WEBVTT - Why AI Will Replace Knowledge Workers 0:00:19.600 --> 0:00:22.120 Hey, what's up? So there's a common narrative right now 0:00:22.160 --> 0:00:24.840 that you don't need to worry about AI. If you 0:00:24.880 --> 0:00:29.000 are a knowledge worker, because humans are special and humans 0:00:29.000 --> 0:00:31.600 are unique in the way they approach work. They are 0:00:31.600 --> 0:00:35.280 better workers than AI. AI is just a text generator. 0:00:35.280 --> 0:00:38.080 It's just an LLM. It has no understanding. It has 0:00:38.080 --> 0:00:41.760 no knowledge. It doesn't really understand anything about what it's 0:00:41.760 --> 0:00:45.200 trying to do. It can't possibly replace human workers. This 0:00:45.200 --> 0:00:48.720 is a very popular narrative for multiple reasons. One, it 0:00:48.720 --> 0:00:50.960 makes us feel better. It makes us feel more safe 0:00:50.960 --> 0:00:53.920 because people are feeling afraid for good reason right now. 0:00:53.920 --> 0:00:58.960 And it's just a very common thread right now, basically saying, relax, 0:00:58.960 --> 0:01:01.760 you don't have to worry, right? AI is actually kind 0:01:01.760 --> 0:01:05.240 of garbage and humans are basically awesome. You don't need 0:01:05.240 --> 0:01:08.000 to worry. I think this is wrong. And more importantly, 0:01:08.000 --> 0:01:10.319 I think it's dangerous. I think it's wrong for a 0:01:10.319 --> 0:01:12.480 bunch of reasons, which I'm about to talk about. I 0:01:12.480 --> 0:01:15.640 think it's dangerous because if you adopt this mindset, if 0:01:15.640 --> 0:01:19.150 you accept this argument, you will not properly prepare to 0:01:19.190 --> 0:01:21.190 get ready for what is coming, right? So this is 0:01:21.190 --> 0:01:24.150 a polar opposite argument to the one that's being made. 0:01:24.150 --> 0:01:26.429 And keep in mind, it's being made by some very 0:01:26.430 --> 0:01:28.710 smart people. Right? I've got a friend we're going to 0:01:28.710 --> 0:01:31.749 talk about Marcus Hutchins, very smart security guy, a friend 0:01:31.750 --> 0:01:34.390 of mine, and we debated this once. Might do it 0:01:34.390 --> 0:01:37.350 again in the future. He's very smart. He's very technical, 0:01:37.390 --> 0:01:41.350 he's very knowledgeable. And he has this argument. He basically says, 0:01:41.350 --> 0:01:45.470 don't worry, it's overinflated. AI doesn't understand anything. And it's 0:01:45.470 --> 0:01:48.070 not just him. There's, you know, hundreds of people out 0:01:48.070 --> 0:01:51.550 here saying this very loudly and more importantly to me 0:01:51.790 --> 0:01:54.550 with a following, right? Marcus has a following. And there's 0:01:54.550 --> 0:01:57.350 people who have much larger followings than Marcus and me 0:01:57.390 --> 0:01:59.829 who are saying pretty much the same thing. And so 0:01:59.870 --> 0:02:02.950 I want to really come at this argument directly, and 0:02:02.950 --> 0:02:06.190 I want to start by talking about what work looks 0:02:06.190 --> 0:02:08.990 like inside of companies and sort of just like break 0:02:09.030 --> 0:02:11.990 down common things that I see all over the place 0:02:11.990 --> 0:02:14.429 and have seen over the course of my career. All right. 0:02:14.430 --> 0:02:17.260 So normally in these videos, I don't talk about my background. 0:02:17.300 --> 0:02:19.940 It's because I just don't care about my background. I 0:02:19.940 --> 0:02:23.260 don't care about other people's background. Usually. I mostly just 0:02:23.260 --> 0:02:24.700 want to hear the content. I want to hear the 0:02:24.700 --> 0:02:27.419 ideas that they're giving me, or if it's a tutorial 0:02:27.419 --> 0:02:29.980 or whatever, I want the content, right? In this case, 0:02:29.980 --> 0:02:32.579 I'm making a lot of claims about how businesses function, 0:02:32.579 --> 0:02:34.740 how they work. And so I think it raises a 0:02:34.739 --> 0:02:37.940 natural question which should be asked, which is, why should 0:02:37.940 --> 0:02:40.820 I believe you? Why do I think you understand anything 0:02:40.820 --> 0:02:44.299 about these companies? And the answer is, I've been working 0:02:44.299 --> 0:02:47.980 inside of companies for like over 25 years. My background 0:02:48.019 --> 0:02:52.299 is strictly information security, but I've held senior roles at Apple. 0:02:52.299 --> 0:02:55.139 I've built a program and a product there which is 0:02:55.139 --> 0:02:58.380 still running. I've been gone from there for a number 0:02:58.380 --> 0:03:01.740 of years now. I've worked at Robinhood, HP. I've worked 0:03:01.739 --> 0:03:06.059 in embedded in very large energy companies. Tons of experience actually, 0:03:06.060 --> 0:03:09.220 in the fortune ten, let alone like the fortune 100. 0:03:09.220 --> 0:03:12.100 And that's just the places that have actually been deeply 0:03:12.100 --> 0:03:15.600 embedded in as a consultant, like basically a full time 0:03:15.600 --> 0:03:19.320 worker there or working full time consulting though, which I 0:03:19.320 --> 0:03:22.239 think is even more important for the context here. I've 0:03:22.239 --> 0:03:24.880 actually worked for hundreds of companies in like the top 0:03:24.880 --> 0:03:28.880 10,000 companies in the world, like hundreds over these 25 years. 0:03:28.880 --> 0:03:31.719 Inside of all these companies, I have seen these problems 0:03:31.720 --> 0:03:35.000 that I'm about to describe over and over and over. 0:03:35.000 --> 0:03:37.400 And in fact, when I describe them, I'm sure you 0:03:37.400 --> 0:03:40.000 have also seen them. If you've worked in multiple companies 0:03:40.000 --> 0:03:41.800 over the course of your career, all of these are 0:03:41.800 --> 0:03:44.760 going to be extremely familiar. They're like not even controversial, 0:03:44.760 --> 0:03:47.240 just a little bit of background. That's where I'm coming from. 0:03:47.240 --> 0:03:49.800 And I just want to jump into this first part 0:03:49.840 --> 0:03:52.640 here talking about the companies and their problems. So the 0:03:52.640 --> 0:03:56.280 first thing to say about working in a, say, a 0:03:56.280 --> 0:03:59.520 top 10,000 company or whatever, just an average company like 0:03:59.520 --> 0:04:03.200 medium size, fairly large, we can talk about startups as well. 0:04:03.200 --> 0:04:06.600 But in general, the natural state of thinking is like, 0:04:06.720 --> 0:04:09.320 everything seems to be fine. Like, I don't even know 0:04:09.320 --> 0:04:12.440 why we're talking about bringing AI into these companies. Isn't 0:04:12.440 --> 0:04:15.310 everything going well? Like we have a good economy. Like, well, 0:04:15.310 --> 0:04:17.910 we have had a good economy, right? We've built plenty 0:04:17.910 --> 0:04:20.950 of companies. We've got plenty of market share. The Dow 0:04:20.990 --> 0:04:23.390 was doing great or whatever for all these years, all 0:04:23.390 --> 0:04:26.310 these decades. So what actually is the problem? The way 0:04:26.310 --> 0:04:29.910 I answer that is most companies are actually a giant 0:04:29.910 --> 0:04:33.110 soup sandwich or a football bat. These are military terms, 0:04:33.110 --> 0:04:37.950 but essentially there is absolute chaos in most companies. Most companies, 0:04:37.950 --> 0:04:40.990 it's actually difficult to find out what they're actually trying 0:04:40.990 --> 0:04:43.790 to do, what they're actually trying to build, especially as 0:04:43.830 --> 0:04:45.990 like a worker, right? It's very hard to get a 0:04:45.990 --> 0:04:48.910 clear message of like, what exactly is the mission? This 0:04:48.910 --> 0:04:51.870 is also the case for like fairly large startups as well. 0:04:51.870 --> 0:04:54.430 It starts to become a problem fairly early, you know, 0:04:54.470 --> 0:04:56.910 a few years into the startup, the vision is not 0:04:56.910 --> 0:04:59.510 super clear. Okay, so the first problem I want to 0:04:59.510 --> 0:05:01.310 talk about that I see in a lot of these 0:05:01.310 --> 0:05:03.790 companies and that I'm sure you've seen as well, is 0:05:03.790 --> 0:05:07.670 the vision isn't always very clear. In fact, sometimes it's non-existent, 0:05:07.670 --> 0:05:10.870 sometimes it's not conveyed all the way through the company. 0:05:10.870 --> 0:05:13.620 So people like aren't really sure exactly what they're working 0:05:13.620 --> 0:05:16.339 on and why. And so that's a huge problem. Another 0:05:16.340 --> 0:05:19.140 one is it's not clear. And this is kind of 0:05:19.180 --> 0:05:21.740 like fighting for top here. It's hard to say. These 0:05:21.740 --> 0:05:24.540 are really big issues. Hard to sort them. But another 0:05:24.540 --> 0:05:27.980 really big one is there aren't really SOPs. There aren't 0:05:27.980 --> 0:05:31.300 really standard operating procedures. That's what we call them in 0:05:31.300 --> 0:05:35.100 the military is SOPs. Most companies just call them processes, 0:05:35.100 --> 0:05:37.180 but like, how do you actually do things? And the 0:05:37.180 --> 0:05:40.260 problem isn't that there isn't some document somewhere that says 0:05:40.260 --> 0:05:42.220 how to do things. The problem is there might be 0:05:42.220 --> 0:05:44.980 multiple of them and they change all the time, or 0:05:44.980 --> 0:05:47.980 they decay over time, or people are referring to the 0:05:47.980 --> 0:05:51.779 wrong one. Okay, so it's like things just keep getting 0:05:51.780 --> 0:05:55.180 done incorrectly throughout the company. And the more layers of 0:05:55.180 --> 0:05:58.380 management you have, the larger the company, the longer the 0:05:58.380 --> 0:06:01.500 company's been around. The more employees you have, the worse 0:06:01.500 --> 0:06:04.860 this problem gets. Now I want to caveat something. There 0:06:04.860 --> 0:06:07.660 are certain companies, there are a few companies, very few 0:06:07.660 --> 0:06:10.410 companies who do this really well. So if you are 0:06:10.450 --> 0:06:13.090 one of these people who works at these companies or 0:06:13.090 --> 0:06:15.210 you're a manager at one of these companies and you're like, well, 0:06:15.250 --> 0:06:17.850 that's not true. I work at Google and it's pretty great. 0:06:17.850 --> 0:06:21.130 Or I work at so-and-so Giant Bank, which has had 0:06:21.130 --> 0:06:23.650 decades to figure this out and sort it out. Also, 0:06:23.650 --> 0:06:26.810 in energy companies, they tend to be extremely risk averse, 0:06:26.810 --> 0:06:29.570 and they have processes that they follow and they're really 0:06:29.570 --> 0:06:33.530 good at processes. I'm talking about the other 99.9%. Most 0:06:33.530 --> 0:06:37.170 companies it's not clear exactly how something is supposed to 0:06:37.170 --> 0:06:39.610 be done. And what ends up happening when this is 0:06:39.610 --> 0:06:42.770 the case is you have people doing work and the 0:06:42.770 --> 0:06:47.730 output is massively inconsistent. Okay? Massively inconsistent, different people doing 0:06:47.730 --> 0:06:51.290 that same job, even using the same document. They do 0:06:51.290 --> 0:06:54.610 the job differently. You know, it's common knowledge within the company. Oh, 0:06:54.650 --> 0:06:57.410 if Sarah does it well, yeah, sure. I love when 0:06:57.410 --> 0:07:00.930 Sarah produces this report or produces this security assessment or 0:07:00.970 --> 0:07:03.370 produces this artifact that we're going to send to the customer, 0:07:03.370 --> 0:07:06.050 she does it really well. Well, guess what, Jim over here, 0:07:06.050 --> 0:07:08.960 he looked at the same process. He got the same training. 0:07:08.960 --> 0:07:12.720 He had the same onboarding. His output is absolute garbage. Okay. 0:07:12.760 --> 0:07:15.680 That's even if there is a process in the company, 0:07:15.680 --> 0:07:19.280 oftentimes there isn't. It was just like, you know, common onboarding, right? 0:07:19.320 --> 0:07:21.120 You work at a company or whatever you come in, 0:07:21.160 --> 0:07:23.320 it's like, hey, you know, here's where you put your keys. 0:07:23.320 --> 0:07:26.000 We get lunch over here, you know, basically your day 0:07:26.000 --> 0:07:27.560 to day is going to be, you've got to take 0:07:27.560 --> 0:07:29.720 these emails, you've got to take it from here. Oh, 0:07:29.760 --> 0:07:32.400 make sure you don't send it to Mr. Johnson like this. 0:07:32.400 --> 0:07:35.000 He really hates that. Right. And so these are all 0:07:35.080 --> 0:07:37.720 being sort of loosely like stored in your mind. Let's 0:07:37.720 --> 0:07:40.040 say you're some sort of admin or something like that, right? 0:07:40.080 --> 0:07:43.080 So this admin named Chris suddenly starts learning all these 0:07:43.080 --> 0:07:46.640 things during onboarding. Yeah. Mr. Johnson doesn't like this. Make 0:07:46.680 --> 0:07:48.760 sure the format, if you send it to the group, 0:07:48.800 --> 0:07:51.280 make sure to include these people. Right. You're starting to 0:07:51.280 --> 0:07:53.720 get all these little things that you're picking up, either 0:07:53.720 --> 0:07:57.000 from the training or someone complaining about something and it 0:07:57.000 --> 0:07:59.840 starts accruing knowledge, right? The problem is that's stuck inside 0:07:59.840 --> 0:08:03.160 that one person's head. And based on how knowledgeable they are, 0:08:03.200 --> 0:08:05.640 based on how good they are at listening, you know 0:08:05.680 --> 0:08:08.710 how good they are at following processes like, did we 0:08:08.710 --> 0:08:10.630 actually write this down or is it stuck in my 0:08:10.630 --> 0:08:13.590 brain and me? Chris, can I just leave the company? 0:08:13.590 --> 0:08:16.270 And now that whole process has to start all over again. 0:08:16.310 --> 0:08:19.110 Very common stuff. You've seen this a million times. Another 0:08:19.110 --> 0:08:21.750 thing you've seen related to this million times is there 0:08:21.750 --> 0:08:24.630 are people in the company all throughout the company who 0:08:24.830 --> 0:08:28.430 just never, ever got good at doing the job. They 0:08:28.430 --> 0:08:31.990 never got good at following the process. They they don't 0:08:31.990 --> 0:08:35.309 produce good output. It's really, really hard to hire people 0:08:35.309 --> 0:08:38.190 in companies. It's also really, really hard to fire people 0:08:38.189 --> 0:08:41.390 at companies. This person, Chris or Mike or Raj or 0:08:41.390 --> 0:08:44.750 Sarah or whoever, can just be horrible at doing this 0:08:44.750 --> 0:08:48.510 for whatever reason. Maybe it's nepotism, maybe it's, you know, 0:08:48.550 --> 0:08:51.990 they're a nice person, maybe they're good at one little thing. And, 0:08:52.030 --> 0:08:54.590 you know, it's just you would hate to give that up. So, 0:08:54.830 --> 0:08:57.429 you know, they can stay on and keep breaking things 0:08:57.430 --> 0:09:00.430 and messing things up when they execute work other places. 