WEBVTT - Introducing Substrate—An Open-source Framework for Human Understanding, Meaning, and Progress 0:00:21.333 --> 0:00:24.093 All right. Welcome to unsupervised learning. This is Daniel Miessler. 0:00:24.093 --> 0:00:27.633 And today I'm super excited to announce a project I've 0:00:27.633 --> 0:00:30.033 been wanting to talk about for a very long time 0:00:30.033 --> 0:00:35.703 called substrate. Okay, let's get into the project itself. So 0:00:35.703 --> 0:00:39.573 what is it exactly? That is really the question. What 0:00:39.573 --> 0:00:43.863 substrate is is an open source framework for human understanding, 0:00:43.863 --> 0:00:47.613 meaning and progress. And you might be inclined to say, 0:00:47.613 --> 0:00:51.463 what the hell does that mean? And it's a great question. Right? 0:00:51.463 --> 0:00:54.223 So the purpose of the project is to make things 0:00:54.223 --> 0:01:01.123 that matter to humans more transparent. Discussable. And ultimately, because 0:01:01.123 --> 0:01:06.043 they're transparent and discussable, they'll be more fixable. So what 0:01:06.043 --> 0:01:08.833 kind of things are we talking about? So we're calling 0:01:08.833 --> 0:01:13.243 these substrate components. And these are the components of human meaning. Right. 0:01:13.243 --> 0:01:16.363 When we talk about understanding, meaning and progress, these are 0:01:16.363 --> 0:01:20.923 the pieces that we're actually talking about. So collections of 0:01:20.923 --> 0:01:24.193 things okay. So the first thing is an idea collections 0:01:24.193 --> 0:01:28.153 of ideas a list of human novel ideas, problems a 0:01:28.153 --> 0:01:32.263 list of our most important human problems, our beliefs, our models, 0:01:32.263 --> 0:01:36.493 which are our ways of conceptualizing reality, frames a list 0:01:36.493 --> 0:01:40.133 of narratives or lenses for perceiving reality. A list of 0:01:40.133 --> 0:01:45.023 solutions that correspond to problems. Information sources. So you'll have, 0:01:45.023 --> 0:01:50.333 like New York Times, The Hill, Breitbart, uh, lots of 0:01:50.333 --> 0:01:57.683 different sources of information, government sources, individual sources, different media organizations. 0:01:57.683 --> 0:02:00.353 It's got to be a comprehensive list, and it's got 0:02:00.353 --> 0:02:03.803 to span different political ideologies. That's really important here as 0:02:03.803 --> 0:02:08.123 we'll see later people. So this is just individuals, organizations 0:02:08.123 --> 0:02:13.913 of different types laws. We're going to collect all different legislation, 0:02:14.003 --> 0:02:17.513 starting with the US government, but expanding out to pretty 0:02:17.513 --> 0:02:20.393 much the world at some point claims. So this is 0:02:20.393 --> 0:02:23.813 a factual claim, a truth claim about the world votes, 0:02:23.843 --> 0:02:28.263 a list of votes and results from laws that were 0:02:28.263 --> 0:02:33.333 submitted and voted on to basically say, here's what the 0:02:33.333 --> 0:02:36.423 votes were from different people. So this is talking about 0:02:36.423 --> 0:02:40.863 representatives voting on things, right? Arguments a list of arguments 0:02:40.863 --> 0:02:44.943 that have been made in favor or against a particular thing. 0:02:44.943 --> 0:02:52.613 Funding sources, lobbyists. So a list of lobbyists in their agendas, missions, donations, 0:02:52.613 --> 0:02:57.743 goals and facts. So these are actually claims from up here, 0:02:57.743 --> 0:03:01.133 but these are just verified ones. And important to note 0:03:01.133 --> 0:03:04.283 that it could become true and then untrue again. So 0:03:04.283 --> 0:03:07.133 you have to kind of keep this updated. Now each 0:03:07.133 --> 0:03:10.583 of these in this list here, each of these will 0:03:10.583 --> 0:03:14.893 be an actual list maintained in a repository within GitHub. 0:03:14.923 --> 0:03:17.833 In fact, we already have one of these okay. So 0:03:17.833 --> 0:03:19.513 I'm going to click in here. And if we go 0:03:19.513 --> 0:03:22.663 into here and look at problems we already have a 0:03:22.663 --> 0:03:25.873 list of problems here okay. It's already starting. So we've 0:03:25.873 --> 0:03:29.053 got we've got this going and I'm just going to 0:03:29.053 --> 0:03:31.513 zoom in a whole bunch for this one. So we've 0:03:31.513 --> 0:03:36.013 got like ransomware attacks on US health care systems. Uh, 0:03:36.013 --> 0:03:40.643 teen depression in the UK, nuclear weapons development in North Korea. 0:03:40.673 --> 0:03:43.733 Each one of them has this PR code so you 0:03:43.733 --> 0:03:47.813 can associate directly to it. And that's basically the structure. 0:03:47.813 --> 0:03:50.183 And you can see the structure here as well. And 0:03:50.183 --> 0:03:52.433 again we have the link here to the GitHub repo 0:03:52.433 --> 0:03:54.593 that we just looked at. And they're all part of 0:03:54.593 --> 0:03:59.123 this substrate organization which is here. So if you go 0:03:59.123 --> 0:04:02.333 into the substrate organization you could see the the main 0:04:02.333 --> 0:04:05.933 flow here. And if you go into repositories you see 0:04:05.933 --> 0:04:08.963 that we've populated a whole bunch of these, and we've 0:04:08.963 --> 0:04:11.213 already got some pretty good activity on some of these 0:04:11.213 --> 0:04:14.753 especially problems. And keep in mind, this just came out 0:04:14.753 --> 0:04:17.633 like two days ago. So this is brand new okay. 0:04:17.633 --> 0:04:21.023 So let's keep going on with the explanation here. So 0:04:21.023 --> 0:04:24.593 the structure will allow the entire open source community which 0:04:24.593 --> 0:04:31.113 is basically the world to contribute their own problems, claims, sources, frames, goals, etc. 0:04:31.113 --> 0:04:35.163 for all those different repositories which are all the different 0:04:35.163 --> 0:04:39.033 substrate components. Okay, I think I'm starting to get it, 0:04:39.033 --> 0:04:42.363 but I need more. Fair enough. So one way to 0:04:42.363 --> 0:04:45.123 think about this is as a way to put handles 0:04:45.123 --> 0:04:49.293 onto things that are hard to discuss, right? The whole 0:04:49.293 --> 0:04:52.803 reason that I created this project is so that we 0:04:52.803 --> 0:04:58.053 can articulate things better to each other, but also to ourselves. Okay, 0:04:58.053 --> 0:05:00.693 so let's get into some actual examples here so you 0:05:00.693 --> 0:05:02.973 can see what I'm talking about. Let's start with an 0:05:02.973 --> 0:05:06.303 argument component okay. This is what this is going to 0:05:06.303 --> 0:05:09.393 look like. Think of a common argument we might hear 0:05:09.393 --> 0:05:12.303 on any given day about whatever topic. And in this 0:05:12.303 --> 0:05:14.673 case it's going to be recycling. So this is just 0:05:14.673 --> 0:05:18.073 somebody that you hear say something right. I don't know 0:05:18.073 --> 0:05:20.593 why you recycle, man. It's a total waste. It costs 0:05:20.593 --> 0:05:23.233 so much to recycle right now. And the programs are 0:05:23.233 --> 0:05:26.863 poorly run. So it's not actually benefiting the environment like 0:05:26.863 --> 0:05:29.833 I do it if it worked, but it doesn't. And 0:05:29.833 --> 0:05:32.773 this is some person watching you put a can in 0:05:32.773 --> 0:05:35.683 a recycling bin. So we're confronted by this type of 0:05:35.683 --> 0:05:39.803 argument constantly. We're being pitched these different things like oh I, 0:05:39.803 --> 0:05:42.983 I would definitely do this because of this boom, that's 0:05:42.983 --> 0:05:45.863 an argument. Or I would definitely stop doing that because 0:05:45.863 --> 0:05:50.033 of this. That's also an argument. So that's things like recycling, 0:05:50.033 --> 0:05:52.613 but it's also things that matter much more about politics 0:05:52.613 --> 0:05:55.793 or whatever. So what substrate does this is the most 0:05:55.793 --> 0:05:59.093 important thing, right. It takes an argument like this. 0:05:59.093 --> 0:05:59.633 And. 0:05:59.633 --> 0:06:02.843 Turns it into something like this. And I'm going to 0:06:02.843 --> 0:06:05.213 zoom in. It's probably going to be a little bit blurry, 0:06:05.213 --> 0:06:10.853 but watch this. This argument consists of these claims, okay, 0:06:10.853 --> 0:06:14.633 effort required cost of progress. So it breaks down the 0:06:14.633 --> 0:06:19.823 argument into these different components. Limited results okay. Then it 0:06:19.823 --> 0:06:25.013 breaks those into further things. Recycling programs require significant effort. 