubber wheels, uh, and all 00:07:23.735 --> 00:07:27.274 of these things are the accumulation, uh, of what leads to us being able to 00:07:27.274 --> 00:07:32.025 build these products for an interesting stat here is for, uh, for any single 00:07:32.215 --> 00:07:34.815 automotive plant, uh, that you hear about. 00:07:35.420 --> 00:07:40.120 There's about 3, 000 suppliers that build all of the subcomponents that feed into 00:07:40.120 --> 00:07:42.200 that, uh, feed into that automotive plant. 00:07:43.650 --> 00:07:46.500 And on the other side, you know, robots are, are available, right? 00:07:46.500 --> 00:07:49.190 Like as I was talking about as an investor, I was investing in robotics. 00:07:49.599 --> 00:07:52.240 Robots actually seem to do a lot very, very well already. 00:07:52.539 --> 00:07:53.719 Uh, they're very capable. 00:07:53.930 --> 00:07:57.110 Uh, there's a lot of tasks that they can already do very reliably. 00:07:59.370 --> 00:08:02.560 But, uh, something that maybe many people may not know is that robot 00:08:02.560 --> 00:08:04.150 adoption is, is abysmal right now. 00:08:04.604 --> 00:08:08.715 Uh, in America, about 90 percent of factories don't have a single robot, 00:08:09.054 --> 00:08:10.794 uh, anywhere to be seen on the floor. 00:08:11.114 --> 00:08:15.254 Uh, and this was a, a stat that, that is I think very, very shocking for most people. 00:08:15.465 --> 00:08:18.255 Um, and it's really only in this kind of the green triangle that you 00:08:18.255 --> 00:08:21.234 remember from my earlier slide, these kind of very advanced plants, you 00:08:21.234 --> 00:08:22.715 know, they're really well automated. 00:08:22.724 --> 00:08:26.354 People who make chips and people who make, uh, cars, uh, are two examples, 00:08:26.655 --> 00:08:29.344 but all of their suppliers, uh, are not. 00:08:30.635 --> 00:08:31.565 Um, and. 00:08:31.915 --> 00:08:35.215 And the real reason for this is that for most manufacturing facilities, 00:08:35.515 --> 00:08:39.575 adopting robotics and automation is kind of out of this world complicated. 00:08:39.705 --> 00:08:45.620 I was in a factory a couple of weeks ago where This is a plant that used 00:08:45.620 --> 00:08:47.360 to make parts for the Ford Model T. 00:08:47.780 --> 00:08:53.320 Uh, it was started, uh, you know, in the early 1900s by the great grandfather 00:08:53.320 --> 00:08:54.460 of the person who's running it now. 00:08:54.700 --> 00:08:55.610 It's a family owned business. 00:08:55.610 --> 00:08:57.640 It's been passed down one generation after another. 00:08:57.849 --> 00:08:59.310 It's still run by the same guys. 00:08:59.569 --> 00:09:02.650 And, uh, if you go and look around in the plant, it looks 00:09:02.650 --> 00:09:04.170 like it's still in the 1900s. 00:09:04.240 --> 00:09:05.880 You know, all the equipment is still from then. 00:09:06.150 --> 00:09:09.060 Uh, and if you're, you know, spent all of your time thinking about how to build 00:09:09.060 --> 00:09:12.224 this specific metal part, Or if you spend all of your time making chocolate chip 00:09:12.224 --> 00:09:16.074 cookies, uh, or you spend all of your time making, uh, you know, dog food, 00:09:16.074 --> 00:09:19.885 which are all, are all factories that we put robots into and you ask them, 00:09:20.125 --> 00:09:22.035 uh, about their journey with robotics. 00:09:22.244 --> 00:09:24.204 What they say is the same every single time. 00:09:24.214 --> 00:09:24.994 They say three things. 00:09:25.015 --> 00:09:28.875 Number one, I know I'm going to go out of business if I don't get robots. 00:09:29.265 --> 00:09:34.145 Um, number two is I've been trying to get robots for the last 10 years. 00:09:34.505 --> 00:09:36.324 Uh, I've been trying to figure out ways to automate. 00:09:36.685 --> 00:09:38.655 Uh, and number three is I still don't have any robots. 00:09:39.385 --> 00:09:42.485 Uh, and so it's really, really challenging and difficult for these 00:09:42.485 --> 00:09:43.825 manufacturing facilities to automate. 00:09:43.885 --> 00:09:46.314 And one of the reasons is that, you know, this, what you see here at the 00:09:46.314 --> 00:09:51.185 bottom, uh, for every 11 factories in America, there's only one, uh, 00:09:51.185 --> 00:09:52.645 you know, robot programmer available. 