0:09:00.430 --> 0:09:04.350 The point is this is endemic throughout companies. In a 0:09:04.350 --> 0:09:08.010 large company, there are dozens or hundreds, maybe even thousands 0:09:08.010 --> 0:09:11.490 of people who don't follow processes. They aren't good at 0:09:11.530 --> 0:09:14.570 their job at all. Right. Now, keep in mind there 0:09:14.570 --> 0:09:18.650 are also a players everywhere, right? There's a players everywhere 0:09:18.650 --> 0:09:20.810 and they do a fantastic job. They don't even need 0:09:20.810 --> 0:09:23.770 a process. They're awesome right. No one's disputing that there 0:09:23.770 --> 0:09:26.970 are awesome humans and that they can do work. That's 0:09:26.970 --> 0:09:29.730 not what this argument is about. This argument is about 0:09:29.730 --> 0:09:33.250 overall work. Okay, just a pop quiz here. Guess how 0:09:33.250 --> 0:09:38.850 much knowledge workers are paid every year? $50 trillion. That 0:09:38.850 --> 0:09:44.090 is the compensation roughly worldwide for knowledge workers. Okay, that 0:09:44.090 --> 0:09:46.730 includes the B players and C players and the D 0:09:46.730 --> 0:09:49.530 players and the F players, right? Forget the A players. 0:09:49.530 --> 0:09:51.850 We're not going to talk about them. Right. They're awesome. 0:09:51.850 --> 0:09:54.330 They're going to continue to be awesome with or without AI. 0:09:54.370 --> 0:09:58.410 Forget that we're talking about most work in most companies. Okay. 0:09:58.450 --> 0:10:02.050 So you've got these people. They can't follow the process. Okay. 0:10:02.210 --> 0:10:04.880 They produce bad work. Now you've got the fact that 0:10:04.880 --> 0:10:07.720 the job isn't that rewarding. They're not exactly sure what 0:10:07.720 --> 0:10:10.520 they're working on. Their bosses are changing all the time, 0:10:10.520 --> 0:10:13.240 by the way, their work that they're supposed to be doing, 0:10:13.240 --> 0:10:16.319 it's also changing all the time, right? Because the the 0:10:16.320 --> 0:10:18.760 company goals, you know, that they were given a long 0:10:18.760 --> 0:10:21.280 time ago, they change all the time. Guess what? That 0:10:21.280 --> 0:10:23.160 has to trickle down. It has to go down to 0:10:23.200 --> 0:10:26.559 a thing. Guess what that means. Meetings. Lots of meetings. Right? 0:10:26.600 --> 0:10:28.720 So your developer, you're trying to code, you're trying to 0:10:28.720 --> 0:10:30.880 do whatever. You're an engineer, you're trying to do a 0:10:30.880 --> 0:10:32.800 bunch of technical stuff. Nope. You got to get on 0:10:32.800 --> 0:10:35.440 this call. Why? Because so and so all these people 0:10:35.440 --> 0:10:37.760 need to have this thing explained to them. You already 0:10:37.760 --> 0:10:40.800 made a document. You already explained it to them 20 times. 0:10:40.800 --> 0:10:44.400 You've already been in 40 meetings about this very same topic. 0:10:44.400 --> 0:10:47.320 Nobody is listening. They keep doing the work. You keep 0:10:47.320 --> 0:10:50.280 having to correct the work that they did because they 0:10:50.280 --> 0:10:54.040 didn't follow the document, which you spent three weeks making, right? 0:10:54.080 --> 0:10:56.760 This is not even special. This is not like a 0:10:56.760 --> 0:10:59.960 surprise to you. If you've worked in any companies, most 0:10:59.960 --> 0:11:04.150 companies across your career, this is endemic now. Add on 0:11:04.150 --> 0:11:07.590 top of this, people are empire building. It's fucking Game 0:11:07.590 --> 0:11:11.070 of Thrones is what's being played out in real time 0:11:11.110 --> 0:11:16.709 across many companies for many employees, right? Managers are empire building. 0:11:16.710 --> 0:11:20.110 They're picking their favorite people also. What about when you 0:11:20.110 --> 0:11:21.950 have a good idea? How many times have you had 0:11:21.950 --> 0:11:24.910 a good idea? You submit it up through management because 0:11:24.910 --> 0:11:28.270 it actually would solve half the stuff that just ruins 0:11:28.270 --> 0:11:30.550 your soul. It crushes your soul to have to do 0:11:30.550 --> 0:11:33.830 this stupid work, which doesn't follow this process, which could 0:11:33.830 --> 0:11:37.230 easily be fixed by these things that you're recommending, right? 0:11:37.230 --> 0:11:39.350 You're like, look, if we just did this, it would 0:11:39.350 --> 0:11:41.710 cut four steps out of the process. It would cut 0:11:41.710 --> 0:11:44.590 four steps out of the process. Everything would be much better. 0:11:44.710 --> 0:11:47.990 I spent three weeks, I put together this proposal. I 0:11:48.030 --> 0:11:50.670 wrote all this documentation and they're like, yeah, yeah, I 0:11:50.670 --> 0:11:52.510 don't think so. Oh, by the way, you need to 0:11:52.510 --> 0:11:55.110 join this meeting. We're changing our goals again. Oh, by 0:11:55.110 --> 0:11:57.390 the way, you're going to report to Sarah now instead 0:11:57.390 --> 0:11:59.950 of Raj. And also we need you to redo all 0:11:59.950 --> 0:12:02.620 that stuff that you just did. And it's like, okay, great. 0:12:02.620 --> 0:12:05.940 Sounds awesome. Why do you think people dread Monday so much? 0:12:05.980 --> 0:12:08.340 Why do you think people on Sunday night are just, like, 0:12:08.380 --> 0:12:11.140 filled with this pit of despair because they get to 0:12:11.180 --> 0:12:13.500 maybe get in a car and drive for an hour 0:12:13.500 --> 0:12:17.060 to commute in, to go to another meeting, to redo 0:12:17.100 --> 0:12:20.660 their plans, hopefully try to get a bonus, right? Meanwhile, 0:12:20.660 --> 0:12:23.380 everyone around them is playing Game of Thrones and like 0:12:23.380 --> 0:12:27.220 breaking everything. Now I'm being a little overdramatic because there 0:12:27.220 --> 0:12:29.819 are some jobs where it's kind of chill and blah 0:12:29.820 --> 0:12:33.140 blah blah. It's everything's fine. But in general, it has 0:12:33.140 --> 0:12:36.540 been known for decades that this is a problem. It 0:12:36.540 --> 0:12:39.980 has been known that this dysfunction has been happening for decades. 0:12:39.980 --> 0:12:43.140 People were complaining about the same stuff. Look at office space. 0:12:43.140 --> 0:12:47.780 We've been making shows about corporate drudgery and stupidity for 0:12:47.780 --> 0:12:53.100 decade upon decade. We recently had severance soul killing work, right? 0:12:53.100 --> 0:12:55.060 You're sitting in the queue, blah blah blah. The manager 0:12:55.060 --> 0:12:58.260 comes by and smiles like nobody's having fun here. That 0:12:58.260 --> 0:13:01.650 manager's not having fun. You're not having fun. Why? Because 0:13:01.650 --> 0:13:04.210 your ideas are not getting into the system. You're not 0:13:04.210 --> 0:13:06.970 even clear exactly what you're working on and why. There's 0:13:06.970 --> 0:13:10.569 multiple layers separating between you and like whatever's happening at 0:13:10.570 --> 0:13:12.929 the top, which seems to change all the time, right? 0:13:12.970 --> 0:13:17.210 So this is not a good state. Okay. Human work 0:13:17.210 --> 0:13:22.610 is not the ideal state. Workflows, processes. What confidence do 0:13:22.610 --> 0:13:25.970 you have that if you have a nine step process, 0:13:26.010 --> 0:13:28.610 each of which has an SOP for it, and you 0:13:28.610 --> 0:13:31.050 give it to the new people that just came on 0:13:31.050 --> 0:13:33.850 and got onboarded and got trained, that the people are 0:13:33.850 --> 0:13:35.730 going to actually read the document and they're going to 0:13:35.730 --> 0:13:38.290 execute those nine steps. And at the end of it, 0:13:38.290 --> 0:13:39.970 you're going to have an artifact. You're going to have 0:13:39.970 --> 0:13:43.610 an outcome, a document project, whatever it is, and it's 0:13:43.610 --> 0:13:45.850 going to meet the standard. And what confidence do you 0:13:45.850 --> 0:13:48.810 have that that will also be the same the next 0:13:48.810 --> 0:13:51.170 time you run it. Because guess what, two of those 0:13:51.170 --> 0:13:53.730 people are on vacation and they didn't share their knowledge 0:13:53.730 --> 0:13:56.730 with anyone. Also, they didn't update the document because there 0:13:56.730 --> 0:14:00.550 was a change. This whole thing is chaos and decay 0:14:00.550 --> 0:14:04.070 and problems. Meanwhile, people are getting sick. People are changing 0:14:04.070 --> 0:14:06.870 jobs like every 2 or 3 years or whatever, depending 0:14:06.870 --> 0:14:09.750 on the industry. Then people are getting so like soul 0:14:09.750 --> 0:14:14.910 crushed that they're doing the bare minimum, right? I think 23% 0:14:14.910 --> 0:14:18.310 of total working time in a week is actually spent. 0:14:18.350 --> 0:14:21.990 Like 40 hours is like 23% of a full week, right? 0:14:22.030 --> 0:14:24.710 So people aren't working the full week. They're only working 0:14:24.710 --> 0:14:27.790 this little slot at work. Many people are doing the 0:14:27.830 --> 0:14:31.750 absolute bare minimum. They're making a fucking sport out of 0:14:31.790 --> 0:14:34.550 doing the least amount of work possible, right? That was 0:14:34.550 --> 0:14:37.470 also part of office space. Like this is extremely well 0:14:37.510 --> 0:14:41.950 known stuff. We are holding this situation up as the 0:14:41.950 --> 0:14:46.710 standard that AI can never achieve. AI can never, never 0:14:46.710 --> 0:14:50.390 pierce this because humans are just so innovative, right? It's 0:14:50.430 --> 0:14:53.990 absolutely insane to me. So another problem here is actually 0:14:54.030 --> 0:14:56.430 because I've been talking mostly about the worker, right? Mostly 0:14:56.430 --> 0:14:59.660 about the human doing this job, getting soul crushed. Let 0:14:59.660 --> 0:15:01.460 me tell you about the other side of it as 0:15:01.460 --> 0:15:04.660 a leader. Okay? Looking down because I've been here as well. 0:15:04.660 --> 0:15:07.580 So you're looking down at this company and you're trying 0:15:07.580 --> 0:15:09.980 to figure out you're looking with the CFO, you're looking 0:15:09.980 --> 0:15:12.780 at your headcount spend and you're like, okay, well, I'm 0:15:12.780 --> 0:15:19.220 paying $94 million on headcount across my entire staff, you know, engineers, admins, 0:15:19.260 --> 0:15:23.100 like all these different people. I'm paying $94 million and 0:15:23.140 --> 0:15:25.020 that's not even a lot of money. Let's just say 0:15:25.020 --> 0:15:28.980 that's a total cost of human compensation. Okay, what exactly 0:15:28.980 --> 0:15:32.060 are we getting? Can I see all of my projects? 0:15:32.100 --> 0:15:34.420 Can I see all of the work that's being done? 0:15:34.460 --> 0:15:37.780 Do I have a list of workflows that show all 0:15:37.780 --> 0:15:41.100 the different processes in my company, what the steps are 0:15:41.180 --> 0:15:44.540 and how they're actually being performed by the different teams? 0:15:44.540 --> 0:15:46.740 Do I know how much that costs? Do I know 0:15:46.740 --> 0:15:50.220 how long it takes to actually execute any of these things? 0:15:50.220 --> 0:15:53.740 Do you know? I think you do. But rhetorically, do 0:15:53.740 --> 0:15:57.330 you know what happens when a leader asks this question 0:15:57.330 --> 0:16:00.010 to their staff, to their company. Hey, I need a 0:16:00.010 --> 0:16:02.330 list of all work being done. I need a list 0:16:02.330 --> 0:16:04.970 of what everyone is doing. I need a list of 0:16:04.970 --> 0:16:08.250 how much this work is costing and how expensive it 0:16:08.250 --> 0:16:11.570 is at the different stages. Oh, also, is it good work? 0:16:11.690 --> 0:16:14.210 What's the average quality of an output of one of 0:16:14.210 --> 0:16:16.770 these processes? Do you know what the answer is? The 0:16:16.770 --> 0:16:19.730 answer is. Well, we could take a look. That's a 0:16:19.730 --> 0:16:22.650 good question, ma'am. We've been thinking about that as well. Yeah. 0:16:22.650 --> 0:16:25.010 If you would like, I could put together a project. 0:16:25.010 --> 0:16:27.050 What we could do is we could put together a 0:16:27.050 --> 0:16:29.130 list of all the stuff that we're doing and we 0:16:29.130 --> 0:16:33.290 could figure out, you know, what exactly our work is 0:16:33.330 --> 0:16:35.290 and how much it costs, and you know how good 0:16:35.290 --> 0:16:38.330 of a job we're doing. That is a major endeavor. Oh, 0:16:38.330 --> 0:16:41.410 by the way, it's separate from their work. This happens 0:16:41.410 --> 0:16:46.050 constantly inside of companies. Constantly. I'm often brought in to 0:16:46.090 --> 0:16:48.650 help them do this. I am often there when they 0:16:48.690 --> 0:16:51.490 are doing it with someone else. So if they spin 0:16:51.490 --> 0:16:53.930 up a project like this. Okay. Do you think this 0:16:53.930 --> 0:17:00.000 takes hours. Nope. Days? Nope. Weeks? Nope. Months. This is multi-week, multi-month, 0:17:00.000 --> 0:17:03.640 or multi-year. This is so work intensive. Keep in mind 0:17:03.640 --> 0:17:06.160 what has to happen to actually gather all this stuff. 0:17:06.280 --> 0:17:08.919 You actually have to call a bunch of meetings. Guess what? 0:17:08.920 --> 0:17:10.399 You have to tell them. Hey, I want you to 0:17:10.440 --> 0:17:13.480 put together the list of everything you're working on. I 0:17:13.480 --> 0:17:15.840 want you to note all the different tasks that you 0:17:15.840 --> 0:17:18.359 do during a day. This is our favorite. This is 0:17:18.359 --> 0:17:20.840 our absolute favorite thing to do. Yeah. Let me tell 0:17:20.840 --> 0:17:23.080 my manager exactly what I do. Why don't you look 0:17:23.080 --> 0:17:25.399 at my output? Why don't you look at my schedule? 0:17:25.520 --> 0:17:30.440 This is happening constantly while I'm speaking. Some managers, hundreds 0:17:30.439 --> 0:17:33.640 of them, thousands of them are telling their people, I 0:17:33.639 --> 0:17:36.200 need a detailed report of what you do and what 0:17:36.199 --> 0:17:38.199 you work on, and how much it costs and how 0:17:38.199 --> 0:17:40.840 long it takes, so I can put it up the chain, 0:17:40.840 --> 0:17:43.199 that's all. So they can get this visibility. And you 0:17:43.199 --> 0:17:45.519 know what's even crazier? If it's a large company, they 0:17:45.520 --> 0:17:48.239 don't even ask their people that. They're like, hey, let's 0:17:48.240 --> 0:17:53.990 bring in McKinsey or KPMG. Let's spend Thousand dollars to 0:17:54.030 --> 0:17:57.589 have a team of smiling 22 year olds come in 0:17:57.590 --> 0:18:01.389 here and do the audit for us. This is ridiculous. 0:18:01.429 --> 0:18:05.430 This is what AI is competing with. Okay, $50 trillion, 0:18:05.510 --> 0:18:08.710 by the way. It's around $10 trillion just for the US. 0:18:08.750 --> 0:18:13.550 $50 trillion worldwide is spent. The CEO and the CFO 0:18:13.590 --> 0:18:16.949 have very little idea what all is happening in the company. 