0:06:25.013 --> 0:06:30.783 Recycling programs are expensive. Most rejected materials end up in landfills. Okay. 0:06:30.783 --> 0:06:34.683 And then it goes and adds research to all of those. 0:06:34.683 --> 0:06:37.083 In this case, it's just pulled it directly out of 0:06:37.083 --> 0:06:40.053 the model. But you can actually add AI on top 0:06:40.053 --> 0:06:42.033 of this, which we're going to be doing with an 0:06:42.033 --> 0:06:45.453 agent framework, either that we build ourselves or that we 0:06:45.453 --> 0:06:48.393 use one of the third parties. And these will actually 0:06:48.393 --> 0:06:51.723 be tools doing this research down here. Okay. Going and 0:06:51.723 --> 0:06:55.413 doing separate lookups to find research. Like you could do 0:06:55.413 --> 0:07:00.003 this at really in-depth level for each stage here, especially 0:07:00.003 --> 0:07:03.033 this one where you're doing the research and then it 0:07:03.033 --> 0:07:06.303 comes back and says supports or supports or weakens. Look 0:07:06.303 --> 0:07:12.033 at this. Supports supports weakens weakens partially supports. Then we 0:07:12.033 --> 0:07:16.163 have a conclusion ends up being a weak argument. Interesting. 0:07:16.163 --> 0:07:18.773 So you could see the reasons this is the most 0:07:18.773 --> 0:07:22.493 important thing. You could see how this argument was formed, 0:07:22.493 --> 0:07:28.013 how it was researched using what sources. Okay, look at this. Uh, 0:07:28.013 --> 0:07:34.463 recycling statistics 2021 National Waste and Recycling Association 2022. So 0:07:34.463 --> 0:07:36.773 we actually have source names here, and we can go 0:07:36.773 --> 0:07:40.383 into even more depth if we do additional AI work 0:07:40.383 --> 0:07:43.953 on this. And that's kind of a foreshadowing for future 0:07:43.953 --> 0:07:46.803 section here. So let's just keep going. Again, what this 0:07:46.803 --> 0:07:50.523 does is it takes an argument like this recycling example, 0:07:50.523 --> 0:07:54.423 and it turns it into a graph like we just saw. Okay, 0:07:54.423 --> 0:07:57.183 the most important thing about this graph is that we 0:07:57.183 --> 0:07:59.283 could throw it up on a board. We could throw 0:07:59.283 --> 0:08:01.083 it on the side of a wall. We could each 0:08:01.083 --> 0:08:04.103 view it inside of our AR glasses or whatever in 0:08:04.103 --> 0:08:06.893 the in the near future. And now both of us 0:08:06.893 --> 0:08:09.683 who are having this discussion can be looking at the 0:08:09.683 --> 0:08:14.693 same thing and saying, uh, yeah, I agree with these two, 0:08:14.693 --> 0:08:17.453 but I don't agree with this one. And the reason 0:08:17.453 --> 0:08:19.673 I don't agree is because I don't trust the source 0:08:19.673 --> 0:08:22.373 that it came from. And then you could do things like, okay, 0:08:22.373 --> 0:08:25.223 let's grab some content from other sources. Boom. You start 0:08:25.223 --> 0:08:28.083 adding other sources to it and maybe it changes based 0:08:28.083 --> 0:08:31.353 on those sources. It changes the conclusion of that subpart, 0:08:31.353 --> 0:08:34.173 of that argument. And now the whole thing updates. And 0:08:34.173 --> 0:08:37.563 the conclusion is maybe, oh, we're not sure. Or maybe 0:08:37.563 --> 0:08:40.023 it switches it from yes to no or up to 0:08:40.023 --> 0:08:44.103 down or whatever. The point is transparency. Okay. Too many 0:08:44.103 --> 0:08:48.753 of our conversations about any topic, it becomes emotional because 0:08:48.753 --> 0:08:51.633 it's too difficult to keep all these things in our 0:08:51.633 --> 0:08:54.783 minds at once. Okay, a basic argument like it was 0:08:54.783 --> 0:08:58.383 just made here about recycling has multiple sub claims in it, 0:08:58.383 --> 0:09:01.863 and those individual sub claims need to be backed by data. 0:09:01.863 --> 0:09:04.443 And it's very hard to do that just in our 0:09:04.443 --> 0:09:07.443 brains when we're in the middle of a conversation, especially 0:09:07.443 --> 0:09:09.873 if you're trying to like blast this out. You want 0:09:09.873 --> 0:09:12.903 to have this conversation, you want to say, hey, recycling 0:09:12.903 --> 0:09:16.463 is good or recycling is bad, or whatever the topic is, 0:09:16.463 --> 0:09:19.103 but you try to put that on social media or somewhere. 0:09:19.103 --> 0:09:22.823 You're basically writing a text thing, right? Which is trying 0:09:22.823 --> 0:09:25.343 to convince someone, which is fine. We've been doing that forever. 0:09:25.343 --> 0:09:28.853 But imagine where we could do this graphically, where we 0:09:28.853 --> 0:09:32.573 could do this visually and show the connections of how 0:09:32.573 --> 0:09:35.933 strong an individual claim is based on how much you 0:09:35.933 --> 0:09:39.893 trust an individual source. Right. So that's the power of 0:09:39.893 --> 0:09:41.833 this thing. And this is why I'm so excited about it. 0:09:41.833 --> 0:09:44.653 And there's many, many more examples we're about to get into. 0:09:44.653 --> 0:09:48.583 So each of those objects in that diagram is another 0:09:48.583 --> 0:09:52.483 substrate component, right. The claims, the sources, they're all in 0:09:52.513 --> 0:09:56.593 there and they're all in there inside of another repository 0:09:56.593 --> 0:09:59.623 inside of the substrate project. So here's an example of 0:09:59.623 --> 0:10:04.273 a source. New York Times, Associated Press, Breitbart. When people 0:10:04.283 --> 0:10:07.013 make truth claims, it's important to be able to fact 0:10:07.013 --> 0:10:10.043 check or research those claims to see their support and substrate. 0:10:10.043 --> 0:10:13.433 Does this by maintaining a list of those sources that 0:10:13.433 --> 0:10:16.373 we may or may not trust. That's the point. Some 0:10:16.373 --> 0:10:18.893 people will trust different sources. Some people will not trust 0:10:18.893 --> 0:10:22.523 different sources. Right. And that's up to them. Then it 0:10:22.523 --> 0:10:25.043 just becomes a question of, okay, well, you should trust 0:10:25.043 --> 0:10:28.313 this source and that becomes an argument by itself. Right 0:10:28.323 --> 0:10:30.843 Now you can make an argument. Well, you're putting too 0:10:30.843 --> 0:10:34.413 much weight onto a particular source or set of sources, 0:10:34.413 --> 0:10:37.233 which let's look at a bunch of claims that they've 0:10:37.233 --> 0:10:41.193 made recently and let's separately research those. And maybe that 0:10:41.193 --> 0:10:44.433 argument is strong enough to make them discount that particular source, 0:10:44.433 --> 0:10:47.133 whether that's The New York Times or Breitbart or whatever 0:10:47.133 --> 0:10:49.893 it is. Right. So when somebody makes an argument or 0:10:49.893 --> 0:10:53.463 a claim within an argument, it can be linked directly 0:10:53.463 --> 0:10:56.073 to those sources that you do or do not trust. 