WEBVTT - Most Companies Aren't Anywhere Near Ready for AI 0:00:00.880 --> 0:00:04.520 Most companies aren't anywhere near ready for AI, and it's 0:00:04.519 --> 0:00:06.880 not that they aren't using AI, it's that they can't 0:00:06.880 --> 0:00:10.319 use AI. Most of the frustration people have with AI 0:00:10.360 --> 0:00:12.399 not being able to do what they want is actually 0:00:12.400 --> 0:00:15.360 them not being able to describe what they want. I've 0:00:15.360 --> 0:00:19.279 consulted for the world's largest companies hundreds of startups, tons 0:00:19.280 --> 0:00:23.360 of mid-sized companies, and lots of companies in the global 1000. 0:00:23.600 --> 0:00:26.960 And the number one issue I see is unclear and 0:00:26.960 --> 0:00:33.000 constantly changing vision and goals. AI is about execution, and 0:00:33.000 --> 0:00:35.760 it's quite powerless when it doesn't know what to execute. 0:00:36.280 --> 0:00:39.479 Companies that know exactly what they want are thriving with AI, 0:00:40.040 --> 0:00:41.800 and the better the AI gets, the more they will 0:00:41.800 --> 0:00:45.120 crush it. But this is a small percentage of companies, 0:00:45.120 --> 0:00:48.440 because only a small percentage of companies are self-aware and 0:00:48.440 --> 0:00:52.720 together enough to give AI proper instructions. You can't optimize 0:00:52.720 --> 0:00:55.720 what you don't understand, and it's foolish to scale something 0:00:55.720 --> 0:00:58.720 that you shouldn't be doing in the first place. People 0:00:58.760 --> 0:01:01.400 talk about companies not being ready for AI, and I 0:01:01.400 --> 0:01:04.540 don't think they realize how bad the problem is. this 0:01:04.540 --> 0:01:07.140 is not a technical maturity issue. It's a whole lot 0:01:07.140 --> 0:01:10.980 more fundamental than that. A massive percentage of companies are 0:01:11.020 --> 0:01:15.220 haphazardly successful despite themselves. It's not super clear what they 0:01:15.220 --> 0:01:18.220 were trying to accomplish or how exactly they're doing it, 0:01:18.220 --> 0:01:20.699 but they have a few tricks that work that they're 0:01:20.700 --> 0:01:24.380 decent enough at executing, that they're still around. But if 0:01:24.380 --> 0:01:25.940 you were to walk into the company and say, okay, 0:01:25.980 --> 0:01:28.020 describe to me what you're trying to do and what 0:01:28.020 --> 0:01:31.420 your strategies are and your challenges are and what your 0:01:31.420 --> 0:01:33.940 work streams are, they would either stare blankly at you 0:01:33.940 --> 0:01:36.100 or laugh in your face. It would take them weeks 0:01:36.100 --> 0:01:39.140 to put together a project to find that out, and 0:01:39.140 --> 0:01:41.979 then months to actually do the project. And of course, 0:01:41.980 --> 0:01:43.700 by the time you got done, it would be out 0:01:43.740 --> 0:01:47.179 of date. I honestly believe the vast majority of companies 0:01:47.180 --> 0:01:50.420 are in grave danger because they are essentially chaotic black 0:01:50.460 --> 0:01:54.260 boxes that barely work. Having the board or leadership tell 0:01:54.260 --> 0:01:57.540 everyone in a company to use AI is like looking 0:01:57.540 --> 0:02:00.580 at a screen of nothing but static and saying, we 0:02:00.620 --> 0:02:04.700 got to scale this thing. So worker says, cool, boss. 0:02:05.380 --> 0:02:06.780 Do me a favor and point to the thing you 0:02:06.780 --> 0:02:09.960 want me to improve. And the boss, like, pauses for 0:02:09.960 --> 0:02:12.400 a little bit and says, yeah, great question. Let's get 0:02:12.400 --> 0:02:16.360 some meetings together to figure that out. AI is mostly 0:02:16.360 --> 0:02:19.880 useless in these companies, and unfortunately, that means many, if 0:02:19.880 --> 0:02:23.839 not most companies. Their narrative right now is that AI 0:02:23.880 --> 0:02:27.040 isn't helping enough companies yet. So what's the problem? When 0:02:27.040 --> 0:02:29.840 will AI get good enough? Maybe AI isn't as good 0:02:29.840 --> 0:02:33.080 as people thought. What's hilarious and easy to see is 0:02:33.080 --> 0:02:36.200 that the company's AI is helping are somehow, magically the 0:02:36.200 --> 0:02:40.120 same companies that already know what they're doing. They can 0:02:40.120 --> 0:02:42.440 tell you