0:18:16.949 --> 0:18:19.070 They don't know where the money is going. They don't 0:18:19.070 --> 0:18:21.350 know what the processes are. They don't know what the 0:18:21.389 --> 0:18:24.510 workflows are. They don't know how well it's going. They 0:18:24.510 --> 0:18:27.189 just have a rough idea based on, you know, performance 0:18:27.189 --> 0:18:29.669 of the company or whatever. Like when they hear from 0:18:29.669 --> 0:18:32.669 their managers and they see the stock doing whatever and 0:18:32.709 --> 0:18:35.429 obviously they have control, they can say, go do this, 0:18:35.429 --> 0:18:38.549 go do that. There is no visibility. They can't see shit. 0:18:38.590 --> 0:18:42.310 This is the state that AI is competing with, which 0:18:42.310 --> 0:18:46.830 is roughly chaos. Okay, so when you hear the argument, 0:18:46.830 --> 0:18:49.270 I just want this to crystallize in your mind. When 0:18:49.270 --> 0:18:52.899 you hear the argument, AI cannot compete with humans inside 0:18:52.899 --> 0:18:55.300 of a company. This is what they have to compete with. 0:18:55.500 --> 0:18:58.020 This is a bar that is not low. It is 0:18:58.020 --> 0:19:00.939 on the floor. It. It burned a hole through the floor. 0:19:00.939 --> 0:19:04.020 It is descending to the. The core of the earth. 0:19:04.020 --> 0:19:07.980 This is a low bar. AI can follow instructions. We're 0:19:07.980 --> 0:19:09.739 not even going to talk about AI yet, but the 0:19:09.740 --> 0:19:12.859 ability to follow instructions over and over, doing the same 0:19:12.859 --> 0:19:17.580 thing in a very uncreative, shitty way is better than 0:19:17.580 --> 0:19:20.179 what is done in most companies most of the time. Like, 0:19:20.219 --> 0:19:23.899 I cannot over express how bad the situation is and 0:19:23.899 --> 0:19:26.779 how low this bar is. So the crazy thing about 0:19:26.780 --> 0:19:31.420 this is everyone knew this before 2022. Everyone knew that 0:19:31.419 --> 0:19:36.020 this was soul crushing, shitty corporate knowledge work. Everyone knew 0:19:36.060 --> 0:19:38.179 that this was just like horrible. And I'm not talking 0:19:38.179 --> 0:19:40.580 about like some people who love their jobs 100%. A 0:19:40.580 --> 0:19:43.219 lot of people love their jobs. They're lucky. Maybe it 0:19:43.219 --> 0:19:45.819 lasts for a while, maybe it's short lived. Whatever. There 0:19:45.820 --> 0:19:48.379 are some people who love their jobs. Some companies are great, 0:19:48.379 --> 0:19:51.999 they tend to be smaller, but whatever. In general, everyone 0:19:52.000 --> 0:19:55.760 knew that this was soul crushing before 2022. They knew 0:19:55.840 --> 0:19:59.040 it before 2022. They knew it in the 90s. They 0:19:59.040 --> 0:20:01.600 knew it in the 80s. Like, that's why it's part 0:20:01.600 --> 0:20:04.639 of culture. Everyone knew this was a problem and guess 0:20:04.639 --> 0:20:07.040 what we were doing? Guess what everyone was doing throughout 0:20:07.040 --> 0:20:09.720 the culture, they're trying to figure out a solution. What 0:20:09.719 --> 0:20:13.000 is a better way? Because this is obviously not the 0:20:13.000 --> 0:20:15.840 way that humans should live. That was the state before 0:20:15.879 --> 0:20:19.760 2022 when AI arrived. Okay, so another argument I'm going 0:20:19.800 --> 0:20:23.559 to continue beating up on us humans here. One last piece. 0:20:23.600 --> 0:20:26.319 I got positive stuff at the end to say about humans. 0:20:26.320 --> 0:20:28.760 So don't worry about that. You can hang in for that. 0:20:28.760 --> 0:20:32.480 This next piece is another negative piece about humans, and 0:20:32.480 --> 0:20:35.319 it's actually something that a lot of people don't think about. 0:20:35.320 --> 0:20:38.720 So and it relates to how AI thinks versus how 0:20:38.719 --> 0:20:42.000 humans think. Another sort of part of this argument is 0:20:42.000 --> 0:20:45.800 that humans just fundamentally think differently. And we just have 0:20:45.840 --> 0:20:49.709 like a lot more integrity of thought. We're more cohesive. 0:20:49.750 --> 0:20:54.670 We have understanding. We know ourselves. You know we can introspect. 0:20:54.709 --> 0:20:58.350 We can be creative. We have intuition. We have all 0:20:58.350 --> 0:21:01.749 these advantages over an AI which is like this, you know, 0:21:01.830 --> 0:21:04.789 dead black box, basically. So I want to sort of 0:21:04.830 --> 0:21:08.069 like dispel this as well. This is another angle to 0:21:08.109 --> 0:21:11.749 this argument, which also makes absolutely no sense. The crux 0:21:11.750 --> 0:21:14.990 of this argument is that humans are better thinking machines. 0:21:14.990 --> 0:21:18.550 Humans are much better thinking machines than AI. If you've 0:21:18.550 --> 0:21:22.509 ever studied or read any meditation documentation, or tried to 0:21:22.550 --> 0:21:24.669 learn it, or tried to study it or whatever, what 0:21:24.669 --> 0:21:28.670 it reveals about humans is absolutely mind blowing. Okay, so 0:21:28.669 --> 0:21:32.109 the first thing that it reveals is that humans have 0:21:32.109 --> 0:21:35.510 no idea what they're thinking at any given moment. Humans 0:21:35.510 --> 0:21:38.629 are basically like, if you could just zoom in and 0:21:38.629 --> 0:21:42.149 listen to the minds of people. Okay, sitting in their cars, 0:21:42.350 --> 0:21:45.310 sitting in their office space or their cube at work 0:21:45.310 --> 0:21:48.179 or wherever, you're just going to zoom in on these 0:21:48.179 --> 0:21:51.379 conversations that are happening in their mind. It's like hamsters 0:21:51.379 --> 0:21:54.340 running on a wheel. You'll just zoom into like Ravi here, 0:21:54.340 --> 0:21:56.660 and Ravi is like, just can't believe he said that. 0:21:56.659 --> 0:21:58.740 All that work I put in in the project, like, 0:21:58.740 --> 0:22:00.539 I hope that guy gets fired. I just hope he 0:22:00.540 --> 0:22:02.500 gets fired. Like, you know what I would love? Here's 0:22:02.500 --> 0:22:05.179 what I would love. Just imagine like, okay, so a 0:22:05.179 --> 0:22:07.739 new boss comes in and the new boss sees my 0:22:07.740 --> 0:22:10.139 stuff and just absolutely loves it. And then he looks 0:22:10.139 --> 0:22:13.260 at Chris's stuff. Can't you know, Chris, I can't believe 0:22:13.260 --> 0:22:15.859 that happened. Yeah, yeah. So looks at Chris's stuff and 0:22:15.859 --> 0:22:18.220 he's like, oh, this is garbage. You know, the one 0:22:18.219 --> 0:22:20.180 I really like is I like this one. And then 0:22:20.179 --> 0:22:22.899 that's me. That's mine. Like that would be so much better. 0:22:22.899 --> 0:22:25.300 And then, oh, did I feed the dog? I think 0:22:25.300 --> 0:22:27.260 I need to feed the dog. Hey, let me text someone. 0:22:27.260 --> 0:22:29.460 I need to feed the dog. Oh, crap. It's almost lunch. 0:22:29.459 --> 0:22:31.580 I got to get ready for lunch. That is the 0:22:31.580 --> 0:22:35.699 mind of a person. That randomness is happening constantly. Now 0:22:35.699 --> 0:22:38.059 check this out. When you try to meditate, what you 0:22:38.060 --> 0:22:40.580 do is you try to observe thoughts. It is the 0:22:40.580 --> 0:22:42.659 craziest thing to try to do if you try to 0:22:42.699 --> 0:22:44.859 observe your thoughts. So you can pause this if you 0:22:44.859 --> 0:22:46.889 want and try to observe your thoughts for a second. 0:22:46.929 --> 0:22:50.169 It's a really fun and crazy exercise. It is like 0:22:50.169 --> 0:22:54.770 this sliding window of chaos. Just scrolling through your brain 0:22:54.770 --> 0:22:56.929 and it is literally the stuff I was just talking about. 0:22:56.969 --> 0:22:59.409 Like you go into this, this woman over here, this 0:22:59.409 --> 0:23:02.250 guy over here, this kid over here, the janitor over here. 0:23:02.250 --> 0:23:05.929 It's just this rumination, this grinding of like, oh yeah, 0:23:05.969 --> 0:23:07.649 I should, I'm going to watch that thing. I've got 0:23:07.689 --> 0:23:09.490 to pick up those things. I don't know why he 0:23:09.490 --> 0:23:12.050 said that. Yeah. You know, I should get a promotion 0:23:12.090 --> 0:23:14.210 like these people don't respect me. Blah blah blah. It's 0:23:14.209 --> 0:23:18.090 just this constant like droning. Where did those thoughts come from? 0:23:18.090 --> 0:23:21.489 Who is making those thoughts and who is observing the thoughts? 0:23:21.490 --> 0:23:23.889 The answer is you have no idea. I have no 0:23:23.889 --> 0:23:26.329 idea what my thoughts are. I have no idea how 0:23:26.330 --> 0:23:28.290 I'm going to finish this sentence. We'll talk about that 0:23:28.290 --> 0:23:31.290 one in a second. Nobody has any idea where their 0:23:31.290 --> 0:23:33.530 thoughts are coming from. Do you know what that reminds 0:23:33.530 --> 0:23:36.770 me of? That reminds me of an LLM. When you 0:23:36.770 --> 0:23:39.689 poke an LLM and you say, hey, say smart stuff 0:23:39.689 --> 0:23:44.170 about logistic workflows and pipelines and you know how to 0:23:44.169 --> 0:23:47.119 move heavy freight from the East Coast to the West 0:23:47.119 --> 0:23:51.039 Coast on an 18 Wheeler. It just goes okay, and 0:23:51.320 --> 0:23:54.680 it starts spewing, starts spewing words, right? And it turns 0:23:54.679 --> 0:23:57.600 out they're really good and they're awesome. And the words 0:23:57.600 --> 0:24:00.600 make sense. And it's actually factual unless it's not right. 0:24:00.639 --> 0:24:03.199 And you could do deep research or whatever and correct that. 0:24:03.199 --> 0:24:05.640 The point is, you could ping a human to do 0:24:05.639 --> 0:24:08.080 that and they start spewing words, or you could ping 0:24:08.119 --> 0:24:10.800 an LLM and start spewing words. Does the LLM have 0:24:10.800 --> 0:24:13.320 any idea where its thoughts are coming from? No. Do 0:24:13.320 --> 0:24:16.159 you have any idea where your thoughts are coming from? No. 0:24:16.159 --> 0:24:19.800 It's a complete black box. Complete black box. We have 0:24:19.800 --> 0:24:23.480 no idea where our thoughts are coming from, where our 0:24:23.480 --> 0:24:27.919 dreams are coming from, where our concerns, our feelings, anything. 0:24:27.919 --> 0:24:31.119 It's bubbling out of us from a black hole. Like 0:24:31.159 --> 0:24:33.480 actually think about this. If you are in the middle 0:24:33.480 --> 0:24:36.999 of a sentence like I have been multiple times here already, 0:24:37.000 --> 0:24:40.359 I have no idea how I'm going to finish the sentence. 0:24:40.359 --> 0:24:42.840 I'm not planning a sentence. First of all, I didn't 0:24:42.840 --> 0:24:45.139 come up with the sentence at all. I didn't plan 0:24:45.139 --> 0:24:47.419 that sentence. I just said, these are spewing out of 0:24:47.419 --> 0:24:50.260 me just like an LLM. I have no idea how 0:24:50.260 --> 0:24:52.619 the sentence is going to end. I am just as 0:24:52.619 --> 0:24:56.139 surprised about anything I say as you are. Who is 0:24:56.139 --> 0:24:59.499 receiving it right now? Guess what that sounds like also AI. 0:24:59.619 --> 0:25:02.499 So going back to the work thing and now this, 0:25:02.500 --> 0:25:06.340 there's this view that we are put together. We are cohesive. 0:25:06.540 --> 0:25:10.659 We are self understanding. It's complete and utter crap. We 0:25:10.659 --> 0:25:12.859 have no idea where our thoughts are coming from. We 0:25:12.859 --> 0:25:15.900 have no idea how we're going to finish sentences. The 0:25:15.899 --> 0:25:19.899 whole thing is like this random generator. So here's another example. 0:25:19.899 --> 0:25:22.019 I'm going to ask you a question right now. What 0:25:22.020 --> 0:25:26.260 are your top seven favorite restaurants in the entire world? 0:25:26.260 --> 0:25:29.899 You've got 30s to answer. What is your brain doing 0:25:29.899 --> 0:25:33.379 right now? You're thinking about these. Who are you receiving 0:25:33.379 --> 0:25:36.739 content from? Where is that content coming from? Now you 0:25:36.740 --> 0:25:41.020 might say, well, I have memories and those memories are me. Yeah, sure. 0:25:41.060 --> 0:25:45.010 AI also has memories. AI has weights and AI has 0:25:45.010 --> 0:25:48.810 context and it goes and retrieves those and brings back stuff, 0:25:48.810 --> 0:25:51.450 but it brings it back in a different way each time. Right. 0:25:51.490 --> 0:25:54.770 When it spews an answer to top seven restaurants, it's 0:25:54.770 --> 0:25:58.929 reviewing content and spewing it in a way that is unpredictable. Right. 0:25:58.969 --> 0:26:02.010 It's random, hallucinated. Some might say, well, guess what? When 0:26:02.010 --> 0:26:04.649 I come up with my list of seven items, I 0:26:04.649 --> 0:26:06.409 have no idea what they're going to be. If you 0:26:06.409 --> 0:26:09.409 ask me multiple days or weeks apart or whatever, and 0:26:09.409 --> 0:26:12.369 I'm not like prepared, the list will be different because 0:26:12.369 --> 0:26:15.449 some I couldn't remember at the time. Actually, some my 0:26:15.449 --> 0:26:19.850 preference might change in between. This is not a static, cohesive, 0:26:19.969 --> 0:26:23.569 dependable process. Okay. And that takes us right into the 0:26:23.570 --> 0:26:27.290 memory one. Okay. One of the most reproduced things in 0:26:27.490 --> 0:26:30.530 psychology is the fact that memories are not what we 0:26:30.530 --> 0:26:33.209 think they are in the human mind. When we store memories, 0:26:33.209 --> 0:26:36.850 we're storing it with like emotion and bias and sort 0:26:36.850 --> 0:26:40.800 of feelings. And when we recall things, it's extremely well 0:26:40.840 --> 0:26:42.959 documented all through science, and I'm going to have a 0:26:42.959 --> 0:26:45.840 whole bunch of science links in the blog that accompanies this. 0:26:45.840 --> 0:26:49.080 So you can actually review this literature, but you actually 0:26:49.080 --> 0:26:51.960 don't store. You're not a hard drive, you're not storing 0:26:52.000 --> 0:26:55.080 unaltered ones and zeros. Okay. First of all, we don't 0:26:55.080 --> 0:26:57.680 know how we're storing things in the brain. That's step one. 0:26:57.679 --> 0:27:01.680 But we do know that when we recall events versus 0:27:01.679 --> 0:27:05.000 what happened, they're massively different and it gets worse over 0:27:05.000 --> 0:27:07.399 time and it gets worse the more you recall it. 0:27:07.399 --> 0:27:11.879 So our memory system, even our memory system is completely flawed. 0:27:11.879 --> 0:27:13.840 So we don't know where our thoughts are coming from. 0:27:13.840 --> 0:27:16.080 We don't know how we're going to finish sentences. And 0:27:16.080 --> 0:27:19.479 our memories are not static. They're not ones and zeros. 0:27:19.520 --> 0:27:23.080 They're actually extremely fluid. And then there's the weird fact 0:27:23.080 --> 0:27:24.800 that when you go to sleep and wake up, you're 0:27:24.840 --> 0:27:28.639 actually like a different person. Your entire being, your entire identity, 0:27:28.679 --> 0:27:31.960 like went offline and came back up, like spun up 0:27:31.959 --> 0:27:35.520 with like mostly the same memories, like mostly the same feelings, 0:27:35.520 --> 0:27:38.159 but you're kind of like a different person. So this 0:27:38.159 --> 0:27:42.310 whole concept of companies are doing great work. They have 0:27:42.310 --> 0:27:46.709 great workflows. People follow the processes. That's garbage. Then you 0:27:46.709 --> 0:27:49.149 have the concept of humans are like these, you know, 0:27:49.189 --> 0:27:53.509 bastions of like perfection. And we are the perfect thinking machines. 