0:10:56.073 --> 0:10:58.683 And you can see the full argument and all of 0:10:58.683 --> 0:11:01.923 its support. In one visual look at this claim. Inflation 0:11:01.923 --> 0:11:06.003 fell by 2% under this particular person's term, right? This 0:11:06.003 --> 0:11:11.073 argument includes this claim. This claim has this source, the AP, 0:11:11.103 --> 0:11:14.943 this claim has this source. New York Times. So this 0:11:14.943 --> 0:11:18.133 argument has a particular strength as a result of that. 0:11:18.133 --> 0:11:22.783 So argument to claims to sources. And this is why 0:11:22.783 --> 0:11:25.633 we're so excited about substrate. It's going to make things 0:11:25.633 --> 0:11:29.803 that used to be almost impossible to discuss actually approachable 0:11:29.803 --> 0:11:34.333 actually discussable, because we could break it down into its 0:11:34.333 --> 0:11:38.353 individual pieces. So before you'd be like, you're just not 0:11:38.353 --> 0:11:41.413 able to counter all my arguments and evidence because I'm 0:11:41.413 --> 0:11:44.533 too smart. And now you could say, look, here's my argument. 0:11:44.533 --> 0:11:46.993 Throw it up on the board. Maybe it's like a 0:11:46.993 --> 0:11:49.723 little bit in the future or, you know, Apple Vision 0:11:49.723 --> 0:11:52.453 Pro or whatever. And now it's actually a 3D thing. 0:11:52.453 --> 0:11:54.343 You could turn it around like a piece of DNA 0:11:54.343 --> 0:11:57.733 or something, right? Show me which claim you disagree with 0:11:57.733 --> 0:12:01.783 or which source you disagree with that backs up those claims. Right. 0:12:01.783 --> 0:12:04.543 So now we can just take this thing apart using 0:12:04.543 --> 0:12:07.503 different sources or whatever and change the output. This will 0:12:07.503 --> 0:12:11.793 enable far more logical and precise discussions that that is 0:12:11.793 --> 0:12:14.433 what we're going for. This is just one example. This 0:12:14.433 --> 0:12:17.193 is just the arguments example. So now I want to 0:12:17.193 --> 0:12:19.443 take and like back up a little bit and think 0:12:19.443 --> 0:12:24.213 about real world use cases for substrate. Overall okay. Sounds 0:12:24.213 --> 0:12:27.723 really cool. What do you actually do with it. Exactly. 0:12:27.723 --> 0:12:30.343 So this is my favorite. Keep in mind, this is 0:12:30.343 --> 0:12:33.163 very early. It literally came out like two days ago. 0:12:33.163 --> 0:12:36.733 But we've already got multiple use cases for this thing okay. 0:12:36.733 --> 0:12:39.193 So let's get into these okay. The first one here 0:12:39.193 --> 0:12:43.723 is describing yourself. This one is massive. Look at this. 0:12:43.723 --> 0:12:46.633 This is pretty cool art. This uh, from Midjourney. It's, uh, 0:12:46.633 --> 0:12:49.453 part of a design that I did for a previous piece. 0:12:49.453 --> 0:12:53.083 But anyway. Yeah. So you can see these are meant 0:12:53.083 --> 0:12:56.913 to represent the different connections of like, uh, goals to 0:12:56.913 --> 0:13:00.033 mission to problems that she thinks are most important in 0:13:00.033 --> 0:13:04.893 the world. Solutions. She thinks, uh, are the answers to 0:13:04.893 --> 0:13:09.333 those problems. Different strategies, different projects she's working on. That's 0:13:09.333 --> 0:13:12.543 what this, uh, AI art is meant to represent with 0:13:12.543 --> 0:13:15.153 this graph. Little function here. But then you can have 0:13:15.153 --> 0:13:17.713 labels assigned to that. And keep in mind this is 0:13:17.713 --> 0:13:20.353 me looking at her in this coffee shop through my 0:13:20.353 --> 0:13:23.803 AR glasses and seeing, okay, first of all, purple is 0:13:23.803 --> 0:13:27.583 associated with engineering. Boom. She's an engineer. She's known for 0:13:27.583 --> 0:13:29.833 being a friend. She's known for being a writer. So 0:13:29.833 --> 0:13:34.003 I'm now seeing this person as she is, which she 0:13:34.003 --> 0:13:37.063 has previously defined. So let's get into that. Many people 0:13:37.063 --> 0:13:41.413 have trouble describing who they are and what they are about. 0:13:41.443 --> 0:13:46.123 This is so critical with substrate, you can basically describe 0:13:46.123 --> 0:13:50.713 yourself in any way you want to write text, audio, video, whatever, 0:13:50.713 --> 0:13:53.053 even have a conversation with AI, which is not part 0:13:53.053 --> 0:13:56.203 of substrate yet. But uh, there's lots of different ways 0:13:56.203 --> 0:13:58.783 to do that, and it will be able to both 0:13:58.783 --> 0:14:03.823 articulate and visualize you as a person. This is absolutely insane. 0:14:03.823 --> 0:14:06.533 People have trouble describing who they are and what they're about. 0:14:06.533 --> 0:14:08.603 This is going to help you do that. And if 0:14:08.603 --> 0:14:12.593 you share your context or your substrate representations with others, 0:14:12.593 --> 0:14:14.483 they'll be able to see what you're about. They'll be 0:14:14.483 --> 0:14:17.753 able to see this. Okay. Now, of course, the visuals here, 0:14:17.753 --> 0:14:19.493 the AR stuff, I mean that's going to take some 0:14:19.493 --> 0:14:23.483 time that that's a separate project, right? That's just visualization 0:14:23.483 --> 0:14:26.603 of data that's already there. Substrate is about creating that 0:14:26.603 --> 0:14:31.433 data for you for organizations for whatever in these individual components. 0:14:31.433 --> 0:14:35.033 So learning a person's values substrate will be a wonderful 0:14:35.033 --> 0:14:38.663 way to start learning about someone or something, what they 0:14:38.663 --> 0:14:41.423 care about, how they see the world, whatever. So imagine 0:14:41.423 --> 0:14:44.663 having something like this available when you're looking at someone 0:14:44.663 --> 0:14:47.513 or you're researching them. Okay, check this out. We got 0:14:47.513 --> 0:14:50.933 a person here. Okay. Got this guy. It's like the 0:14:50.933 --> 0:14:55.323 same coffee shop programmer, gamer, organizer. And you've got this 0:14:55.323 --> 0:14:58.053 grid here, right? Or this graph here. So check this out. 0:14:58.053 --> 0:15:03.813 Mission merged gaming and programming, continuous learning, community building innovation 0:15:03.813 --> 0:15:09.663 and game dev values. Goals, projects, annual Pixel Jam AI, 0:15:09.663 --> 0:15:12.843 MPC framework. These are things that he's working on to 0:15:12.843 --> 0:15:16.923 further these particular projects. And these are the goals that 0:15:16.923 --> 0:15:20.803 he's shooting for, for the mission of merging gaming and programming. 0:15:20.803 --> 0:15:24.103 That's insane. So now I know about this person, and 0:15:24.103 --> 0:15:26.773 if there's matches, then I could walk up to him 0:15:26.773 --> 0:15:29.143 and be like, hey, I hear you're into this, or 0:15:29.143 --> 0:15:31.033 I hear here you're into that. And now I have 0:15:31.033 --> 0:15:33.943 a conversation to start. Or maybe my eye is having 0:15:33.943 --> 0:15:36.733 that conversation with his eye. Whatever. It depends how far 0:15:36.733 --> 0:15:38.683 in the future it is. But this will be a 0:15:38.683 --> 0:15:42.143 wonderful way to learn about what somebody really cares about 0:15:42.143 --> 0:15:44.843 and how they see the world. So check this out. 0:15:44.843 --> 0:15:48.713 They believe the most important problems are these three problems. 