00:09:52.925 --> 00:09:56.035 Uh, and honestly, you know, to, to automate a plant, you need, you know, 00:09:56.165 --> 00:10:00.355 probably 50 or 100, uh, robot programmers to be able to, uh, to get these jobs done. 00:10:00.495 --> 00:10:06.235 Um, but you know, the good news is, you know, we're at a very important moment 00:10:06.255 --> 00:10:11.244 in a turning point in this, in this story where AI is making robots much 00:10:11.245 --> 00:10:15.324 more intelligent, much more capable, uh, but also all the subcomponents that you 00:10:15.325 --> 00:10:18.165 need in order to make a robot work, you need to evaluate the site and you need 00:10:18.165 --> 00:10:21.075 to do 3D scanning, you need to generate code, you need to figure out all the 00:10:21.075 --> 00:10:24.565 parts, you need to be able to reconfigure that robot, all of those are becoming 00:10:24.565 --> 00:10:26.625 faster and easier, uh, by the day. 00:10:26.670 --> 00:10:32.140 Um, and so, you know, this is, I think, an example where, uh, this is 00:10:32.140 --> 00:10:35.339 a phrase you hear a lot in Silicon Valley, where the future is here, uh, 00:10:35.370 --> 00:10:36.769 but it's not evenly distributed yet. 00:10:37.000 --> 00:10:39.819 For technology to really get into the hands of the people who need it the 00:10:39.819 --> 00:10:43.329 most, there's a huge divide that needs to be crossed, and there's this bridge. 00:10:45.550 --> 00:10:48.430 Uh, and so, you know, I started thinking about, you know, where, where has this 00:10:48.439 --> 00:10:50.550 bridge been, been, been crossed before? 00:10:50.810 --> 00:10:52.780 And there's a lot of industries where this has happened. 00:10:53.100 --> 00:10:55.740 Uh, and I think this is something for those of us here in Silicon Valley who 00:10:55.780 --> 00:10:57.659 think about technology on a regular basis. 00:10:58.125 --> 00:11:01.515 Uh, you know, just inventing the technology is really not nearly enough. 00:11:01.825 --> 00:11:04.665 Uh, we really have to figure out how it gets adoption in the real world. 00:11:05.095 --> 00:11:08.824 Um, so in the case of, of email is, is one of my favorite examples. 00:11:09.085 --> 00:11:12.185 You know, for those who may remember, there was a time where if you wanted to 00:11:12.194 --> 00:11:16.550 have an email address, Your only option was to, you know, have a giant air 00:11:16.550 --> 00:11:19.810 conditioned room in your office and have a bunch of, you know, large server racks 00:11:19.810 --> 00:11:23.650 and have a bunch of people who manage the IT of it and pay hundreds of thousands of 00:11:23.650 --> 00:11:27.010 dollars of software licensing fees before you could actually have an email address. 00:11:27.859 --> 00:11:30.039 And as a result, most people just didn't do it. 00:11:30.449 --> 00:11:32.520 For the vast majority of people, we just didn't have email. 00:11:32.890 --> 00:11:38.280 And it wasn't until this kind of generation of operators came around 00:11:38.280 --> 00:11:40.750 that said, we're going to create hosted services for you, we're going 00:11:40.750 --> 00:11:42.519 to manage it on your behalf, and you're going to be able to access it. 00:11:42.810 --> 00:11:46.110 access these technologies, um, that we saw massive adoption. 00:11:46.240 --> 00:11:49.550 And the same thing has happened in industry after industry, uh, from solar 00:11:49.550 --> 00:11:51.970 panel adoption to aircraft adoption. 00:11:51.970 --> 00:11:55.630 And if you look at the history of technology, uh, there's always this gap 00:11:55.640 --> 00:11:58.689 where, you know, something's invented and it takes a long time for people to figure 00:11:58.689 --> 00:12:00.199 out how to actually use it at scale. 00:12:00.430 --> 00:12:02.439 Uh, and then you see a massive adoption. 00:12:03.340 --> 00:12:05.540 And so I think robots are, are at that turning point. 00:12:05.954 --> 00:12:09.805 Uh, we're starting to see, uh, government and private enterprise come in and 00:12:09.805 --> 00:12:13.435 start to take on the role of centrally managing the operations of robotics 00:12:13.435 --> 00:12:15.344 and helping manufacturers to automate. 00:12:15.604 --> 00:12:18.864 We work with, uh, you know, a lot of state governments, and we see that, uh, 00:12:18.904 --> 00:12:22.654 you know, there are massive programs in place to help, uh, with the adoption of 00:12:22.654 --> 00:12:24.704 this incredibly, uh, critical technology. 00:12:25.664 --> 00:12:28.564 Uh, and why is this, you know, important to all of us who are sitting here? 00:12:28.564 --> 00:12:30.579 You know, the hope is that this really leads us to a path forward. 00:12:30.939 --> 00:12:34.099 Where we have a world of abundance, uh, we want to be able to have, you 00:12:34.099 --> 00:12:36.780 know, all of the things that we need on a, need to rely on on a daily 00:12:36.780 --> 00:12:41.099 basis available to, uh, you know, our, our children and our grandchildren. 00:12:41.420 --> 00:12:44.609 Uh, and, uh, in order to do that, we really need to remove a lot of the 00:12:44.609 --> 00:12:46.339 stigmas around adopting technology. 00:12:46.599 --> 00:12:49.619 We need to create pathways, uh, for those who need this technology 00:12:49.619 --> 00:12:50.589 to be able to figure out how to. 00:12:52.250 --> 00:12:52.589 Thank you.