0:27:53.510 --> 0:27:56.670 And AI can't possibly hold a candle to us. Also 0:27:56.669 --> 0:28:00.109 highly flawed. The whole thing is built on just absolute madness. 0:28:00.109 --> 0:28:05.149 So once again, as thinking machines, the bar is extremely low. Okay, 0:28:05.189 --> 0:28:06.790 I want to show you this. I came up with 0:28:06.790 --> 0:28:09.990 this capability stack, whatever. It's just an image. And by 0:28:09.990 --> 0:28:13.189 the way, this is not like settled science. Like this 0:28:13.189 --> 0:28:15.310 is all still up in the air. People are still 0:28:15.310 --> 0:28:17.270 trying to figure all this out. I read a lot. 0:28:17.310 --> 0:28:20.350 I've got pretty decent mental models. Just work with me 0:28:20.350 --> 0:28:23.990 on this. Think of these blue pill boxes here, these 0:28:23.990 --> 0:28:28.710 blue layers here as fundamentals of a knowledge worker. Okay? 0:28:28.909 --> 0:28:31.429 Things that make us good at doing knowledge work. So 0:28:31.429 --> 0:28:36.749 first of all, it's knowledge. So collecting, organizing and accessing facts, right? 0:28:36.790 --> 0:28:43.100 Knowledge Facts. Easy understanding. Turning knowledge into concepts. Okay, so 0:28:43.140 --> 0:28:45.979 it's the ability to actually like, let it seep in 0:28:45.980 --> 0:28:48.259 and actually be able to say, okay, oh, that knowledge 0:28:48.260 --> 0:28:51.820 is related to that knowledge. Oh, okay. That forms a concept, right? 0:28:51.860 --> 0:28:54.500 So now we have concepts. So that's, that's what I'm 0:28:54.500 --> 0:28:58.100 calling understanding here. Again, we could use lots of different definitions. 0:28:58.100 --> 0:29:02.019 Just work with these knowledge understanding. Now on top of that, 0:29:02.020 --> 0:29:04.340 in addition to having a whole bunch of facts stored, 0:29:04.340 --> 0:29:07.340 that's not enough to be a good knowledge worker. Understanding. Okay, 0:29:07.380 --> 0:29:09.739 now we're getting closer, right? You have to have understanding 0:29:09.740 --> 0:29:12.500 of all these different concepts and everything and how they relate. 0:29:12.540 --> 0:29:15.940 Intelligence is the next one above. This is the ability 0:29:15.940 --> 0:29:20.219 to face obstacles, right? You're navigating obstacles. You're navigating the 0:29:20.219 --> 0:29:24.180 world around you, and you have goals that you are pursuing. Okay, 0:29:24.219 --> 0:29:27.499 so you're pursuing goals. So basically, if you're more intelligent, 0:29:27.500 --> 0:29:30.820 you're able to navigate the world better and better achieve 0:29:30.820 --> 0:29:33.740 your goals. So that's the definition of intelligence that we're 0:29:33.740 --> 0:29:37.930 using here. And the final one is creativity. Now this one. 0:29:37.930 --> 0:29:40.769 I would even argue not to get ahead of myself here, 0:29:40.770 --> 0:29:43.730 but I would argue that this is truly something that 0:29:43.730 --> 0:29:47.050 humans have that AI doesn't have right now. And who 0:29:47.050 --> 0:29:49.050 knows when they're going to get it or how much 0:29:49.050 --> 0:29:51.690 they're going to get. But even the first three are 0:29:51.690 --> 0:29:55.770 really solid and most knowledge work, okay? Most knowledge work 0:29:55.770 --> 0:29:59.290 that people are doing is actually the knowledge, the understanding 0:29:59.290 --> 0:30:03.530 and the intelligence creativity. You can have some people have 0:30:03.650 --> 0:30:05.690 some people have a lot of creativity. Some people have 0:30:05.690 --> 0:30:07.930 a little, some people have none, but they're still really 0:30:07.930 --> 0:30:10.490 good knowledge workers. I would argue you could even have 0:30:10.490 --> 0:30:14.650 an A player who crushes knowledge, understanding and intelligence and 0:30:14.730 --> 0:30:16.890 is still really great at their job. Then you have 0:30:16.890 --> 0:30:21.410 creativity on top of that. So this is like intuition art, 0:30:21.450 --> 0:30:25.810 you know, liberal arts thinking, philosophy, you know, logic, like 0:30:25.850 --> 0:30:28.370 all the different things, the ability to come up with 0:30:28.370 --> 0:30:32.130 novel things, net new things. That's ultimately what we're talking 0:30:32.130 --> 0:30:35.510 about here. So taking just the left side, just the 0:30:35.510 --> 0:30:39.749 blue layers. One of the arguments for why humans cannot 0:30:39.750 --> 0:30:42.670 be replaced is that they have expertise. This is a 0:30:42.670 --> 0:30:45.350 thing that my friend Marcus Hutchins is always talking about. 0:30:45.390 --> 0:30:48.510 He's like, hey, look, there's like, expertise is the thing 0:30:48.510 --> 0:30:51.150 that AI can't do. Like it can have facts and 0:30:51.150 --> 0:30:54.950 it can regurgitate them. But like, that's not real useful. 0:30:54.950 --> 0:30:57.470 That's not what work is. That's ultimately what we're trying 0:30:57.469 --> 0:31:00.030 to answer here. What is work, right? And how does 0:31:00.030 --> 0:31:04.190 it relate to human capabilities versus AI capabilities? So where 0:31:04.190 --> 0:31:07.350 is expertise on this list? And maybe you think I 0:31:07.350 --> 0:31:09.510 missed a layer. I don't think I did, but I'm 0:31:09.510 --> 0:31:13.070 challenging you to tell me what I missed. Where is 0:31:13.070 --> 0:31:16.870 the expertise layer here if it's not one of these four? 0:31:16.990 --> 0:31:21.550 So my answer to that question is expertise is knowledge, 0:31:21.590 --> 0:31:25.870 understanding and intelligence combined. Okay, so let's just go through 0:31:25.870 --> 0:31:29.390 some different work tasks. Okay? You've got engineering, you've got 0:31:29.430 --> 0:31:33.820 like civil engineering, you've got software developers, you've got site 0:31:33.860 --> 0:31:38.980 reliability engineer. You've got product manager, you've got project manager, 0:31:38.980 --> 0:31:42.299 you've got. Head of finance. You've got what do we 0:31:42.340 --> 0:31:45.580 got lots of other jobs right. Million different jobs. Admin. 0:31:45.620 --> 0:31:49.500 You're an admin for the company. You you handle you know, meals. 0:31:49.500 --> 0:31:52.339 You handle emails, you handle events. You handle all that 0:31:52.340 --> 0:31:54.900 kind of stuff. Maybe there's marketing as well. If I 0:31:54.900 --> 0:31:57.779 didn't already mention that, tons of different roles, thousands of 0:31:57.780 --> 0:32:00.900 different roles. Here's the question. When somebody is good at 0:32:00.940 --> 0:32:03.740 that thing, which of these are they good at? Which 0:32:03.780 --> 0:32:06.579 of these do they have in combination and which of 0:32:06.580 --> 0:32:10.700 these are required? Obviously, some require more creativity than others. 0:32:10.700 --> 0:32:13.820 Some don't require any creativity at all. I would argue 0:32:13.860 --> 0:32:17.020 they pretty much all require intelligence. So this is really 0:32:17.020 --> 0:32:22.739 interesting actually. Here's the question. Why hasn't existing automation previous automation, 0:32:22.740 --> 0:32:26.180 why has it not killed off knowledge workers for all 0:32:26.180 --> 0:32:30.219 these decades? We've had automation for decades. We've had automation 0:32:30.219 --> 0:32:33.930 for hundreds of years, and we've had computers for decades 0:32:33.930 --> 0:32:38.730 and decades, and we've had automation scripts, you know, programming. 0:32:38.730 --> 0:32:41.569 We've had these things for decades as well with really, 0:32:41.570 --> 0:32:46.330 really advanced automations, excel, all these different things. Why can't 0:32:46.370 --> 0:32:50.290 they replace a knowledge worker? Why have they not replaced 0:32:50.290 --> 0:32:55.050 a knowledge worker? My answer to that is understanding and intelligence. Okay. 0:32:55.090 --> 0:32:59.729 Previous automation basically just churned it executed it followed a 0:32:59.770 --> 0:33:03.650 list of steps deterministically, right? What it can't do is 0:33:03.650 --> 0:33:06.890 the thing that everyone must do as a knowledge worker 0:33:06.890 --> 0:33:09.930 inside of the chaos that I described earlier. Remember I 0:33:09.930 --> 0:33:13.210 was describing how everything is chaos. Everything's changing all the time. Oh, 0:33:13.250 --> 0:33:15.290 the goals are different. This is different when you come 0:33:15.290 --> 0:33:17.890 into work on an average day and an average company 0:33:17.890 --> 0:33:19.850 as a knowledge worker, you don't know what you're going 0:33:19.890 --> 0:33:22.690 to get. They might be like, hey, we're moving offices. Hey, 0:33:22.730 --> 0:33:25.850 our remit is completely different. All of our jobs are different. 0:33:25.850 --> 0:33:28.450 We now report to a different person. They have different goals. 0:33:28.450 --> 0:33:31.320 Maybe the goals are the same, but they like everything 0:33:31.360 --> 0:33:35.079 to be done differently. They like blue templates. So you 0:33:35.080 --> 0:33:37.800 need to put all your stuff in brand new templates. Well, 0:33:37.800 --> 0:33:41.279 if it's base automation, okay, it didn't learn how to 0:33:41.280 --> 0:33:44.400 use blue templates. It didn't learn how to have new goals. 0:33:44.400 --> 0:33:47.440 All that stuff needs to be rewritten. Intelligence is the 0:33:47.440 --> 0:33:52.080 ability to adapt, to change, to navigate, to work around 0:33:52.080 --> 0:33:55.040 and basically say, okay, well, here are my actual goals, 0:33:55.040 --> 0:33:57.000 here's what I know and I'm supposed to do. Here's 0:33:57.000 --> 0:34:00.160 my knowledge, here's my understanding of what is meant by 0:34:00.160 --> 0:34:02.520 all of this. So here are the changes I'm going 0:34:02.560 --> 0:34:04.959 to make to the way I do my work based 0:34:04.959 --> 0:34:09.840 on these new requirements that intelligence piece previous automation didn't have. Okay, 0:34:09.880 --> 0:34:12.600 now you add creativity. On top of that, the ability 0:34:12.600 --> 0:34:15.799 to come up with new ideas and innovate and stuff 0:34:15.800 --> 0:34:19.480 like that, that's way beyond automation. And I would say 0:34:19.680 --> 0:34:24.280 arguably today, AI, it's also way beyond today's AI. Now, 0:34:24.280 --> 0:34:26.919 there are some edge cases where it's starting to seep in, 0:34:26.920 --> 0:34:30.310 but let's just set those aside. So this is what 0:34:30.310 --> 0:34:34.670 makes a human valuable as a knowledge worker. Expertise is 0:34:34.670 --> 0:34:37.670 all for. But I would say mostly it's knowledge, understanding 0:34:37.670 --> 0:34:40.749 and intelligence. So give me an example. Try to think 0:34:40.750 --> 0:34:43.230 of an example. And this is sort of me arguing 0:34:43.230 --> 0:34:46.430 with Marcus here. Give me an example of expertise that 0:34:46.430 --> 0:34:50.750 isn't this. Think about what expertise is. It's lots of experience. 0:34:50.750 --> 0:34:53.910 It is the knowledge, right? It's the understanding of the concepts. 0:34:53.910 --> 0:34:57.190 It's the ability to adapt with the intelligence and navigate 0:34:57.190 --> 0:35:00.830 and change and everything. The combination of those combined with 0:35:00.830 --> 0:35:05.910 experience is ultimately what gives you your expertise right now. 0:35:05.950 --> 0:35:10.310 Maybe you would argue, well, what about experience? Maybe experience 0:35:10.310 --> 0:35:13.510 should be on the list, but experience is memory, right? 0:35:13.549 --> 0:35:19.189 Experience is recall experiences, accrued iterations on seeing different problems. 0:35:19.190 --> 0:35:22.390 And ultimately this is why I have the model. This way. 0:35:22.390 --> 0:35:26.469 It is actually improving your knowledge or your understanding. Okay, 0:35:26.509 --> 0:35:29.450 your intelligence is not really improving. That's just a thing 0:35:29.450 --> 0:35:31.290 that you have. It's a tool that you can use, 0:35:31.290 --> 0:35:35.690 but your knowledge improves with the experience. An experienced person 0:35:35.690 --> 0:35:39.689 has more knowledge. An experienced person has more understanding. They 0:35:39.690 --> 0:35:43.969 have better understanding, okay? Which means they have linked more 0:35:43.969 --> 0:35:47.209 and more facts and more and more situations and more 0:35:47.210 --> 0:35:51.170 and more patterns to interlock, to connect to each other 0:35:51.170 --> 0:35:53.129 so they could just see patterns better. And this is 0:35:53.130 --> 0:35:57.569 related to intelligence as well. But somebody with 25 years experience, 0:35:57.569 --> 0:36:02.930 30 years experience, 50 years experience of doing financial audits, 0:36:02.930 --> 0:36:06.329 they just immediately see things. They see patterns. Okay, here's 0:36:06.330 --> 0:36:09.450 the question. What part of that does AI not have? 0:36:09.489 --> 0:36:13.569 First of all, how is AI a knowledge? Okay, insane. Okay. 0:36:13.610 --> 0:36:16.249 It's read every book. It's a deep expert. I don't 0:36:16.250 --> 0:36:20.010 want to lead the witness here. It's deeply knowledgeable about 0:36:20.009 --> 0:36:23.770 many fields, tens of thousands of fields and subdomains. It 0:36:23.770 --> 0:36:27.760 can rattle off all the details, explain everything to you 0:36:27.799 --> 0:36:30.039 whatever way you want to receive it. It could write 0:36:30.080 --> 0:36:32.960 a full book on the topic by itself. It could 0:36:32.960 --> 0:36:36.919 make flashcards for you about this very specific thing that 0:36:36.920 --> 0:36:39.519 is like very few people on earth actually know about, 0:36:39.520 --> 0:36:41.960 but the AI knows about it. It's also read most 0:36:41.960 --> 0:36:48.160 every book effectively. Okay, so knowledge, forget about it. Okay. Understanding. Yes. Okay. 0:36:48.319 --> 0:36:51.759 You could argue AI's don't have understanding. There is a 0:36:51.759 --> 0:36:55.840 type of understanding that AI's don't have and that is experience. Okay. 0:36:55.880 --> 0:37:00.839 If you argue that understanding requires subjective experience, well, then 0:37:00.839 --> 0:37:03.560 they don't have it right because they don't experience anything. 0:37:03.600 --> 0:37:09.800 My definition of understanding is connecting concepts, finding patterns across domains, 0:37:09.799 --> 0:37:14.239 and AI is unbelievably good at this. It is mind 0:37:14.279 --> 0:37:19.560 blowingly good at connecting things, cross linking patterns, identifying a 0:37:19.560 --> 0:37:22.840 pattern in a piece of content. Oh, this relates to 0:37:22.880 --> 0:37:26.710 the 1930s Russia in the following way. Oh well, how 0:37:26.710 --> 0:37:31.669 does that relate to chicken feed farming practices from Idaho? 0:37:31.710 --> 0:37:34.430 It's like, oh, well good question. Here's how it relates. 