0:15:48.713 --> 0:15:51.653 This is really cool, and they believe the best solutions 0:15:51.653 --> 0:15:55.883 are these solutions to those problems. And they intend to 0:15:55.883 --> 0:15:59.933 track progress using the following KPIs. Boom, boom boom. Here 0:15:59.933 --> 0:16:02.933 are the KPIs. So imagine if you match up with 0:16:02.933 --> 0:16:06.833 somebody across all these different axes right? You match with 0:16:06.833 --> 0:16:10.673 them on values, goals, beliefs, preferences. You could find friends, 0:16:10.673 --> 0:16:13.493 you can find people to have conversations with. You can 0:16:13.493 --> 0:16:16.013 find a mate, you could find a partner. This way 0:16:16.013 --> 0:16:18.863 you could find business partners this way. So we're very 0:16:18.863 --> 0:16:21.983 excited about the potential to spawn more human connection in 0:16:21.983 --> 0:16:26.333 this way. Okay, so the next one here is visualizing arguments. 0:16:26.333 --> 0:16:29.273 So when you have a given narrative or rumor or 0:16:29.273 --> 0:16:32.723 conspiracy theory going viral, you'll be able to use substrate 0:16:32.723 --> 0:16:35.783 to analyze the argument or claim and publish the results. 0:16:35.993 --> 0:16:37.973 We already talked about this one. We already showed it. 0:16:37.973 --> 0:16:40.523 So I'm just going to show another argument here and 0:16:40.523 --> 0:16:42.923 zoom in. Pretty heavy. Look at this. We never went 0:16:42.923 --> 0:16:45.353 to the moon. Look at this. Look at this. We 0:16:45.353 --> 0:16:48.863 never went to the moon. Contradicted by contradicted by leads 0:16:48.863 --> 0:16:52.883 to claim moon landings were faked. Look at this claim. 0:16:52.883 --> 0:16:55.353 We never went to the moon. It leads to moon. 0:16:55.353 --> 0:17:00.093 Landings were faked by NASA. Okay. And we have actions 0:17:00.093 --> 0:17:03.963 which is contradicted by leads to. And then we have 0:17:03.963 --> 0:17:06.693 another claim which is the red one. And then look 0:17:06.693 --> 0:17:11.133 at this provides attempts to explain demonstrate. So again the 0:17:11.133 --> 0:17:14.163 actions or the verbs. And then we have results. So 0:17:14.163 --> 0:17:19.093 reflectors on the moon consistently reflect lasers back from the Earth. 0:17:19.123 --> 0:17:22.603 I didn't think of this one. I actually dynamically created 0:17:22.603 --> 0:17:25.213 this one just just for the launch of the project. 0:17:25.213 --> 0:17:28.483 And that was a great one. Okay, what is bouncing 0:17:28.483 --> 0:17:31.693 lasers off the moon if there aren't reflectors on the moon, 0:17:31.693 --> 0:17:34.483 which means somebody put them there. So I guess maybe 0:17:34.483 --> 0:17:37.213 that would take you into aliens, but that's a separate topic. Okay. 0:17:37.213 --> 0:17:41.833 Multiple countries have independently verified moon landings. Over £800 of 0:17:41.833 --> 0:17:45.163 moon rocks have been studied worldwide, so you've got a 0:17:45.163 --> 0:17:48.223 collection of evidence that starts to add up. Conclusion. False. 0:17:48.253 --> 0:17:52.753 Overwhelming evidence supports moon landings, and this is just one 0:17:52.753 --> 0:17:55.303 level deep. You can go even deeper with sub claims 0:17:55.303 --> 0:17:58.153 and sub claims and pointing to the various sources with 0:17:58.153 --> 0:18:00.763 actual citations, like you can go as far as you 0:18:00.763 --> 0:18:03.523 want with this and have it all stored right there 0:18:03.523 --> 0:18:06.473 inside of substrate. So you'll be able to see, for example, 0:18:06.473 --> 0:18:10.913 that they're making the following arguments, which include the following claims, 0:18:10.913 --> 0:18:14.513 which we fact checked using the following sources which resulted 0:18:14.513 --> 0:18:19.253 in the following results which, using the following methodology, leads 0:18:19.253 --> 0:18:22.073 us to this conclusion. So I don't know if you 0:18:22.073 --> 0:18:25.973 remember Snopes, but Snopes was a basically a rumor denier 0:18:25.973 --> 0:18:29.393 or approver. It was like, yes, this sounds like it's 0:18:29.393 --> 0:18:31.763 not true, but it's actually true. Or it would be like, 0:18:31.763 --> 0:18:33.563 that sounds like it's true, but it's actually not true. 0:18:33.563 --> 0:18:36.773 But this is like Snopes, except for in a way 0:18:36.773 --> 0:18:41.063 that you can visually explore and you can individually validate 0:18:41.063 --> 0:18:46.163 each particular component. So any particular background can now evaluate 0:18:46.163 --> 0:18:49.193 this with more transparency than it's ever been possible. And 0:18:49.193 --> 0:18:51.713 they could see the pieces. And of course people will 0:18:51.713 --> 0:18:54.053 be able to add all their favorite sources. Right. They 0:18:54.053 --> 0:18:57.633 could build arguments, they could evaluate the same argument using 0:18:57.633 --> 0:18:59.703 a different set of sauces and see if the conclusion 0:18:59.703 --> 0:19:02.133 comes up different. So this is why we're excited about 0:19:02.133 --> 0:19:04.833 this particular argument piece, which we've spent a lot of 0:19:04.833 --> 0:19:07.413 time on, because it's really important. This has the potential 0:19:07.413 --> 0:19:12.453 to significantly strengthen our shared understanding of reality, and will 0:19:12.453 --> 0:19:14.883 allow us to disagree with each other in a far 0:19:14.883 --> 0:19:17.523 healthier way. Here's one for the claim that there's a 0:19:17.523 --> 0:19:21.543 tiny teapot orbiting the sun. Okay, so there's a tiny 0:19:21.543 --> 0:19:26.163 teapot orbiting the sun investigated by space mission data claimed 0:19:26.163 --> 0:19:28.563 the teapot is too small to be detected by our 0:19:28.563 --> 0:19:33.423 current instruments. Result no unusual objects detected. No evidence of 0:19:33.423 --> 0:19:39.213 teapot found. Support. Support. Conclusion. False. Insufficient evidence to support 0:19:39.213 --> 0:19:42.033 the claim, which is exactly what it should say, right? 0:19:42.033 --> 0:19:44.423 You can't prove that it's not there, but not being 0:19:44.423 --> 0:19:46.553 able to prove that it's not there is not evidence 0:19:46.553 --> 0:19:48.863 that it is there. Right. So and that's what we 0:19:48.863 --> 0:19:51.833 end up getting to okay. So this is all additive. 0:19:51.833 --> 0:19:56.723 It all starts compounding and adding in with itself okay. 0:19:56.723 --> 0:20:02.573 So now we're talking about substrate plus I leading to 0:20:02.573 --> 0:20:07.433 actual action. Yeah yeah yeah I this I that uh, 0:20:07.433 --> 0:20:10.433 totally get it. But this is different. This is not 0:20:10.433 --> 0:20:14.933 about I okay. The substrate is not about I. It's 0:20:14.933 --> 0:20:18.053 about human meaning and progress. I is just a tool 0:20:18.053 --> 0:20:21.233 for helping that along. So think about this with everything 0:20:21.233 --> 0:20:24.893 you've heard so far about substrate and what's simultaneously happening 0:20:24.893 --> 0:20:27.833 inside the world of I. So context sizes, which is 0:20:27.833 --> 0:20:30.143 basically the size of prompts that you can use, are 0:20:30.143 --> 0:20:34.283 increasing massively. And inference costs, which is basically a fancy 0:20:34.313 --> 0:20:38.363 way of saying the cost to run individual AI pulls 0:20:38.363 --> 0:20:42.803 or questions are massively falling. Okay, what this means is 0:20:42.803 --> 0:20:46.133 we can basically like chocolate and peanut butter this thing 0:20:46.163 --> 0:20:50.483 together with Substrate's ability to have all of these things 0:20:50.483 --> 0:20:55.653 stored in very neat, structured ways and being seen together 0:20:55.653 --> 0:21:00.153 in graphs