very quickly the problem or problems that they're 0:02:42.440 --> 0:02:45.639 solving for customers. The problem with existing solutions that their 0:02:45.639 --> 0:02:48.280 solution solves, the goals that they have for the company, 0:02:48.280 --> 0:02:51.359 the metrics associated with those goals, the challenges that are 0:02:51.360 --> 0:02:54.640 stopping them from reaching their goals, the strategies that they 0:02:54.680 --> 0:02:58.040 are pursuing to overcome those challenges, the projects that they're 0:02:58.040 --> 0:03:01.120 doing to implement those strategies, the work that's being done 0:03:01.120 --> 0:03:04.200 within those projects by what people and how much it's 0:03:04.200 --> 0:03:07.840 costing them. And crucially, companies that know what they're doing 0:03:07.840 --> 0:03:10.780 give largely the same answers to these questions across different 0:03:10.780 --> 0:03:14.380 quarters and years. A sure sign of a flailing company 0:03:14.380 --> 0:03:17.420 is when their departments mean massive amounts of time claiming 0:03:17.419 --> 0:03:20.380 to have these answers. But the answers significantly change every 0:03:20.380 --> 0:03:23.820 quarter because everything is constantly in flux. So people are 0:03:23.820 --> 0:03:26.299 basically just writing down whatever they think makes them look 0:03:26.300 --> 0:03:29.180 good as a manager or as a department or whatever. 0:03:29.740 --> 0:03:32.180 It takes them weeks to prepare, only to be discarded 0:03:32.180 --> 0:03:34.700 a few weeks later when the process starts over again. 0:03:36.020 --> 0:03:38.700 AI can basically do nothing for these companies. In fact, 0:03:38.700 --> 0:03:40.620 it could even make it worse because now it helps 0:03:40.620 --> 0:03:44.180 people flail more impressively, like with more backflips and charts 0:03:44.180 --> 0:03:47.500 and stuff. And you may ask, well, if they're so bad, 0:03:47.740 --> 0:03:50.020 how are these companies still around? And the answer is 0:03:50.020 --> 0:03:54.100 simple most of their competition is just as bad. So 0:03:54.460 --> 0:03:56.740 what's the point of all this? A few things I 0:03:56.740 --> 0:03:59.900 think we can take away from this. AI has barely 0:03:59.900 --> 0:04:02.660 even started in the enterprise, because only a tiny fraction 0:04:02.660 --> 0:04:06.780 of companies have the self-knowledge and ability to articulate themselves 0:04:06.980 --> 0:04:10.700 well enough to be ready for AI. Two, we should 0:04:10.700 --> 0:04:13.380 stop looking at AI as the problem, or even technology 0:04:13.380 --> 0:04:15.640 as the problem. The issue is not being able to 0:04:15.640 --> 0:04:20.880 describe one's company clearly, including its goals, workflows, operations, decisions, teams, 0:04:20.880 --> 0:04:26.279 and spending in a clear and consumable fashion. Three companies 0:04:26.279 --> 0:04:28.600 who are unable to do this are in grave danger 0:04:28.720 --> 0:04:32.919 from the companies who can. For the primary danger to large, 0:04:32.960 --> 0:04:35.799 unwieldy companies is that it's now possible for a smaller 0:04:35.800 --> 0:04:38.880 company to function with the strength of a much larger one. 0:04:39.120 --> 0:04:41.120 And it's far easier for a small company to be 0:04:41.160 --> 0:04:44.320 able to answer the questions above. And because of this, 0:04:44.320 --> 0:04:48.240 all existing companies are about to face extraordinary downward force, 0:04:48.720 --> 0:04:51.200 and only companies who have their shit together will stay 0:04:51.200 --> 0:04:55.680 afloat and rise. AI is a remarkably small part of 0:04:55.720 --> 0:04:58.720 who will win or lose in this initial phase. AI 0:04:58.720 --> 0:05:00.960 is what the winners will use to fight each other 0:05:00.960 --> 0:05:03.160 in the new world. But right now, the game is 0:05:03.160 --> 0:05:06.120 figuring out who's going to make it there. The number 0:05:06.120 --> 0:05:08.320 one question you should be asking yourself as a company 0:05:08.600 --> 0:05:10.640 is not what AI can do for you, but whether 0:05:10.640 --> 0:05:12.560 or not your company is in the state where AI 0:05:12.600 --> 0:05:15.440 can help at all. And if not, you need to 0:05:15.440 --> 0:05:17.520 get into that state as quickly as possible.