0:37:34.430 --> 0:37:37.390 It can cross-reference anything. It holds it all in its 0:37:37.390 --> 0:37:40.029 mind all at once and can reference this. It can 0:37:40.029 --> 0:37:43.470 also do that supplemented by tools to go and research 0:37:43.469 --> 0:37:45.989 anything and add to it. It could also do that 0:37:45.989 --> 0:37:49.310 by looking at Rag or at a giant composite list 0:37:49.310 --> 0:37:52.149 of additional knowledge that you give it. Right? So it 0:37:52.150 --> 0:37:54.629 doesn't have to actually be in the model weights. It 0:37:54.630 --> 0:37:58.270 could actually be part of a unified system, which includes 0:37:58.270 --> 0:38:01.430 going to get new information and cross-referencing to make sure 0:38:01.430 --> 0:38:06.270 it's real. Right. So knowledge and understanding, crushing it, absolutely 0:38:06.270 --> 0:38:09.989 crushing it. Let's go to intelligence. Can AI navigate the 0:38:09.989 --> 0:38:15.350 world and achieve goals? Yes, absolutely. And this is fairly modern, right? 0:38:15.390 --> 0:38:17.149 It's been going on for a couple of years. But 0:38:17.150 --> 0:38:20.190 the last year, the last six months. Are you kidding me? 0:38:20.190 --> 0:38:23.700 An agent infrastructure that has access to all these tools 0:38:23.700 --> 0:38:27.580 and all your company's knowledge, all this context and a really, 0:38:27.580 --> 0:38:30.900 really smart model and a system of agents that actually 0:38:30.900 --> 0:38:33.700 can use that context. It actually has a list of 0:38:33.700 --> 0:38:36.859 rules to follow. And okay, so it's proceeding down a 0:38:36.860 --> 0:38:40.219 track and you're like, okay, well, now imagine that the 0:38:40.219 --> 0:38:43.100 customer says, we actually can't do the deal in Japan. 0:38:43.100 --> 0:38:46.460 We're actually need to do it in the UK. And actually, 0:38:46.500 --> 0:38:49.180 here's another restraint. So the size of the project has 0:38:49.180 --> 0:38:51.579 to be larger than this. By the way, you need 0:38:51.580 --> 0:38:56.300 to employ only people who have residency in Singapore or 0:38:56.299 --> 0:38:59.980 the UK. Also, we need some people who have understanding 0:39:00.020 --> 0:39:03.580 of Scotland and also everyone should be over the age 0:39:03.580 --> 0:39:07.580 of 35, right? And also make sure the report format 0:39:07.580 --> 0:39:10.660 has to follow this particular template. And also we have 0:39:10.660 --> 0:39:14.299 to take into account these other 90 things which I'm 0:39:14.299 --> 0:39:17.259 about to give you. Those are new requirements. That is 0:39:17.259 --> 0:39:21.980 dynamic changing requirements on the ground. Here's another requirement. Also 0:39:22.049 --> 0:39:25.809 go and research all things related to this. Find all 0:39:25.810 --> 0:39:29.610 scientific documents related to this and do thorough research for 0:39:29.610 --> 0:39:32.169 a day and a half using all your resources. Bring 0:39:32.170 --> 0:39:35.129 in all the sources to support the document that you're 0:39:35.170 --> 0:39:38.089 actually going to produce for this customer. That's going to 0:39:38.089 --> 0:39:44.010 cost $300,000. Those are dynamic, changing situations. Everyone who's used 0:39:44.049 --> 0:39:48.569 AI recently, who is a decent user, knows that you 0:39:48.569 --> 0:39:51.410 could throw anything at these eyes. You could throw anything 0:39:51.450 --> 0:39:53.370 at them. And it's like, oh, well, in that case, 0:39:53.370 --> 0:39:55.770 you should do this. And you're like, yeah, but here's 0:39:55.770 --> 0:39:58.890 another requirement. Oh, well, in that case, we should adjust. 0:39:58.890 --> 0:40:02.890 So can it do intelligence? Do modern AIS have intelligence 0:40:02.890 --> 0:40:09.129 according to this definition, 100% expertise. Like I said, expertise, knowledge, 0:40:09.130 --> 0:40:14.209 understanding and intelligence, mostly knowledge and understanding. Okay. Intelligence, I 0:40:14.210 --> 0:40:17.930 would say definitely. Sure. Let's include that creativity. Again, I'm 0:40:17.969 --> 0:40:20.609 not sure you need it for expertise. You can have 0:40:20.719 --> 0:40:23.799 somebody who is just the craziest bookworm and is just 0:40:23.799 --> 0:40:26.799 like locked in. You give them this task. They just 0:40:26.799 --> 0:40:28.719 crush it. You give them a different task. They just 0:40:28.759 --> 0:40:32.200 absolutely crush it. Are they inventing anything new? Not really. 0:40:32.200 --> 0:40:34.440 Are they coming up with a new way to do it? Nope. 0:40:34.480 --> 0:40:36.479 The way that they were trained or the way they 0:40:36.480 --> 0:40:38.600 decided they wanted to do it, that's the way they 0:40:38.600 --> 0:40:42.239 execute every single time doesn't require creativity. Okay, so now 0:40:42.239 --> 0:40:46.640 let's slide over here. Let's look at the AI one intelligence, understanding, 0:40:46.640 --> 0:40:49.880 knowledge creativity. Let's just take it off the list. Let's 0:40:49.880 --> 0:40:53.719 just pretend it doesn't have it. AI has the expertise. 0:40:53.839 --> 0:40:56.880 So this argument that a bunch of people are making 0:40:56.880 --> 0:41:01.120 very smart people, very knowledgeable, very technical, they're saying that 0:41:01.120 --> 0:41:06.480 you can't possibly replicate human expertise, human thinking, the quality 0:41:06.480 --> 0:41:09.680 of human thought, the quality of human thinking, the quality 0:41:09.680 --> 0:41:13.120 of human execution of work, everything that we've been talking 0:41:13.120 --> 0:41:16.400 about tells us that is just not true. In fact, 0:41:16.400 --> 0:41:20.180 the polar opposite is actually more true that even decent 0:41:20.219 --> 0:41:24.459 AI today, given these types of tasks, will just crush 0:41:24.500 --> 0:41:28.580 an average knowledge worker. Just absolutely crush them on tasks 0:41:28.580 --> 0:41:31.979 that are there defined. Okay, what are we doing all day? 0:41:32.020 --> 0:41:36.220 We're accepting emails. We're responding to emails, we're producing reports, 0:41:36.219 --> 0:41:41.140 we're writing code. These basic commodity tasks make up. I'm 0:41:41.140 --> 0:41:44.700 making up a number 95% of all knowledge work. It's 0:41:44.700 --> 0:41:49.300 common stuff. The only thing that stops somebody from walking 0:41:49.299 --> 0:41:52.580 in off the street who has the requisite, like education 0:41:52.580 --> 0:41:55.380 or training in this field, which again is just knowledge 0:41:55.380 --> 0:41:59.219 and understanding, mostly knowledge. Someone walking in brand new to 0:41:59.259 --> 0:42:01.060 do that in the company, they just don't have the 0:42:01.060 --> 0:42:04.700 company knowledge. Well guess what? They accrue it over time. 0:42:04.700 --> 0:42:07.419 They go through onboarding, they read the documentation, they do 0:42:07.420 --> 0:42:09.219 all that well guess what? AI can do that in 0:42:09.219 --> 0:42:12.820 a few seconds. Here's a core point about all of this. 0:42:12.860 --> 0:42:16.380 A big reason that people think there's a gap between 0:42:16.380 --> 0:42:18.930 what the humans have and what A's have. A big 0:42:18.930 --> 0:42:21.810 reason they think there's a gap is because the humans 0:42:21.810 --> 0:42:26.009 have not articulated the stuff. It's not written down. Think 0:42:26.009 --> 0:42:28.850 about this. The way that AI has gotten so good 0:42:28.850 --> 0:42:31.770 in the last year is largely because of cloud code. 0:42:31.810 --> 0:42:34.689 What did cloud Code do? Did it have a dramatically 0:42:34.690 --> 0:42:38.050 better model than the other models? No. What it did 0:42:38.049 --> 0:42:42.249 was scaffolding a harness around the model. What it did 0:42:42.250 --> 0:42:47.530 is it built context and context management relations between documents, 0:42:47.529 --> 0:42:52.289 allowing it to have knowledge, understanding and intelligence. It linked 0:42:52.290 --> 0:42:55.489 together everything that the model needs to be able to 0:42:55.529 --> 0:42:59.609 perform to actually do intelligence tasks, to actually do work. 0:42:59.650 --> 0:43:01.930 That is what this thing did. That is what cloud 0:43:01.930 --> 0:43:05.290 code did. Okay. That stuff being written down. Okay, so 0:43:05.330 --> 0:43:08.410 another thing related to this that's extremely hot right now. 0:43:08.410 --> 0:43:11.610 It's been hot for a number of months is skills. 0:43:11.610 --> 0:43:17.840 What are skills? Skills are a combination of resources, tools, Context. Knowledge. 0:43:18.040 --> 0:43:20.200 It's a whole bunch of stuff written down. This is 0:43:20.200 --> 0:43:24.279 the center piece. It's essentially the core benefit of a 0:43:24.319 --> 0:43:27.160 harness versus a model. A model is just like a 0:43:27.160 --> 0:43:31.040 whole bunch of knowledge, right? Once you realize how powerful 0:43:31.080 --> 0:43:34.200 these skills are, once you realize how powerful it is 0:43:34.200 --> 0:43:36.959 to have all this context in one place, that sort 0:43:36.960 --> 0:43:40.080 of wraps an AI model, you start to realize that 0:43:40.080 --> 0:43:43.720 the problem is not the AI. The gap between human 0:43:43.719 --> 0:43:48.200 expertise and AI expertise is not the model, okay? It's 0:43:48.200 --> 0:43:52.000 not the fact that humans have expertise and AI doesn't. 0:43:52.040 --> 0:43:56.600 It's the fact that we have not articulated our expertise 0:43:56.719 --> 0:44:01.680 down into files, into tools, into contexts. I've been doing 0:44:01.680 --> 0:44:04.440 this for over a year now. I've built a system 0:44:04.440 --> 0:44:08.000 called Pi. It is a context management system. It is extraordinary. 0:44:08.000 --> 0:44:10.080 It's open source. You can check it out. Everyone is 0:44:10.080 --> 0:44:12.719 building with skills now. It is becoming the most powerful 0:44:12.719 --> 0:44:16.230 concept in all of AI is skills. Why is that? 0:44:16.230 --> 0:44:20.990 Because it is literally expertise packaged for this very smart 0:44:20.989 --> 0:44:24.629 model to then be able to use the problem inside companies. 0:44:24.870 --> 0:44:29.310 That makes people think there's a disconnect between human capability 0:44:29.310 --> 0:44:32.390 and AI capability. Nobody has any of this written down. 0:44:32.390 --> 0:44:35.430 All the stuff that is in Robbie's mind and Susie's 0:44:35.430 --> 0:44:38.749 mind and Kristy's mind and Chris's mind, all the stuff 0:44:38.750 --> 0:44:41.710 that's in their minds, it's not written down. Their expertise 0:44:41.750 --> 0:44:44.630 is not captured the way this is supposed to be done, 0:44:44.630 --> 0:44:47.790 this particular task, this sending of the emails, the creation 0:44:47.790 --> 0:44:50.189 of the report, the writing of the code. It's not 0:44:50.190 --> 0:44:54.190 written down anywhere. It's passed from human brain to human brain. 0:44:54.190 --> 0:44:56.710 And they go on vacation, they go on maternity leave, 0:44:56.870 --> 0:45:00.709 they retire, they jump companies every couple of years. That 0:45:00.710 --> 0:45:04.230 knowledge just dies. What's starting to happen now is knowledge 0:45:04.230 --> 0:45:10.350 is dispersing from brains, from very smart, high expertise, people 0:45:10.350 --> 0:45:13.510 into skills. This is why there was a giant stock 0:45:13.620 --> 0:45:16.939 drop when anthropic came out with a whole bunch of 0:45:16.940 --> 0:45:21.100 legal stuff, right? Those are just processes. Those are expertise 0:45:21.100 --> 0:45:25.180 captured in documents, which AI can then follow. So the 0:45:25.180 --> 0:45:29.980 expertise gap between humans and AI is actually the failure 0:45:30.020 --> 0:45:34.459 so far of us to articulate all the different chaos 0:45:34.500 --> 0:45:37.899 things that we talked about in the beginning, all that chaos, 0:45:37.940 --> 0:45:41.499 all those random pieces of knowledge inside of people's brains. 0:45:41.660 --> 0:45:44.980 Like there's some super old guy who's 62 years old, 0:45:45.020 --> 0:45:48.100 you know, his name's Cliff. He's the one you call 0:45:48.219 --> 0:45:50.980 when the things go down and there's no docs, you 0:45:50.980 --> 0:45:52.540 can't find a thing. You don't know how to get 0:45:52.540 --> 0:45:54.700 the server back up. People don't even know what servers 0:45:54.700 --> 0:45:57.620 are anymore. You call Cliff. Cliff has never written the 0:45:57.620 --> 0:46:00.700 stuff down. Cliff has never created a skill. He's never 0:46:00.700 --> 0:46:04.419 done a full debrief interview to capture his knowledge. Now 0:46:04.420 --> 0:46:05.660 you're not going to get all of it. It's not 0:46:05.660 --> 0:46:08.820 going to be perfect. But if Cliff could document that, 0:46:08.819 --> 0:46:10.859 and by the way, lots of people are. Lots of 0:46:10.860 --> 0:46:15.200 people are documenting what they know about Cliff's topic and 0:46:15.200 --> 0:46:18.360 putting it into skills and their open source. This stuff, 0:46:18.360 --> 0:46:21.279 this is P that's seeping into the pool and it's 0:46:21.279 --> 0:46:23.279 not going to come out again, but inside of a 0:46:23.319 --> 0:46:26.320 given company, if that stuff were to be captured and 0:46:26.319 --> 0:46:29.400 brought into this context system like I have with Pi, 0:46:29.440 --> 0:46:32.479 like lots of other people have like are captured in 0:46:32.480 --> 0:46:36.759 skills that's just absolutely flooding the internet right now. Flooding, 0:46:36.799 --> 0:46:40.360 especially the AI space. That is what bridges the gap 0:46:40.360 --> 0:46:46.200 between human expertise and AI expertise. And this is happening fast. 0:46:46.200 --> 0:46:49.080 It's happening extremely fast. All right. So, so far I've 0:46:49.080 --> 0:46:51.640 been talking about the problem. I've been talking about what 0:46:51.640 --> 0:46:53.520 I believe to be the problem with the way people 0:46:53.520 --> 0:46:57.400 are thinking about this, right? Comparing humans to AI, thinking 0:46:57.400 --> 0:47:01.120 that work is like easily done by humans and it's like, 0:47:01.120 --> 0:47:03.759 it's working fine. Like what's even the problem now I 0:47:03.759 --> 0:47:06.719 want to transition into and we also talked about the 0:47:06.719 --> 0:47:09.879 whole expertise thing and like what that really means and 0:47:09.880 --> 0:47:11.709 what I think the gap is there. What I want 0:47:11.750 --> 0:47:14.629 to pivot to now is essentially talking about what I 0:47:14.630 --> 0:47:18.590 think the solution is and what I think is inevitable. Okay, 0:47:18.630 --> 0:47:21.110 so this is kind of getting to why I think 0:47:21.110 --> 0:47:23.590 this is going to happen and why I think people 0:47:23.589 --> 0:47:25.989 need to prepare and get ready. So all of that 0:47:25.989 --> 0:47:28.390 was sort of a setup to what I think is 0:47:28.390 --> 0:47:31.229 about to happen. And this is an architecture that I've 0:47:31.230 --> 0:47:33.830 come up with. I call it the lattice architecture. I've 0:47:33.830 --> 0:47:36.230 been working on this for quite a while. I'm going 0:47:36.230 --> 0:47:39.190 to release it as an open source project as well. 0:47:39.190 --> 0:47:41.870 It's the same system that I use with customers to 0:47:41.910 --> 0:47:46.109 basically build their back end systems to help them solve 0:47:46.110 --> 0:47:48.350 the stuff that we talked about on the company side. 0:47:48.350 --> 0:47:50.910 But the system itself, I'm actually going to release, just 0:47:50.910 --> 0:47:53.150 like I did with Pi, I'm going to release this 0:47:53.150 --> 0:47:56.870 as kind of a sister project for actually helping companies, 0:47:56.910 --> 0:48:01.150 actually any type of entity basically solve that chaos mess. 0:48:01.190 --> 0:48:06.710 So essentially, the concept is addressing this transparency problem, addressing 0:48:06.710 --> 0:48:10.019 the fact that all these different groups are opaque. The 0:48:10.060 --> 0:48:14.460 work is opaque. The processes, the SOPs are not clear. 