combined with AI's ability to hold multiple things 0:21:00.153 --> 0:21:03.243 in its mind at once, and then perform really cool 0:21:03.243 --> 0:21:08.793 actions and answer questions and make recommendations. So this combination 0:21:08.793 --> 0:21:13.473 is absolutely insane. So we could feed AI with our goals. Okay, 0:21:13.473 --> 0:21:16.263 us as an individual, us as a county, us as 0:21:16.263 --> 0:21:20.953 a city, us as a company. Us as a country. Okay, goals, KPIs, 0:21:20.953 --> 0:21:25.873 risks and have it help us untangle these, discuss them, 0:21:25.873 --> 0:21:30.403 debate about them, vote on them, whatever, and then take action. Okay, 0:21:30.403 --> 0:21:34.153 so here's some of the examples that we're most excited 0:21:34.153 --> 0:21:38.173 about for for this combination of substrate plus I. So 0:21:38.173 --> 0:21:43.783 first one science automated hypothesis to results workflows. So one 0:21:43.783 --> 0:21:46.243 big problem with science is that it takes so long 0:21:46.273 --> 0:21:49.033 to do science. And this is why it takes so long. 0:21:49.033 --> 0:21:51.373 Because look at look at all the different things you 0:21:51.373 --> 0:21:53.953 have to go through. It's hard to come up with ideas. 0:21:53.953 --> 0:21:56.683 It's hard to design experiments. It's hard to find funding 0:21:56.713 --> 0:21:59.863 to do the experiments. It's hard to interpret the results. 0:21:59.863 --> 0:22:02.203 It's hard to publish the results, and it's hard to 0:22:02.203 --> 0:22:04.963 get those results in front of the right people who 0:22:04.963 --> 0:22:07.673 actually have influence and might want to do something. So 0:22:07.673 --> 0:22:10.313 now imagine that we have our list of problems, our 0:22:10.313 --> 0:22:13.733 list of proposed experiments, all within substrate, a list of 0:22:13.733 --> 0:22:17.933 funding sources, also within substrate. They're all there. Now I 0:22:17.963 --> 0:22:21.383 can help us to do almost every step in that 0:22:21.383 --> 0:22:25.013 difficulty chain we just talked about above. Okay. So I 0:22:25.043 --> 0:22:29.543 can help us come up with or collect ideas and hypotheses. 0:22:29.543 --> 0:22:32.093 It can help us design experiments. It can help us 0:22:32.093 --> 0:22:34.943 collect and evaluate the best funding sources, because it's got 0:22:34.943 --> 0:22:37.193 the full list right there. So it could be like, okay, 0:22:37.193 --> 0:22:39.923 based on this, based on what your goals and your 0:22:39.923 --> 0:22:42.383 mission and what you're trying to do here, these are 0:22:42.383 --> 0:22:45.413 the types of groups that are most likely to want 0:22:45.413 --> 0:22:48.503 to fund you. Okay. Requesting funding now that it knows 0:22:48.503 --> 0:22:50.903 those groups, maybe it knows how to contact them because 0:22:50.903 --> 0:22:53.483 that information is inside a substrate. Okay. So we can 0:22:53.483 --> 0:22:56.433 write the perfect pitch for you to get funding, help. 0:22:56.433 --> 0:22:59.193 You set up the experiments, and we're going to need 0:22:59.193 --> 0:23:01.353 some humans and or robots to help with that part. 0:23:01.353 --> 0:23:05.943 But whatever. Running and monitoring the experiments, interpreting the results, 0:23:05.943 --> 0:23:09.723 all of this really, really right in the wheelhouse of AI, 0:23:09.753 --> 0:23:13.323 writing the paper and sharing the paper also possible with 0:23:13.323 --> 0:23:16.113 AI and getting better all the time. So in other words, 0:23:16.113 --> 0:23:19.893 we're talking about hypothesis to propose experiment to looking at 0:23:19.893 --> 0:23:24.333 funding sources, to acquiring funding, to running experiments, to publishing 0:23:24.333 --> 0:23:28.773 the results, to actually making progress. So in the beginning, 0:23:28.773 --> 0:23:31.083 this is still going to require a lot of human help, right? 0:23:31.083 --> 0:23:35.193 Especially at the idea and the running of the experiment phases. 0:23:35.193 --> 0:23:37.683 But over time, I will get better at that as well. 0:23:37.683 --> 0:23:41.553 But what we're talking about ultimately with this right here 0:23:41.553 --> 0:23:46.073 is the acceleration of science. Right. Because this cycle right 0:23:46.073 --> 0:23:48.173 here is leading to a whole bunch of failures and 0:23:48.173 --> 0:23:50.693 making us move on to do something else instead. But 0:23:50.693 --> 0:23:52.913 this whole pipeline, and I want to give credit to 0:23:52.913 --> 0:23:56.573 Joseph Thacker for thinking about a very similar thing at 0:23:56.573 --> 0:24:00.353 a similar time, like a year ago, talking about experiments 0:24:00.353 --> 0:24:04.853 and hypotheses and testing and stuff like that. So a 0:24:04.853 --> 0:24:09.313 lot of people probably thinking very similar things along these lines. 0:24:09.313 --> 0:24:11.983 But we are so excited about this. The idea of 0:24:11.983 --> 0:24:16.303 just being able to experiment and advance our knowledge forward 0:24:16.303 --> 0:24:18.823 through the most powerful mechanism that we're aware of, which 0:24:18.823 --> 0:24:24.103 is science accountability. Okay, this one's insane. Monitoring crime and 0:24:24.103 --> 0:24:27.943 corruption okay. So the reason it's so easy to get 0:24:27.943 --> 0:24:30.853 away with corruption and crime right now is because there 0:24:30.853 --> 0:24:36.113 aren't enough people watching gangs, cartels, embezzlers, dirty politicians. They're 0:24:36.113 --> 0:24:40.223 actually dropping evidence all the time. There's receipts, there's like travel, 0:24:40.223 --> 0:24:44.363 there's tickets, there's cameras. There's lots of different ways to 0:24:44.363 --> 0:24:47.093 know that a particular person was in a particular place, 0:24:47.093 --> 0:24:50.603 and they're not even being that careful, because it's actually 0:24:50.603 --> 0:24:53.243 so difficult to go and collect that stuff and bring 0:24:53.243 --> 0:24:56.133 it together into a narrative. So it usually takes a 0:24:56.133 --> 0:25:00.063 major journalist team or a massive law enforcement operation to 0:25:00.063 --> 0:25:04.383 dump thousands of hours of highly skilled work to collect 0:25:04.383 --> 0:25:06.873 all this different evidence. Okay, then you have to do 0:25:06.873 --> 0:25:10.203 the analysis. Then you have to formulate the conclusions. Then 0:25:10.203 --> 0:25:12.033 you have to document all of this. Then you have 0:25:12.033 --> 0:25:13.563 to like take it to the media. You have to 0:25:13.563 --> 0:25:16.833 get in front of people. And most crime and corruption 0:25:16.833 --> 0:25:20.613 slips by because nobody is simply watching. There aren't enough journalists. 0:25:20.613 --> 0:25:24.093 There aren't enough law enforcement teams who have the skills 0:25:24.093 --> 0:25:26.043 to do this stuff. And even if they had the skills, 0:25:26.043 --> 0:25:28.773 they don't have the resources and the time. So substrate 0:25:28.773 --> 0:25:32.253 plus AI versus dirty politician, this is a use case. 0:25:32.253 --> 0:25:35.313 So let's take substrate with some AI added on. And 0:25:35.313 --> 0:25:38.043 let's think about a dirty politician who is taking massive 0:25:38.043 --> 0:25:41.283 gifts from a particular lobbyist. Let's say this is some 0:25:41.283 --> 0:25:46.193 dirty Democrat, some dirty Republican. Independent doesn't matter. This has 0:25:46.193 --> 0:25:48.683 nothing to do with that. Problem is, there are so 0:25:48.683 --> 0:25:52.853 many donations. There are so many lobbyists, so many representatives, 0:25:52.853 --> 0:25:56.813 so many actual laws and bills and so many votes. 