0:48:14.460 --> 0:48:18.100 So what Latisse does is that it actually has a 0:48:18.100 --> 0:48:22.019 single unified daemon that everything reports up into. And what 0:48:22.020 --> 0:48:24.819 you have are the different tiers of organization. You have 0:48:24.819 --> 0:48:28.300 the company level, you have like a unified context of 0:48:28.299 --> 0:48:31.700 documents and processes and SOPs and everything, all the different 0:48:31.739 --> 0:48:35.060 knowledge for a particular company in a unified place. And 0:48:35.060 --> 0:48:38.020 then you have a set of things for each company. 0:48:38.020 --> 0:48:41.460 So you have SOPs, you have metrics, you have work, 0:48:41.500 --> 0:48:45.060 you have goals, you have budget, you have all these 0:48:45.060 --> 0:48:48.779 different components of a particular entity all the way from 0:48:48.779 --> 0:48:51.580 the company level. But you also have that same exact 0:48:51.580 --> 0:48:55.140 thing for the department. You also have it for a team, 0:48:55.140 --> 0:48:58.860 and you also have it for individual people working inside 0:48:58.860 --> 0:49:01.779 the company as a worker. So what this does is 0:49:01.779 --> 0:49:04.779 it takes a system like pie and you can have 0:49:04.779 --> 0:49:07.499 your own version. People will have their own versions. This 0:49:07.500 --> 0:49:09.489 is the one that I use. And again, it's open 0:49:09.489 --> 0:49:11.850 source project. You can go and check out. And what 0:49:11.850 --> 0:49:15.410 it is, is as an individual, you will have your 0:49:15.529 --> 0:49:18.530 personal SOPs the way that you like to get things 0:49:18.529 --> 0:49:22.370 done that you follow. You will have your personal budget 0:49:22.410 --> 0:49:24.689 of what you spend things on. You will have your 0:49:24.690 --> 0:49:28.050 own personal metrics, which maybe you got from management or whatever. Actually, 0:49:28.049 --> 0:49:29.850 you're about to see where you got it from. You 0:49:29.850 --> 0:49:33.729 have all these different components. You have your knowledge in skills, 0:49:33.730 --> 0:49:36.969 you have all these capabilities that you have when you 0:49:37.009 --> 0:49:39.489 learn something new or whatever, you capture it into the 0:49:39.489 --> 0:49:45.450 system and most importantly, your individual unit. You are broadcasting APIs. 0:49:45.730 --> 0:49:49.689 You as a person, you, Sarah, you, Ravi, you, Chris, 0:49:49.730 --> 0:49:53.129 you are broadcasting what your SOPs are. You are broadcasting 0:49:53.130 --> 0:49:56.290 what your metrics are. You are broadcasting what your budget is. 0:49:56.290 --> 0:50:00.610 You're broadcasting your work items. They are available for any 0:50:00.650 --> 0:50:04.490 agent or any other team member or the agent for 0:50:04.489 --> 0:50:07.200 the team or the agent for the department or the 0:50:07.200 --> 0:50:10.920 agent for the company can ping these things. Can query 0:50:10.960 --> 0:50:14.360 you and actually pull the stuff in real time. Most importantly, 0:50:14.360 --> 0:50:17.919 this allows the CEO and the CFO actually to be 0:50:17.960 --> 0:50:20.959 able to look down and say, okay, this is the 0:50:20.960 --> 0:50:22.919 work that we're doing. Okay, so you're going to have 0:50:22.920 --> 0:50:26.919 workflows also documented in this exact same way. I want 0:50:26.960 --> 0:50:29.160 to show you some other diagrams here. Okay, so here's 0:50:29.160 --> 0:50:33.600 another way of looking at it. So again, goals, plans, SOPs, metrics, 0:50:33.600 --> 0:50:36.200 work cost quality. The thing I want to express to 0:50:36.200 --> 0:50:38.600 you here, like we talked about already, so you probably 0:50:38.600 --> 0:50:40.880 already have this point. So I won't belabor it too much. 0:50:40.880 --> 0:50:46.160 These things goals, plans, SOPs, metrics, work, cost quality, these things, 0:50:46.160 --> 0:50:50.000 the workflows for the company. This stuff is extremely opaque 0:50:50.040 --> 0:50:52.680 to the leaders of a company. This is extremely hard 0:50:52.680 --> 0:50:56.400 to gather and this is going to turn it into seconds. 0:50:56.400 --> 0:50:59.960 It's going to turn it into less than a second oftentimes, right? 0:51:00.000 --> 0:51:03.479 It's going to turn it into minutes at the absolute maximum. 0:51:03.480 --> 0:51:05.910 Like if you have a super large company, right? And 0:51:05.910 --> 0:51:08.029 it depends on the implementation. Like I want to give 0:51:08.029 --> 0:51:10.709 actual numbers or anything. That's just ridiculous. It depends on 0:51:10.710 --> 0:51:14.629 the architecture that you use. But the point is transparency. 0:51:14.670 --> 0:51:17.110 First of all, and the other core point is that 0:51:17.110 --> 0:51:21.789 each layer from the worker to the team to the department, 0:51:21.790 --> 0:51:26.470 to the company is independently authoritative over its stuff. So 0:51:26.630 --> 0:51:29.390 one of the problems that I see tons with customers 0:51:29.509 --> 0:51:33.070 is they have no idea what their other departments are doing, right. 0:51:33.110 --> 0:51:36.469 They have no visibility into how they're interacting with the customer, 0:51:36.469 --> 0:51:39.070 what they've said to the customer, what's been done with them. 0:51:39.150 --> 0:51:42.629 You can also have context silos like this for each 0:51:42.630 --> 0:51:46.109 customer inside of a customer container, right? And you can 0:51:46.110 --> 0:51:49.029 have other agents because I'm a security person, there's going 0:51:49.029 --> 0:51:52.310 to be another agent flow here, basically a gateway that 0:51:52.310 --> 0:51:55.870 basically says, okay, which agents can ask which questions about 0:51:55.870 --> 0:51:59.390 what context zones, right? So there's going to be permissions. 0:51:59.390 --> 0:52:01.270 You've got identity stuff, you've got a whole bunch of 0:52:01.270 --> 0:52:03.890 stuff that's all being worked on. But the core concept 0:52:03.890 --> 0:52:08.330 of lattice is visibility up and down and across. I 0:52:08.330 --> 0:52:10.410 could just query, first of all, I'm just I'm not 0:52:10.410 --> 0:52:13.410 going to be actually running any APIs myself. I just 0:52:13.410 --> 0:52:16.650 talk to my agent system as an individual worker on 0:52:16.650 --> 0:52:18.850 the front end team over here. And I say, hey, 0:52:18.850 --> 0:52:21.330 what are my coworkers working on? I'm thinking about working 0:52:21.330 --> 0:52:23.770 on this. And they're like, oh, yeah, you know, Julie's 0:52:23.770 --> 0:52:26.089 already working on that. You should talk to, you know, 0:52:26.130 --> 0:52:28.850 Modi or whatever. Yeah, hit them up. They're already working 0:52:28.850 --> 0:52:30.450 on it. Maybe you can do a collab. Do you 0:52:30.450 --> 0:52:33.370 want me to set up a meeting? Boom. Okay, cool. Well, 0:52:33.410 --> 0:52:35.970 now you just posted that you're in a meeting now 0:52:35.969 --> 0:52:38.969 it just posted that you're going to collaborate about this 0:52:38.969 --> 0:52:41.930 particular feature that could go into your API. It's now 0:52:41.930 --> 0:52:45.289 available when someone asks, hey, is anyone working on this? Well, 0:52:45.290 --> 0:52:48.410 now it includes you that you're actually also working on this. 0:52:48.450 --> 0:52:52.049 It's now available to other teams, other departments, all the 0:52:52.049 --> 0:52:54.769 way up to the company level. And this unified sort 0:52:54.770 --> 0:52:58.569 of context layer here, the lattice daemon is where you 0:52:58.569 --> 0:53:01.049 can query and get it for everyone. Now I want 0:53:01.089 --> 0:53:03.600 to talk about this whole concept of the SOPs and 0:53:03.600 --> 0:53:07.000 the goals. Think of what happens right now in a company. 0:53:07.000 --> 0:53:08.719 I didn't talk about this too much in the sort 0:53:08.719 --> 0:53:12.039 of negative state. When you go to create goals and 0:53:12.040 --> 0:53:14.319 change your goals, how many people are triggered by the 0:53:14.319 --> 0:53:16.920 word OKR? For a good reason. The reason is because 0:53:16.920 --> 0:53:19.439 they change all the time. People have different systems for 0:53:19.440 --> 0:53:23.479 doing them. Imagine a giant all hands meeting where the 0:53:23.759 --> 0:53:27.040 CEO gets together with everyone and says, hey, check it out. 0:53:27.040 --> 0:53:29.320 These are the new goals. This is what we're doing. 0:53:29.319 --> 0:53:31.560 Here are our new metrics that go with the goals. 0:53:31.560 --> 0:53:34.400 By the way, we've got some new SOPs we just published. Boom. 0:53:34.440 --> 0:53:36.999 Guess what they say. They have now been published. All 0:53:37.000 --> 0:53:40.600 documentation is now updated, goes all the way down. Updates 0:53:40.600 --> 0:53:44.560 the documentation, updates the goals. Now inside of the goals 0:53:44.560 --> 0:53:47.759 of each group. That content is now updated. We're talking 0:53:47.759 --> 0:53:50.560 about text files here. This is not like magic stuff. 0:53:50.560 --> 0:53:52.839 This is not difficult stuff. We are doing much more 0:53:52.839 --> 0:53:55.799 difficult stuff with AI already. This is just not a 0:53:55.799 --> 0:53:58.279 system that has been built yet and it will take 0:53:58.319 --> 0:54:00.520 time to roll out. Some companies will be able to 0:54:00.520 --> 0:54:03.830 adopt this very quickly. Some companies will take, you know, 0:54:03.870 --> 0:54:05.910 five years or ten years or more, or they'll just 0:54:05.910 --> 0:54:08.150 die off because they couldn't get there fast enough. And 0:54:08.150 --> 0:54:10.669 to be clear, this is one architecture. This is the 0:54:10.670 --> 0:54:15.109 architecture that I'm using. But the concept is actually more important. 0:54:15.150 --> 0:54:19.190 The functions that this architecture allows is the most important piece. 0:54:19.190 --> 0:54:23.229 And I think the broadcasting of the demon, the broadcasting 0:54:23.230 --> 0:54:27.550 of the APIs for each tier level, for each object 0:54:27.589 --> 0:54:31.390 type within the company. That is what makes this all work, right. 0:54:31.430 --> 0:54:34.149 So the protocols that we use, who knows? Like it's 0:54:34.150 --> 0:54:36.269 hard to predict exactly how it's going to play out. 0:54:36.310 --> 0:54:39.029 Maybe it's going to be something better than HTTP. Maybe 0:54:39.029 --> 0:54:41.549 it's going to be something better than rest. It doesn't matter. 0:54:41.549 --> 0:54:44.469 The concept is what I want you to think about here. Okay, 0:54:44.509 --> 0:54:46.870 so here's another view of this. And just another way 0:54:46.870 --> 0:54:48.469 to think about it. I've got a whole bunch of 0:54:48.469 --> 0:54:51.549 these diagrams, by the way, that I put together over 0:54:51.589 --> 0:54:53.589 quite a decent amount of time, and I'm going to 0:54:53.589 --> 0:54:55.749 make them all available there in the blog post. They're 0:54:55.750 --> 0:54:58.310 also in, I think the link is in the description 0:54:58.310 --> 0:55:00.379 to the video as well. But look at this. All 0:55:00.380 --> 0:55:03.540 these different queries can go up to the latest statement. 0:55:03.540 --> 0:55:06.739 You basically post your status. It could be push or pull, right? 0:55:06.779 --> 0:55:09.780 A combination of the both. But again you have the tiers. 0:55:09.779 --> 0:55:13.459 You have goals, budget, plans for the department container, the 0:55:13.460 --> 0:55:16.819 team container, the individual node, right? And they all are 0:55:16.819 --> 0:55:20.380 speaking to one central place, but they're all able to 0:55:20.580 --> 0:55:23.540 post to each other and or receive from each other. 0:55:23.540 --> 0:55:26.499 And the other thing that's part of different diagrams is 0:55:26.500 --> 0:55:30.060 essentially a big unified context as well. Like kind of 0:55:30.100 --> 0:55:32.500 like a context data lake, which I've talked about in 0:55:32.500 --> 0:55:35.379 previous videos. Okay, so here's another view. And this is 0:55:35.380 --> 0:55:38.500 absolutely critical. This is huge. A big part of the 0:55:38.500 --> 0:55:41.500 problem that we have is not being able to see 0:55:41.660 --> 0:55:45.100 the work that we're doing. We don't understand the workflows. 0:55:45.100 --> 0:55:48.100 We don't understand who's doing which part. So I've got 0:55:48.100 --> 0:55:51.100 another video where I talked about companies are just graphs 0:55:51.100 --> 0:55:54.859 of algorithms. Okay? This is a really core concept. The 0:55:54.860 --> 0:55:58.330 idea that someone's doing like insurance claims or whatever. Well, 0:55:58.330 --> 0:56:01.009 it's not wizardry. They don't make it up each time. 0:56:01.009 --> 0:56:04.130 They should be following a set of steps, right? But 0:56:04.130 --> 0:56:06.290 it's not clear like we talked about in the beginning. 0:56:06.290 --> 0:56:08.570 It's not clear what those steps are. It's not clear 0:56:08.569 --> 0:56:11.330 if they're being followed correctly. It's not clear exactly what 0:56:11.330 --> 0:56:13.689 is happening inside the company. There's not even a list 0:56:13.690 --> 0:56:15.969 of them. And like I said, it's actually really difficult 0:56:15.969 --> 0:56:18.890 to go and collect a list. So look at this here. 0:56:18.890 --> 0:56:21.169 This is the type of view that somebody is going 0:56:21.210 --> 0:56:23.610 to have. They're going to have the ability, anyone, whoever 0:56:23.610 --> 0:56:26.489 has access, right, with the security permissions to be able 0:56:26.489 --> 0:56:27.930 to look at this stuff and they're going to be 0:56:27.930 --> 0:56:31.450 able to say, okay, well, we still got human components here. 0:56:31.450 --> 0:56:35.049 We got automated components here. Okay, we got review over here. Okay, 0:56:35.089 --> 0:56:37.649 we've got audit over here. We've got a quality check 0:56:37.650 --> 0:56:40.850 over here. This is the type of view that CEOs, 0:56:40.890 --> 0:56:44.890 CFOs like senior management, really any manager really wants to 0:56:44.890 --> 0:56:47.009 be able to see because you can't optimize what you 0:56:47.009 --> 0:56:50.090 don't understand. You can't optimize what you don't see. Right. 