0:25:56.813 --> 0:25:59.693 But guess what? It's all public. We're talking about the 0:25:59.693 --> 0:26:03.593 US here. This is all public information. It is required 0:26:03.593 --> 0:26:06.533 by law that all the stuff is posted. The lobbyist 0:26:06.533 --> 0:26:10.463 groups must actually register themselves. The donations that they make 0:26:10.463 --> 0:26:14.093 to any particular representative, they have to be public records 0:26:14.093 --> 0:26:17.663 of meetings. I believe those are also public, and so 0:26:17.663 --> 0:26:22.703 are all the votes that representatives make on bills where 0:26:22.703 --> 0:26:26.453 lobbyists have been donating and trying to influence. So a 0:26:26.453 --> 0:26:29.543 nonprofit or even just a project, a small project that 0:26:29.543 --> 0:26:31.853 comes out of substrate or whatever, just a bunch of 0:26:31.853 --> 0:26:34.523 open source people could use AI to collect all of 0:26:34.523 --> 0:26:37.823 these different things, continuously put them in substrate, and they're 0:26:37.823 --> 0:26:40.313 already going in substrate because we're about to start dumping 0:26:40.313 --> 0:26:43.283 all the laws, all the different voting records, all the 0:26:43.283 --> 0:26:48.053 different histories for each lobbyist and also for each representative, 0:26:48.053 --> 0:26:51.323 all that goes right into substrate to inspect. And then 0:26:51.323 --> 0:26:54.593 I can ingest all of that at any given moment 0:26:54.593 --> 0:26:57.903 and basically tell us this. Here are all the bills 0:26:57.903 --> 0:27:00.423 written by that person. Here are all the summaries of 0:27:00.423 --> 0:27:03.393 those bills. Basically, what are they trying to do? Here's 0:27:03.393 --> 0:27:08.733 who those bills helped and who they hurt. Right? Here 0:27:08.733 --> 0:27:12.003 are all the lobbyists that care about those particular issues. 0:27:12.003 --> 0:27:16.563 Here are all the donations that those groups made to 0:27:16.563 --> 0:27:20.673 those representatives or that particular representative. And here's how the 0:27:20.673 --> 0:27:24.213 representative voted on every single bill. Then guess what? The 0:27:24.213 --> 0:27:27.453 AI is really good at, which we already know. Okay. 0:27:27.453 --> 0:27:31.443 Perform a comprehensive analysis of all legislation created and voted 0:27:31.443 --> 0:27:34.503 on from Bill Myers, Senator from Arkansas. I made that up. 0:27:34.503 --> 0:27:37.593 Hopefully there is no one like that cross-referenced with every 0:27:37.593 --> 0:27:41.133 single donation ever made to him, every dinner he's ever 0:27:41.133 --> 0:27:45.163 attended with them, every gift he's ever received, etc. finally, 0:27:45.163 --> 0:27:47.263 give me your assessment of whether or not he is 0:27:47.263 --> 0:27:51.463 being unduly influenced by this lobbyist and give me your 0:27:51.463 --> 0:27:54.583 reasons for this conclusion. Then it comes back with something 0:27:54.583 --> 0:27:59.293 like this assessment. This is a compromised politician reasoning. Osint 0:27:59.293 --> 0:28:03.913 reveals you can have Osint going and researching various things 0:28:03.913 --> 0:28:07.843 that this person is doing. Again, all public, all legal. Right? 0:28:07.843 --> 0:28:11.243 We're not talking about nasty stuff here. We're talking about legal. 0:28:11.243 --> 0:28:16.913 Public documents reveals that he was illegally gifted a small 0:28:16.913 --> 0:28:20.063 yacht last year, which he tweeted about and later deleted. 0:28:20.063 --> 0:28:22.853 He's had 31 dinners in the last 18 months with 0:28:22.853 --> 0:28:27.203 them totaling over like almost $15,000. Osint reveals that the 0:28:27.203 --> 0:28:32.483 Lobbyist's president used considerable influence to get Bill Meyer's daughter 0:28:32.483 --> 0:28:35.813 admitted to an exclusive private school that she wasn't actually 0:28:35.813 --> 0:28:39.263 qualified for. Every vote he's made about this particular issue 0:28:39.293 --> 0:28:42.443 has been in the direction that the lobbyist actually wants 0:28:42.443 --> 0:28:46.613 to happen, and previous votes before they started the relationship, 0:28:46.613 --> 0:28:50.543 the politician used to vote in the opposite direction. Therefore, 0:28:50.573 --> 0:28:53.903 the conclusion from this I. And by the way, you 0:28:53.903 --> 0:28:56.663 published the algorithm for the AI as well. We've got 0:28:56.663 --> 0:29:00.353 another project called fabric, where you publish the actual thoughts 0:29:00.353 --> 0:29:04.433 and directions given to the AIS so that they're inspectable. Right? 0:29:04.433 --> 0:29:06.833 You you don't want people to think, oh, they came 0:29:06.833 --> 0:29:09.383 to that conclusion because it's biased. No. You can look 0:29:09.383 --> 0:29:12.353 and see the actual instructions given to the AI to 0:29:12.353 --> 0:29:15.503 evaluate this thing impartially. So now you could trust this 0:29:15.503 --> 0:29:18.833 thing with a very high rate of confidence, because you 0:29:18.833 --> 0:29:21.213 could see how it's thinking and you could see all 0:29:21.213 --> 0:29:24.753 the data sources, just like before with arguments. So basically, 0:29:24.753 --> 0:29:30.303 the incredibly important objects of legislation, votes, etc. are all 0:29:30.303 --> 0:29:34.623 things that can be monitored and collected and stored within substrate. 0:29:34.623 --> 0:29:39.303 So that's that one next one here. Leadership. This is absolutely, 0:29:39.303 --> 0:29:43.953 absolutely powerful okay. I've been a consultant for a couple 0:29:43.953 --> 0:29:46.933 of decades now and I've worked with hundreds of startups, 0:29:46.933 --> 0:29:51.823 so many large corporations, and a big, big problem for 0:29:51.823 --> 0:30:00.043 most organizations. And that includes governments, families, individuals, startups, corporations. 0:30:00.043 --> 0:30:03.883 Everything is that they don't have clarity. Just like a person. 0:30:03.883 --> 0:30:06.163 It's hard to know what they think the issues are, 0:30:06.163 --> 0:30:09.463 what they specifically plan on doing, and how they plan 0:30:09.463 --> 0:30:13.003 to measure progress. So we see this with business leaders 0:30:13.003 --> 0:30:16.093 and we see this with politicians. So with substrate, we 0:30:16.093 --> 0:30:19.543 intend to make it so that check this out. Every 0:30:19.543 --> 0:30:22.783 leader will need to have a full detailed plan that 0:30:22.783 --> 0:30:27.043 has the following components. Imagine if everyone had this when 0:30:27.043 --> 0:30:30.433 they were pitching some plan for the high school, or 0:30:30.433 --> 0:30:33.073 some plan to be a principal, or some plan to 0:30:33.083 --> 0:30:35.993 be a community leader, or some plan to be a 0:30:35.993 --> 0:30:38.993 politician or whatever it is. Here's what I think the 0:30:38.993 --> 0:30:42.293 problems are. Remember, this is a problem object inside of 0:30:42.293 --> 0:30:44.483 substrate so you can actually look at them. Here's what 0:30:44.483 --> 0:30:47.363 I think the problems are. And coming off of the 0:30:47.363 --> 0:30:50.363 problem here's what I think the solutions are. Here are 0:30:50.363 --> 0:30:54.683 my proposed strategies for accomplishing that. Here are the KPIs. 0:30:54.683 --> 0:30:57.923 These are the metrics. Here's how actually going to measure myself. 0:30:57.923 --> 0:31:01.043 And how about this fire me if I don't get 0:31:01.043 --> 0:31:05.093 the KPIs to this number by this date. So at 0:31:05.093 --> 0:31:07.343 the end of the three and a half years when 0:31:07.343 --> 0:31:10.883 I'm going back up for reelection, I expect whatever the 0:31:10.883 --> 0:31:13.403 thing is, uh, the number of kids who don't go 0:31:13.403 --> 0:31:16.733 hungry in this high school, the literacy level in this 0:31:16.733 --> 0:31:20.213 high school, the unemployment rate, whatever it is, doesn't matter 0:31:20.213 --> 0:31:23.803 if I don't improve that number by X amount. By 0:31:23.803 --> 0:31:26.533 this date, you have my permission to vote me out. 0:31:26.563 --> 0:31:30.943 Imagine having that level of clarity and accountability for any 0:31:30.943 --> 0:31:34.903 leader trying to get any job doing anything. So really 0:31:34.903 --> 0:31:38.413 excited about that one. Okay, next one here. This is 0:31:38.413 --> 0:31:40.603 the best one okay. I saved the best for last. 0:31:40.603 --> 0:31:43.183 These are all adding on each other. Watch this. Look 0:31:43.183 --> 0:31:46.753 at this. I did a post a while back about 0:31:46.753 --> 0:31:51.073 how companies are essentially graphs of algorithms, and I encourage 0:31:51.073 --> 0:31:52.933 you to go check that out. I've got the link 0:31:52.933 --> 0:31:55.783 down here below, but it's like, okay, you have a 0:31:55.783 --> 0:31:59.203 company that processes bad images come in, they send them 0:31:59.203 --> 0:32:02.263 to you on the website, and you then do a 0:32:02.263 --> 0:32:04.873 number of things to that photo, right? You do a 0:32:04.873 --> 0:32:07.873 high quality scan, you repair it, you stylize it, you 0:32:07.873 --> 0:32:10.183 caption it, you send it to the user. Then there's 0:32:10.183 --> 0:32:13.603 a marketing group. A marketing group has an idea. It 0:32:13.603 --> 0:32:15.943 shares it with a team. They decide on the best version. 0:32:15.943 --> 0:32:18.433 They do this, they do a final vote. They do 0:32:18.433 --> 0:32:22.543 a launch campaign. Okay, or the uploading process. Visit the 0:32:22.543 --> 0:32:25.813 website create account, click the upload link. These are all 0:32:25.813 --> 0:32:29.503 individual steps okay, these are just algorithms. And this is 0:32:29.503 --> 0:32:32.863 a graph of algorithms where they're all connected. And you 0:32:32.863 --> 0:32:35.813 can break these into smaller and smaller pieces until you 0:32:35.813 --> 0:32:39.473 eventually see the world in this way. Okay. I'm just 0:32:39.473 --> 0:32:42.263 going to click into this thing just to highlight this 0:32:42.263 --> 0:32:45.353 a little bit. This is what it ends up looking like. Okay. 0:32:45.353 --> 0:32:48.833 It looks like this is what your companies are going 0:32:48.833 --> 0:32:52.673 to look like before too long okay. Not your companies. 0:32:52.673 --> 0:32:56.213 Everyone's company okay. AI consultancies are going to come in. 0:32:56.213 --> 0:32:58.683 They're going to see the world in this way. Everything 0:32:58.683 --> 0:33:02.163 is a process. Everything is a set of components. Okay? 0:33:02.163 --> 0:33:05.043 You got human decisions here. You got human teams here. 0:33:05.043 --> 0:33:07.383 Maybe the red are all the places that it's human 0:33:07.383 --> 0:33:09.963 and AI wants to replace that. Maybe the red is 0:33:09.963 --> 0:33:12.723 it's a little bit too manual. Maybe the red is 0:33:12.723 --> 0:33:16.593 the efficiency numbers are too low or it takes too long. 0:33:16.593 --> 0:33:18.693 Doesn't matter. You could drill in. You could see this 0:33:18.693 --> 0:33:21.033 is actually a human team. They actually need more people. 0:33:21.033 --> 0:33:24.093 We need to hire more people, etc. the point is, 0:33:24.093 --> 0:33:27.663 you can look at any company or any process as 0:33:27.663 --> 0:33:29.973 a series of these components, which is part of a 0:33:29.973 --> 0:33:33.243 larger process which has a flow. Okay. That's the important 0:33:33.243 --> 0:33:36.273 part here. Okay. Yeah. This is the piece here. Companies 0:33:36.273 --> 0:33:38.853 are just graphs of algorithms. But think of it this 0:33:38.853 --> 0:33:41.163 way I don't think I went far enough with that. 0:33:41.163 --> 0:33:45.333 Everything can be conceptualized in this way. Everything can be 0:33:45.333 --> 0:33:49.173 conceptualized in this way as a process. So essentially what 0:33:49.173 --> 0:33:51.963 you have is you have a current state of things, right? 0:33:51.963 --> 0:33:54.963 State of the universe, but smaller down to a scale 0:33:54.963 --> 0:33:57.423 that we're dealing with, like a company or a process 0:33:57.423 --> 0:33:59.823 or a team or a department. Okay. Then we have 0:33:59.823 --> 0:34:03.693 an action or event, like a decision is made. Um, 0:34:03.693 --> 0:34:07.413 you know, customer does a particular thing. They buy a thing. 0:34:07.413 --> 0:34:10.143 We need to pay out an insurance policy, whatever the 0:34:10.163 --> 0:34:12.263 thing is. Right? And then that results in a new 0:34:12.263 --> 0:34:17.843 state of things. So previous state action thing happens a cause. 0:34:17.843 --> 0:34:19.943 And then you have like an effect which is the outcome, 0:34:19.943 --> 0:34:22.463 which is the result. And if we add human components 0:34:22.463 --> 0:34:26.393 into this like people's jobs or making decisions, and we 0:34:26.393 --> 0:34:28.313 do this for like running a business or a country 0:34:28.313 --> 0:34:30.923 or family, we have additional pieces, right? We have people, 0:34:30.923 --> 0:34:33.983 we have decisions, we have strategies. We have lessons learned. 0:34:33.983 --> 0:34:37.733 We have conclusions. We have reasons. All of these are 0:34:37.733 --> 0:34:42.023 substrate components. There aren't infinite numbers of these. These are 0:34:42.023 --> 0:34:45.143 all look, what was the decision made? There aren't that 0:34:45.143 --> 0:34:49.163 many decisions okay. Buy more stock. Um, hire more people. 0:34:49.163 --> 0:34:52.103 This is the type of decision in there, right? Again, 0:34:52.103 --> 0:34:54.653 which we can connect and see visually. And what that 0:34:54.653 --> 0:34:57.443 results in is a way to tie together all of 0:34:57.443 --> 0:35:00.743 this into much larger graphs, graphs that we could use 0:35:00.743 --> 0:35:05.093 to describe the operations of anything a family, company, or 0:35:05.093 --> 0:35:08.813 even a country. So here's one for a small company. 0:35:08.813 --> 0:35:11.993 And I made this live. This is actual like I 0:35:11.993 --> 0:35:14.603 put in some fake company stuff. Look at this. No 0:35:14.603 --> 0:35:18.773 customers website visitor quality lead. Yes. No. Okay. Goes down. 0:35:18.773 --> 0:35:21.383 Look at this. Breaks down the thing. This is how 0:35:21.383 --> 0:35:24.243 we're about to see our companies. We're about to see 0:35:24.243 --> 0:35:27.123 all of anything that humans do as a set of 0:35:27.123 --> 0:35:30.303 processes like this. And it's fine if you want to 0:35:30.303 --> 0:35:34.533 resist this, but trust me, this is coming because the 0:35:34.533 --> 0:35:38.553 consultancies are coming. I got to stop for you. 40% 0:35:38.553 --> 0:35:43.113 of McKinsey's business right now is already in AI consulting. 40%. 0:35:43.113 --> 0:35:44.973 This thing happened like a year and a half ago. 0:35:44.973 --> 0:35:48.603 They are already doing almost half their business doing this. 0:35:48.603 --> 0:35:51.333 They're doing this. They're coming in. They're looking at companies 0:35:51.333 --> 0:35:54.033 and saying, what can I optimize? What can I improve? 