0:56:50.130 --> 0:56:53.129 All right. So this is actually another really cool aspect 0:56:53.130 --> 0:56:56.689 of once you have those visuals, once you have those workflows, 0:56:56.830 --> 0:57:00.270 things just fundamentally change. You know how like someone comes up? 0:57:00.310 --> 0:57:02.189 You know, they take you to a steak dinner, they 0:57:02.190 --> 0:57:04.189 meet you for some meeting or whatever, and they're like, hey, 0:57:04.230 --> 0:57:08.350 you know, we have the best image background removal process. 0:57:08.350 --> 0:57:10.350 I don't know what you're using for that today, but 0:57:10.350 --> 0:57:12.629 ours is way better. And so it's like, okay, well, 0:57:12.670 --> 0:57:14.510 I guess we'll do a pilot. You know, I really 0:57:14.509 --> 0:57:16.870 liked the steak. So I guess we'll do a pilot 0:57:16.870 --> 0:57:18.830 and I'll tell part of my team, you know, you're 0:57:18.830 --> 0:57:21.550 going to spend six weeks coming up with test criteria 0:57:21.550 --> 0:57:23.230 and we're going to test this thing and a B 0:57:23.270 --> 0:57:25.669 test or whatever. And if they like it, you know, 0:57:25.710 --> 0:57:27.829 I guess we'll take a look at it that goes 0:57:27.830 --> 0:57:31.430 away when you actually have a lattice over here of 0:57:31.470 --> 0:57:34.950 like all the different pieces that make up your company. 0:57:34.950 --> 0:57:37.870 These are the different processes, these are the different components. 0:57:37.870 --> 0:57:40.230 This is the tool, this is the decision. This is 0:57:40.230 --> 0:57:43.830 the human making the decision. This is an AI making decision. 0:57:43.830 --> 0:57:46.870 This is an AI processing work. This is the graph 0:57:46.870 --> 0:57:51.510 of algorithms that makes up the company. One particular node, 0:57:51.510 --> 0:57:54.710 this little node right here, one of hundreds or thousands 0:57:54.710 --> 0:57:57.820 or tens of thousands or millions of nodes inside your company. 0:57:57.820 --> 0:58:03.540 Is the process the algorithm for doing image background removal. Today, 0:58:03.540 --> 0:58:06.300 we don't have metrics on that. Really, most people don't. 0:58:06.340 --> 0:58:08.180 What do we have? We have like, yeah, it works 0:58:08.180 --> 0:58:10.499 pretty good. I like it, yeah, it's pretty good. Maybe 0:58:10.500 --> 0:58:12.260 we could pull some kind of metrics, but it's not 0:58:12.260 --> 0:58:15.019 like really good what we're going to have, what we're 0:58:15.020 --> 0:58:18.620 moving towards and most importantly, what people want to get to, 0:58:18.660 --> 0:58:21.820 what leadership wants to get to is all of these 0:58:21.820 --> 0:58:25.860 have metrics for them. They're all clearly defined. So now 0:58:26.100 --> 0:58:30.660 a vendor doesn't come with a steak dinner. They cannot anymore. Okay. 0:58:30.700 --> 0:58:34.499 Because we're going from wizardry to excel, okay? We show 0:58:34.500 --> 0:58:36.820 them our metrics. We show them, hey, look, this is 0:58:36.820 --> 0:58:40.420 how we do things. This is our current lattice structure. 0:58:40.420 --> 0:58:44.580 This is our current flow structure. This is image background removal. 0:58:44.580 --> 0:58:46.499 This is how much it costs us. This is how 0:58:46.500 --> 0:58:49.260 good it is. This is our quality ratings according to 0:58:49.300 --> 0:58:51.980 so and so standard. What are your ratings? What are 0:58:51.980 --> 0:58:55.450 your metrics? What are your cost numbers for this? If 0:58:55.450 --> 0:58:57.450 they can't bring that and all they brought was the 0:58:57.490 --> 0:59:02.050 steak dinner, that's not a conversation anymore. Once this transition happens, right? 0:59:02.090 --> 0:59:06.890 This is absolutely essential. Absolutely essential. Essentially, it's turning the 0:59:06.890 --> 0:59:11.530 company from these sort of black box, sort of vibey 0:59:11.650 --> 0:59:14.690 things that are happening that hopefully work out pretty well 0:59:14.690 --> 0:59:21.290 into actual pipelines, actual workflows with metrics and reliability and 0:59:21.290 --> 0:59:23.890 testing and all this sort of stuff as part of 0:59:23.890 --> 0:59:25.930 the system. All right. Real quick, this is just a 0:59:25.930 --> 0:59:28.210 little teaser. I'm not going to talk too much about this. 0:59:28.210 --> 0:59:30.810 This is a screenshot that I took earlier, I think 0:59:30.850 --> 0:59:34.090 last week. And this is actually just an example of 0:59:34.090 --> 0:59:37.930 the lattice system working. And it's actually one agent system 0:59:37.930 --> 0:59:41.930 reporting up into the lattice daemon basically describing its work. 0:59:41.930 --> 0:59:44.970 So this is a particular system called the algorithm. And 0:59:44.970 --> 0:59:47.209 it's running, I guess if you know, you know, but 0:59:47.210 --> 0:59:50.050 it's part of Pi and it runs work. This is 0:59:50.050 --> 0:59:52.209 the type of thing I'm just trying to show. This 0:59:52.210 --> 0:59:55.200 is not a theoretical thing. It is quite easy. And 0:59:55.200 --> 0:59:57.600 if you're an experienced user of AI, you already know 0:59:57.600 --> 0:59:59.280 you could do this type of stuff. This is the 0:59:59.280 --> 1:00:02.200 type of dashboard that line managers want to be able 1:00:02.200 --> 1:00:04.600 to see, that engineering managers want to be able to 1:00:04.600 --> 1:00:07.240 see that anybody should be able to drill into and 1:00:07.240 --> 1:00:10.320 actually see the work that's proceeding through the system. And 1:00:10.320 --> 1:00:14.120 this is making simple Rest calls and pushing this content 1:00:14.120 --> 1:00:17.559 up into the system and giving you a dashboard like this. 1:00:17.600 --> 1:00:19.680 All right. Essentially, what I want you to have up 1:00:19.680 --> 1:00:22.120 to this point is sort of dispelling some of the 1:00:22.120 --> 1:00:25.519 myths about how perfect humans are. We are not perfect. 1:00:25.520 --> 1:00:28.360 I'm not trying to talk crap about anybody. Some of 1:00:28.360 --> 1:00:31.439 my best friends are actually humans. So like, I'm not 1:00:31.440 --> 1:00:34.600 trying to throw stones here, right? I'm trying to describe 1:00:34.720 --> 1:00:38.080 the reality that's actually happening according to my view. Like 1:00:38.120 --> 1:00:40.480 you could watch all this and still think I'm wrong, 1:00:40.480 --> 1:00:43.040 but hopefully you won't. So where is this taking us? 1:00:43.040 --> 1:00:45.800 It seems like this is just horribly bad news. So 1:00:45.800 --> 1:00:47.880 what I want you to do have had so far 1:00:48.000 --> 1:00:51.309 is essentially the current state of work. The current state 1:00:51.310 --> 1:00:54.150 of human work, the current state of companies is not good. 1:00:54.150 --> 1:00:58.189 The perfect human thinking machine is not correct, right? The 1:00:58.190 --> 1:01:01.830 way that we actually process information, store information, it is 1:01:01.830 --> 1:01:05.070 much more like an AI than is actually comfortable to 1:01:05.110 --> 1:01:08.749 think about. And all that combined basically means the bar 1:01:08.750 --> 1:01:11.310 is actually extremely low. And then when you add on 1:01:11.310 --> 1:01:14.550 top of that, this lattice system, which is just one 1:01:14.550 --> 1:01:17.910 implementation that one person came up with, I'm telling you 1:01:18.110 --> 1:01:20.870 that is the direction that things are going. This is 1:01:20.870 --> 1:01:24.789 the architecture that companies are moving to that they are 1:01:24.790 --> 1:01:27.350 going to have. A lot of them haven't figured it 1:01:27.350 --> 1:01:30.350 out yet. Like this is still very early stuff, but 1:01:30.350 --> 1:01:32.910 people will catch on to this very soon and they 1:01:32.910 --> 1:01:36.229 will start building this type of architecture. And it's going 1:01:36.270 --> 1:01:39.230 to happen fast. It's going to happen extremely fast. Now, 1:01:39.470 --> 1:01:41.990 one other point I want to mention about this is 1:01:42.150 --> 1:01:45.070 a way to think about automation in a completely different way. 1:01:45.070 --> 1:01:47.310 If you think about it in terms of like, okay, 1:01:47.350 --> 1:01:50.980 we have the humans. Now let's think about adding AI. 1:01:51.180 --> 1:01:54.420 Oh my God, that's crazy. I have the same instinct. Okay, 1:01:54.460 --> 1:01:58.860 it feels icky because they're coming into our processes. Here's 1:01:58.860 --> 1:02:01.340 another way to think about this. Imagine you have a 1:02:01.340 --> 1:02:06.060 factory production line that produces iPhones. So it's soldering two 1:02:06.300 --> 1:02:11.180 nanometer transistors for iPhones. And this thing is amazing. It's 1:02:11.180 --> 1:02:13.900 producing whatever, $1 million a minute. Who knows what the 1:02:13.900 --> 1:02:16.100 numbers are, but it's just fully automated. It's a whole 1:02:16.100 --> 1:02:19.500 bunch of robots and it's just flowing perfectly. Who would say, 1:02:19.540 --> 1:02:21.580 you know what this thing really needs? It lacks a 1:02:21.580 --> 1:02:24.060 human touch. It lacks a human touch. We need to 1:02:24.060 --> 1:02:26.740 get people need jobs. People need jobs. We need to 1:02:26.740 --> 1:02:30.499 get some humans in here because AI's don't have the 1:02:30.500 --> 1:02:33.260 right thinking. They're just not as good as humans at 1:02:33.260 --> 1:02:36.220 doing work. Do you have any idea what would happen 1:02:36.780 --> 1:02:40.260 if you tried to ask a human to solder two 1:02:40.300 --> 1:02:44.300 nanometer transistors onto circuit boards? It would not go well. 1:02:44.340 --> 1:02:47.900 Also known as impossible. Okay. Another one which I think 1:02:47.940 --> 1:02:51.160 can't remember who mentioned this. Somebody mentioned an example of 1:02:51.160 --> 1:02:54.560 a amazing Excel sheet. We've all seen one of these before. 1:02:54.600 --> 1:02:57.520 It's got a million tabs. Looks like an actual application. 1:02:57.520 --> 1:03:00.320 The thing is just fantastic. Maybe it handles the entire 1:03:00.320 --> 1:03:04.480 company's finances and just produces the best results. It never 1:03:04.480 --> 1:03:07.600 messes up. It's just awesome. It's like, you know, 36 1:03:07.600 --> 1:03:10.919 months of work goes into this crown jewel of an 1:03:10.920 --> 1:03:14.400 Excel spreadsheet, and it's just processing millions and millions of 1:03:14.400 --> 1:03:18.880 lines for the entire enterprise globally. Where on the spreadsheet 1:03:19.000 --> 1:03:22.360 should we add a human to help the spreadsheet? Where 1:03:22.360 --> 1:03:26.520 do we add it when automation works, when AI workflows work, 1:03:26.520 --> 1:03:29.600 when the lattice system that I was showing you works, 1:03:29.600 --> 1:03:33.240 the answer will not be to add humans to the thing. 1:03:33.240 --> 1:03:37.960 It becomes extremely obvious that adding humans actually makes it worse. Okay, 1:03:38.000 --> 1:03:40.360 with an exception, which we're going to talk about. So 1:03:40.520 --> 1:03:44.480 think about that inversion when you're considering the uncomfortable feeling 1:03:44.480 --> 1:03:47.590 that you're getting when you're thinking about adding automation, think 1:03:47.590 --> 1:03:51.310 of the opposite. The default is opposite automation first. And 1:03:51.310 --> 1:03:53.709 there's another reason to think that one saying that I 1:03:53.710 --> 1:03:56.150 have that I think is very, very true is the 1:03:56.150 --> 1:03:59.830 ideal number of human employees for a company is zero. Now, 1:03:59.830 --> 1:04:02.229 I'm not talking about small boutique startups. I'm not talking 1:04:02.230 --> 1:04:05.310 about collaborations with your friends. I'm not talking about like 1:04:05.350 --> 1:04:07.630 sole projects that you just love to do. You want 1:04:07.630 --> 1:04:10.229 to work with other humans? I do too, that's always 1:04:10.230 --> 1:04:13.070 going to be there. I'm talking about for large companies, 1:04:13.070 --> 1:04:16.550 medium sized companies, companies that need to do things at scale. 1:04:16.550 --> 1:04:21.630 These companies, again, they are paying $50 trillion in knowledge 1:04:21.630 --> 1:04:25.750 work compensation. They want to do that work internally. Okay. 1:04:25.990 --> 1:04:28.870 If I had an ice cream truck business and all 1:04:28.870 --> 1:04:30.990 I had was my ice cream truck, and I had 1:04:30.990 --> 1:04:33.750 a machine that could actually make the perfect ice cream cone. 1:04:33.750 --> 1:04:37.430 And maybe I have three trucks and I'm the only employee. 1:04:37.430 --> 1:04:40.070 And the trucks, they have X number of machines. They 1:04:40.070 --> 1:04:42.070 can pump out the things. I can hand them off 1:04:42.070 --> 1:04:44.990 to the kids with a lollipop or whatever, and I'm 1:04:44.990 --> 1:04:48.060 making X amount of money and I'm happy. I'm just happy. 1:04:48.060 --> 1:04:50.260 This is a great amount of money. That is the 1:04:50.260 --> 1:04:52.620 state that companies want to be in. Of course they 1:04:52.620 --> 1:04:54.979 want to grow, right? So they'll want more and more 1:04:54.980 --> 1:04:57.780 bigger trucks, you know, more stuff like that. But they 1:04:57.780 --> 1:05:01.180 won't say, I need more humans to add to this. 1:05:01.220 --> 1:05:03.459 If I'm sitting here with my ice cream truck and 1:05:03.460 --> 1:05:05.939 I've got my ice cream cone makers, and I am 1:05:05.940 --> 1:05:08.460 perfectly happy with the amount of money that I'm making, 1:05:08.460 --> 1:05:10.459 if a group of people come and they pick at 1:05:10.460 --> 1:05:13.900 me and they're like, you're against humans. Humans need jobs, 1:05:13.900 --> 1:05:16.580 humans need jobs. I'm like, what are you talking about? 1:05:16.580 --> 1:05:19.220 I got my truck. I've got my ice cream cone makers. 1:05:19.220 --> 1:05:23.460 It's just me. This is literally me. Those machines are me. 1:05:23.460 --> 1:05:25.939 And this ice cream truck is me. This, as a 1:05:25.940 --> 1:05:30.140 unit is me by myself doing the work. Well, guess what? 1:05:30.140 --> 1:05:32.499 That diagram I showed you of the lattice system of 1:05:32.500 --> 1:05:35.540 the entire company. Let's just say nothing in there was human, 1:05:35.540 --> 1:05:38.220 which that'll be forever before that happens. Or a long 1:05:38.220 --> 1:05:41.140 time anyway. Let's just say it's fully automated. Everything in 1:05:41.140 --> 1:05:43.700 the whole lattice structure, it's all automated, or let's say 1:05:43.700 --> 1:05:47.330 it's a small company or whatever. Fully automated. Right? Fully automated. 1:05:47.330 --> 1:05:51.050 It's working great. That is their company. They consider that 1:05:51.050 --> 1:05:54.249 to be them doing the work. All that infrastructure, all 1:05:54.290 --> 1:05:57.770 that automation, all the machines, all the tech. It is 1:05:57.770 --> 1:06:00.450 part of the company now. It's like the walls and 1:06:00.450 --> 1:06:03.610 the floors and the ceilings and the chairs and the desks. 1:06:03.610 --> 1:06:06.330 It's literally just part of the core of the company 1:06:06.330 --> 1:06:09.730 that is the preferred mode. They are currently paying $50 1:06:09.730 --> 1:06:14.209 trillion for outside help. They want that help to be inside. 1:06:14.210 --> 1:06:17.209 They want it to be all internal, okay? That is 1:06:17.210 --> 1:06:20.250 the desired state. And going back to the whole setup 1:06:20.250 --> 1:06:23.410 with like the factory and adding humans, one argument I've 1:06:23.410 --> 1:06:26.410 heard is like, well, yeah, but if you want to scale, right? 1:06:26.450 --> 1:06:29.249 That's there's no problem. You'll just hire more humans. Like 1:06:29.250 --> 1:06:32.450 why wouldn't you hire more humans? Because humans can also 1:06:32.450 --> 1:06:35.090 do the job. Well, no, they can't do the job 1:06:35.090 --> 1:06:38.410 as good as it's already being done by the automation. Right? 1:06:38.450 --> 1:06:41.570 So that's where that argument breaks down. You don't add 1:06:41.570 --> 1:06:45.040 humans to a working system that doesn't have humans in it, 1:06:45.040 --> 1:06:47.920 because the working system without humans in it is likely 1:06:47.920 --> 1:06:51.200 to be many times better and many times faster for 1:06:51.200 --> 1:06:53.999 a lot of tasks, maybe most tasks. So just keep 1:06:54.000 --> 1:06:56.200 that in mind. All right, so here we are. I 1:06:56.200 --> 1:06:58.560 promise some good news and this is where we get 1:06:58.560 --> 1:07:00.600 to it. So we've talked about the bad news. We've 1:07:00.640 --> 1:07:03.440 talked about what the problems are and what I think 1:07:03.440 --> 1:07:07.240 the inevitable solution is. And I do believe that is true. 1:07:07.240 --> 1:07:09.999 Now let's talk about, okay, what are humans supposed to do? 1:07:10.000 --> 1:07:12.680 What is the good news? This sounds like horribly bad 1:07:12.680 --> 1:07:16.280 news with no possible upside. All the humans are screwed. 1:07:16.280 --> 1:07:18.320 What are we going to do? And for that, I 1:07:18.320 --> 1:07:21.360 want to bring you over to this content here. All right, 1:07:21.360 --> 1:07:23.200 so for that I want to bring you over to 1:07:23.240 --> 1:07:26.280 this content here. So remember when we talked about the 1:07:26.280 --> 1:07:31.999 four layers and we said, okay, knowledge, understanding, intelligence and 1:07:32.000 --> 1:07:35.480 creativity and we said, look, this is the stuff that 1:07:35.520 --> 1:07:39.240 humans can do. But there were two layers underneath. And 1:07:39.240 --> 1:07:42.660 these are the layers that eyes don't have. Okay. They 1:07:42.660 --> 1:07:48.380 don't have subjective experience. They don't have desires to do anything. Okay, 1:07:48.420 --> 1:07:53.020 so imagine that whole workflow again, fully automated. It's amazing. 1:07:53.020 --> 1:07:56.100 And it doesn't need any humans because it's just better 1:07:56.100 --> 1:08:00.020 and faster and more consistent, higher quality, all of that 1:08:00.020 --> 1:08:03.700 cheaper way, way cheaper. Fine. Who decides what to make? 1:08:03.740 --> 1:08:06.380 Who decides what company to make in the first place? 1:08:06.380 --> 1:08:09.100 Here's what AI doesn't have, and this should be a 1:08:09.100 --> 1:08:12.300 source of hope. Hopefully what I've done here is show 1:08:12.300 --> 1:08:15.740 you how bad one part of the situation is, and 1:08:15.940 --> 1:08:19.140 it should be equally clear. Hopefully, after what I'm about 1:08:19.140 --> 1:08:22.179 to show you that the opposite is also true. What 1:08:22.180 --> 1:08:25.460 this unlocks for humans is actually what we should have 1:08:25.460 --> 1:08:29.420 been doing all along. We should not be the row 1:08:29.420 --> 1:08:32.380 in the Excel sheet. We should not be the section 1:08:32.380 --> 1:08:36.100 of the robot army that you know solders the transistor. 1:08:36.100 --> 1:08:39.140 We should not be the one in the car dreading 1:08:39.180 --> 1:08:42.370 going into the office Monday morning that all should never 1:08:42.370 --> 1:08:46.290 have happened. Which, like I said, we knew that before 2022. 1:08:46.330 --> 1:08:49.570 It's totally understandable to freak out when the jobs are 1:08:49.570 --> 1:08:52.130 going away and AI is taking them. The natural tendency 1:08:52.130 --> 1:08:54.450 of someone yanx something out of your arms is to 1:08:54.490 --> 1:08:58.090 pull it closer, right? Understandable. Like not mad at anyone 1:08:58.090 --> 1:09:00.370 for that. I'm just trying to help you sort of 1:09:00.410 --> 1:09:04.650 think through this. That work was garbage. It is garbage. 1:09:04.650 --> 1:09:07.930 It was garbage for decades. It still is garbage. And 1:09:07.970 --> 1:09:10.370 AI's are coming to take it. Fine. The place that 1:09:10.370 --> 1:09:12.810 we're going, the place that we need to get to, 1:09:12.850 --> 1:09:16.770 is where humans actually have ideas. People are actually broadcasting 1:09:16.770 --> 1:09:19.610 their own capabilities and their own plans and their own 1:09:19.610 --> 1:09:22.770 creativity and their own beliefs and their own ideas and 1:09:22.770 --> 1:09:28.250 their own creations. Nurturing people, helping people grow, creating new art, 1:09:28.290 --> 1:09:31.650 new experiences for other people. And guess what? Coming up 1:09:31.650 --> 1:09:34.850 with companies to do this right. How will the company 1:09:34.850 --> 1:09:37.930 be implemented? How will all the work tasks get done 1:09:37.970 --> 1:09:40.800 using a system similar to Latisse. It's going to be 1:09:40.800 --> 1:09:44.080 an all AI system doing the actual crunching of the work. 1:09:44.080 --> 1:09:46.920 But what can the AI not do? It doesn't have 1:09:46.920 --> 1:09:50.200 a desire. It doesn't have any goals. It doesn't have ambitions. 1:09:50.240 --> 1:09:53.000 AI doesn't see a problem in the world and go, yeah, 1:09:53.000 --> 1:09:54.920 I just don't like that. I just don't like that. 1:09:54.960 --> 1:09:57.000 You know, that's not the way they should be treated. 1:09:57.040 --> 1:09:59.760 You know, this is a horrible experience. I don't like 1:09:59.760 --> 1:10:01.960 this product. We should make a better one. AI is 1:10:01.960 --> 1:10:04.720 not doing that. Why? Because it doesn't experience anything. It 1:10:04.720 --> 1:10:08.160 doesn't experience problems and feel bad. It doesn't come up 1:10:08.160 --> 1:10:10.440 with new ideas. If you ask it to come up 1:10:10.440 --> 1:10:12.680 with a new idea, it can. But it's not sitting 1:10:12.680 --> 1:10:16.960 around dormant in like unhappy. We are unhappy. We have 1:10:16.960 --> 1:10:21.960 these desires and we have this subjective experience for another reason, right? 1:10:22.000 --> 1:10:25.160 It's because we are powered by evolution. This is kind 1:10:25.160 --> 1:10:28.040 of a big idea as well. Might be slightly controversial, 1:10:28.040 --> 1:10:30.040 I think if you're watching the channel, might not be 1:10:30.040 --> 1:10:33.760 as controversial as it is to most people, but evolution 1:10:33.760 --> 1:10:36.320 is the thing that gave us our desires. When we 1:10:36.360 --> 1:10:39.990 sit around ruminating, when that whole freight train of ideas 1:10:39.990 --> 1:10:42.910 is just scrolling across the things that's in our mind 1:10:42.950 --> 1:10:47.310 is generated by evolution. The reason we have these desires 1:10:47.310 --> 1:10:51.710 and goals and ambitions is because evolution is what designed us. 1:10:51.710 --> 1:10:55.070 We are basically mech suits, we're giant mechs. We're like 1:10:55.070 --> 1:10:59.470 giant transformer mech suits. Around evolution, which is trying to 1:10:59.470 --> 1:11:02.350 do things and it's trying to do things through us. 1:11:02.390 --> 1:11:08.390 Our entire pleasure system, our chemicals, dopamine, serotonin, cortisol, all 1:11:08.390 --> 1:11:11.430 these things are designed. The whole purpose that they're there 1:11:11.430 --> 1:11:15.350 is to steer us, to steer us towards desires. Do 1:11:15.350 --> 1:11:17.910 we have any control over our desires? Can you say 1:11:17.910 --> 1:11:21.310 I suddenly like broccoli when you didn't a second ago? 1:11:21.350 --> 1:11:23.630 Can you decide to do that? No you can't. You 1:11:23.630 --> 1:11:27.510 also can't really control what you desire. You can't really 1:11:27.510 --> 1:11:30.470 control your ambitions. You kind of have them. You could 1:11:30.470 --> 1:11:32.990 try to use your intellect and steer them, I'll grant 1:11:32.990 --> 1:11:35.780 you that. But in general, we are being powered. We 1:11:35.820 --> 1:11:39.700 are being powered by a very ancient, powerful thing, which 1:11:39.700 --> 1:11:45.340 is our upbringing, our evolution, our hardware, our mammal minds, right? 1:11:45.380 --> 1:11:51.700 Our brains, our minds, our identity. It is driving forward, survive, reproduce, 1:11:51.700 --> 1:11:58.340 become attractive, become powerful, become respected. That drive makes us fundamentally, 1:11:58.340 --> 1:12:01.660 I would say better, definitely different, but I would say better. 1:12:01.660 --> 1:12:05.540 It gives us an advantage over eyes. Eyes are dormant. 1:12:05.540 --> 1:12:08.500 They just sit there. They just execute things that we want. 1:12:08.540 --> 1:12:11.300 When they act like they feel things, when they act 1:12:11.300 --> 1:12:13.700 like they have goals and they act like, hey, I'm 1:12:13.700 --> 1:12:16.340 going to generate ideas that is not coming from their 1:12:16.340 --> 1:12:19.100 true center of self the way that it does for humans. 1:12:19.100 --> 1:12:21.860 So what does this mean? What this means is we 1:12:21.860 --> 1:12:25.020 are still the centerpiece of this whole thing. Everything we've 1:12:25.020 --> 1:12:28.460 been talking about in this video, everything around why AI 1:12:28.460 --> 1:12:32.260 is going to replace us doing work is around the execution. 1:12:32.260 --> 1:12:36.370 It's all around the implementation of someone's idea of a 1:12:36.370 --> 1:12:40.090 founder's idea of a company builders idea. Well guess what? 1:12:40.090 --> 1:12:42.810 There's a trillion more ideas out there that need to 1:12:42.810 --> 1:12:45.730 be made, and AI makes it so that it's easy 1:12:45.730 --> 1:12:48.530 to go and actually implement them, because you could just 1:12:48.530 --> 1:12:51.410 implement a system like lattice and stand up your company 1:12:51.410 --> 1:12:53.650 and you have some humans there, you have your buddies there, 1:12:53.650 --> 1:12:56.210 you have a bunch of A's and a bunch of agents. 1:12:56.210 --> 1:12:59.689 You have transparency. You say, okay, I want to execute this. 1:12:59.690 --> 1:13:01.890 I want to do this. Where are humans in this 1:13:01.890 --> 1:13:05.650 process of like starting companies, managing companies, doing all this stuff? 1:13:05.690 --> 1:13:08.010 They're at the top. We are at the top. Do 1:13:08.010 --> 1:13:10.690 you remember how you saw all those workflows and all 1:13:10.690 --> 1:13:13.930 those diagrams, all those interconnected pieces? What is the role 1:13:13.930 --> 1:13:17.130 of the human when the AI does all that work better? 1:13:17.130 --> 1:13:19.370 The role the human is to say, I don't want 1:13:19.370 --> 1:13:22.530 you to do that task anymore, okay, I can't improve 1:13:22.530 --> 1:13:25.690 how you do the task. Your AI, you're great. Do 1:13:25.690 --> 1:13:29.290 the workflow. You got it faster, cheaper, great. Who determines 1:13:29.290 --> 1:13:32.050 what the company should be doing, why they're doing it, 1:13:32.090 --> 1:13:35.630 what problem they're solving that is only human, right? This 1:13:35.630 --> 1:13:40.710 is incredibly positive. Incredibly optimistic. This is what we do 1:13:40.750 --> 1:13:44.230 that they can't. Okay. And this is where my recommendation 1:13:44.230 --> 1:13:46.550 and everything I talk about in all my videos, all 1:13:46.550 --> 1:13:49.710 my open source projects, my recommendation is to make the 1:13:49.710 --> 1:13:54.190 switch from focusing on the execution and the implementation, which 1:13:54.190 --> 1:13:56.310 is all the first part of this video. It is 1:13:56.310 --> 1:13:59.230 all this debate about can AI do it? Can AI 1:13:59.310 --> 1:14:02.430 not do it? It's all it's all a distraction. AI 1:14:02.430 --> 1:14:05.710 is going to crush that stuff. Those jobs are gone. 1:14:05.710 --> 1:14:09.550 That's why I consider this threat so dangerous. That's why 1:14:09.550 --> 1:14:12.750 I made this video to address this narrative that our 1:14:12.750 --> 1:14:15.870 jobs are not in danger from AI. It's because if 1:14:15.910 --> 1:14:18.910 you don't realize this, then you will keep focusing your 1:14:18.910 --> 1:14:22.030 energy on trying to maintain this job that is going away. 1:14:22.030 --> 1:14:25.510 Instead of transitioning all your effort to getting to the 1:14:25.510 --> 1:14:28.150 other side. This thing I call human 3.0, it is 1:14:28.150 --> 1:14:33.420 basically becoming your full self, Magnifying your true self, understanding 1:14:33.420 --> 1:14:37.540 your true self, broadcasting it and interconnecting that with other 1:14:37.540 --> 1:14:39.780 humans who are doing the same. So we can have 1:14:39.780 --> 1:14:42.900 value exchange from human to human without the need for 1:14:42.900 --> 1:14:46.460 this corporate hierarchy, all these corporations, all that kind of stuff. 1:14:46.460 --> 1:14:49.100 It's all, it's all going away. First of all, but 1:14:49.100 --> 1:14:51.100 it's not good in the first place. We don't need 1:14:51.100 --> 1:14:52.740 to hold on to it. We need to get to 1:14:52.780 --> 1:14:55.020 the other side where we can look down at a 1:14:55.020 --> 1:14:58.140 lattice like system where all the AI is executing, but 1:14:58.140 --> 1:15:00.860 we got to decide what to build. We built that company, 1:15:00.860 --> 1:15:03.500 we started that business. We can have 20 of them, 1:15:03.500 --> 1:15:05.460 and we have a dashboard that looks down at all 1:15:05.460 --> 1:15:07.500 20 of them, and they're all doing awesome stuff. We 1:15:07.500 --> 1:15:09.140 could drill in and we could say, hey, we're not 1:15:09.140 --> 1:15:11.460 doing this anymore. We're doing it in a different way. Hey, 1:15:11.460 --> 1:15:13.100 I want to optimize this. I want to do it 1:15:13.100 --> 1:15:16.340 that way. And the AI goes and executes. This is 1:15:16.340 --> 1:15:19.140 why I care so much about this. This is why 1:15:19.420 --> 1:15:24.100 I counter this argument that AI will not replace knowledge work. 1:15:24.100 --> 1:15:26.220 It is going to replace it. And we need to 1:15:26.220 --> 1:15:29.260 get ready. And it's not bad news. It's actually really 1:15:29.260 --> 1:15:30.820 good news. We'll see you in the next one.