0:35:54.033 --> 0:35:57.123 So that's pretty cool. We can create that. But that's 0:35:57.123 --> 0:35:59.553 not the full power of this yet. Check this out. 0:35:59.553 --> 0:36:02.643 The smarter the AI gets, it's not only going to 0:36:02.643 --> 0:36:06.603 see this and describe it and make some recommendations, but no, 0:36:06.603 --> 0:36:09.753 it can optimize the stuff. In other words, this is 0:36:09.753 --> 0:36:12.133 just the current state. What about future state? What about 0:36:12.133 --> 0:36:15.643 recommended state? Should this company merge departments? Where can we 0:36:15.643 --> 0:36:18.073 add more people? What kind of people should we hire? 0:36:18.073 --> 0:36:21.673 Which processes here are inefficient which can be replaced by AI? 0:36:21.673 --> 0:36:24.133 That's going to be the number one question somebody is asking. 0:36:24.133 --> 0:36:26.593 Where can we use more human decision making? If we 0:36:26.593 --> 0:36:30.103 wanted to grow, where would that happen? Now imagine this 0:36:30.103 --> 0:36:35.063 for a family, a corporation, a church, a city, a county, etc. 0:36:35.063 --> 0:36:37.553 and keep in mind, the more data that you have here, 0:36:37.553 --> 0:36:40.463 the better, right? You can. You can pull in all 0:36:40.463 --> 0:36:44.543 the data about conversion rates, churn rates, like whatever data 0:36:44.543 --> 0:36:47.543 you have that comes in. That's part of this whole equation. 0:36:47.543 --> 0:36:50.903 So here's an example for a security team because my 0:36:50.903 --> 0:36:54.803 background is in security. So it currently takes 3.5 business 0:36:54.803 --> 0:36:59.003 days to complete a security assessment. Delays in security assessment 0:36:59.003 --> 0:37:02.873 turnaround are the number one complaint in the engineering survey. 0:37:02.873 --> 0:37:06.863 So the security group sends a survey to engineering and says, hey, 0:37:06.863 --> 0:37:09.263 what do you love and what do you hate about security? 0:37:09.263 --> 0:37:11.603 This is their number one complaint. It takes too long 0:37:11.633 --> 0:37:14.423 to get stuff back. So you're slowing us down. So 0:37:14.423 --> 0:37:17.393 if we switch to the new flex scan model, which 0:37:17.393 --> 0:37:19.943 is a new model for doing security assessments in a 0:37:19.943 --> 0:37:23.923 more flexible way using fewer generalist security testers were able 0:37:23.923 --> 0:37:27.673 to complete type B and type C assessments 94% faster. 0:37:27.703 --> 0:37:30.673 This will give our senior testers two extra days to 0:37:30.673 --> 0:37:33.823 do a high impact security assessments, and this will also 0:37:33.823 --> 0:37:37.393 likely make engineering much happier with security and make them 0:37:37.393 --> 0:37:40.903 more likely to cooperate on our goals. So there's multiple 0:37:40.903 --> 0:37:44.623 steps the full articulation and breakdown of how a process 0:37:44.623 --> 0:37:47.853 is currently running. Visualization of that process to help with 0:37:47.853 --> 0:37:51.723 human understanding, and then analysis by AI of how to 0:37:51.723 --> 0:37:55.863 optimize the process to optimize the stated goals of the entity. 0:37:55.863 --> 0:37:59.313 That's the three pieces. And remember, the AI will also 0:37:59.313 --> 0:38:02.583 have access to the mission of the organization as well, 0:38:02.583 --> 0:38:05.193 and its goals and its strategies and its team members 0:38:05.193 --> 0:38:08.283 and its projects and its budget. Everything. So we'll have 0:38:08.283 --> 0:38:11.863 the full context on how resources are being spent relative 0:38:11.863 --> 0:38:14.473 to the desired outcomes, and we'll be able to see 0:38:14.473 --> 0:38:17.833 how the actual KPIs that we care about, the actual numbers, 0:38:17.833 --> 0:38:21.853 the actual metrics, are moving as we adjust these. So 0:38:21.853 --> 0:38:24.253 it'll be able to do all sorts of recommendations hiring 0:38:24.253 --> 0:38:28.453 new people, hiring people with specific skills, using more AI 0:38:28.453 --> 0:38:32.653 and high volume and low creativity areas, adjusting strategies based 0:38:32.653 --> 0:38:36.223 on goals and market conditions. Canceling projects like this one 0:38:36.223 --> 0:38:38.803 and that one. Like they're not even related to the goals. 0:38:38.803 --> 0:38:41.173 They're costing way too much. Way too much of the 0:38:41.173 --> 0:38:43.333 team is working on them. Let's get rid of those. 0:38:43.333 --> 0:38:47.293 This project over here, this human powered project, it's way better. 0:38:47.293 --> 0:38:50.713 It's producing way more revenue. It's way more tied to 0:38:50.713 --> 0:38:55.453 our actual goals and KPIs. So we're taking those people 0:38:55.453 --> 0:38:58.633 and putting them over here instead. So ultimately we're talking 0:38:58.633 --> 0:39:03.203 about the ability to continuously analyze and optimize any system 0:39:03.203 --> 0:39:06.653 using full knowledge of its goals and progress. And the 0:39:06.653 --> 0:39:09.683 more data about the system it has, the better it 0:39:09.683 --> 0:39:12.143 will actually work. And the smarter the AI gets, the 0:39:12.143 --> 0:39:15.683 better it actually works. Completely insane. All right, to summarize, 0:39:15.683 --> 0:39:18.803 the world is hard to understand, and things that are 0:39:18.803 --> 0:39:21.863 hard to understand are hard to discuss and improve. The 0:39:21.863 --> 0:39:24.733 goal of Substrate is to address this problem by making 0:39:24.733 --> 0:39:29.173 the things humans care about more visible. Discussable and Improvable. 0:39:29.173 --> 0:39:32.713 The framework is open source, lives on GitHub. At its core, 0:39:32.713 --> 0:39:36.583 it's a collection of crowdsourced lists of the things humans 0:39:36.583 --> 0:39:39.283 care about and that make up our discourse and society. 0:39:39.283 --> 0:39:42.403 One major problem that people and organizations have is not 0:39:42.403 --> 0:39:45.493 knowing and or being able to communicate what they are about. 0:39:45.493 --> 0:39:48.613 Using this framework, people and organizations will be able to 0:39:48.613 --> 0:39:51.763 articulate their values and purpose more clearly, which will not 0:39:51.763 --> 0:39:54.793 only help them, but everyone that they interact with. Substrate 0:39:54.793 --> 0:39:58.213 is magnified by I because I can or will soon 0:39:58.243 --> 0:40:01.213 be able to hold all of substrate in its mind 0:40:01.213 --> 0:40:03.253 at once. And from there, we'll be able to ask 0:40:03.253 --> 0:40:06.463 all sorts of meaningful questions such as what does that 0:40:06.463 --> 0:40:10.063 person or organization about? Are we pursuing the best path 0:40:10.063 --> 0:40:12.793 towards our goals, or what are the most critical mistakes 0:40:12.793 --> 0:40:15.823 I'm currently making? Ultimately, this will allow us to take 0:40:15.823 --> 0:40:19.123 action on these things. What action should I do right 0:40:19.123 --> 0:40:21.643 now to optimize this workflow? What should I do right 0:40:21.643 --> 0:40:24.493 now to achieve the best possible outcome that's aligned with 0:40:24.493 --> 0:40:28.183 my goals? In short, substrate is a way to better 0:40:28.183 --> 0:40:32.503 understand and optimize the things that we care about as humans. 0:40:32.503 --> 0:40:35.233 So we would love for you to get involved. I 0:40:35.233 --> 0:40:38.353 want to give thanks to people who are already involved 0:40:38.353 --> 0:40:41.563 and are like, really, really into this. Along with myself, 0:40:41.563 --> 0:40:46.393 Jonathan Dunn, Jason Haddox, Clint Gibler, Joseph Thacker, Joel Parrish 0:40:46.393 --> 0:40:49.993 and Robert Hansen, some of the few that are already 0:40:49.993 --> 0:40:53.563 involved and excited about this. And if you are interested, 0:40:53.563 --> 0:40:57.253 please go to the Substrate Project and, uh, let us 0:40:57.253 --> 0:41:00.523 know that you'd like to contribute. Thanks for your time.