WEBVTT 00:00:00.000 --> 00:00:02.339 Think about the sheer scale of the new economy 00:00:02.339 --> 00:00:05.700 for a second. Beat. We are looking at a truly 00:00:05.700 --> 00:00:09.039 wild reality right now. Oh, absolutely. One person 00:00:09.039 --> 00:00:12.300 can run a $10 million company entirely alone. 00:00:12.460 --> 00:00:15.759 They have absolutely zero human employees to 00:00:15.759 --> 00:00:17.879 manage. Yeah, they just orchestrate smart systems 00:00:17.879 --> 00:00:20.940 in the cloud. It is a profound shift in human 00:00:20.940 --> 00:00:23.179 potential. Welcome to the deep dive, by the way. 00:00:23.300 --> 00:00:25.699 I am super excited for this one. Today, we're 00:00:25.699 --> 00:00:28.579 unpacking a fascinating new roadmap for building 00:00:28.579 --> 00:00:31.699 an AI business. This is a step -by -step framework 00:00:31.699 --> 00:00:34.759 for starting from absolute zero. Right, specifically 00:00:34.759 --> 00:00:38.140 as a solo founder. Our mission today is to deeply 00:00:38.140 --> 00:00:40.399 deconstruct this entire process. We're going 00:00:40.399 --> 00:00:42.840 to explore how to find the exact right problem 00:00:42.840 --> 00:00:45.700 to solve. We'll look at how to validate it without 00:00:45.700 --> 00:00:48.439 writing a single line of code. And finally, we'll 00:00:48.439 --> 00:00:51.359 map out how to scale a team of AI agents. And, 00:00:51.359 --> 00:00:53.359 you know, AI agents are just smart programs that 00:00:53.359 --> 00:00:56.140 work without human help to complete complex tasks. 00:00:56.359 --> 00:00:58.500 They're essentially stepping in to replace human 00:00:58.500 --> 00:01:00.820 employees entirely. Right, and it completely 00:01:00.820 --> 00:01:02.799 rewrites the fundamental rules of business. You 00:01:02.799 --> 00:01:05.140 just don't need a massive team to scale anymore. 00:01:05.390 --> 00:01:07.870 Let's start by dismantling the old idea of a 00:01:07.870 --> 00:01:11.189 good company. We really have to break that concept 00:01:11.189 --> 00:01:13.810 before building the new one. Yeah, the old metric 00:01:13.810 --> 00:01:16.670 of success was always headcount. Founders loved 00:01:16.670 --> 00:01:19.409 bragging about having over 100 employees. They 00:01:19.409 --> 00:01:22.590 wanted huge, shiny offices and massive corporate 00:01:22.590 --> 00:01:25.230 teams. It was all about how big your empire looked 00:01:25.230 --> 00:01:27.790 from the outside. But, you know, more people 00:01:27.790 --> 00:01:30.030 usually just means a lot more problems. Well, 00:01:30.090 --> 00:01:33.269 exactly. You have to constantly manage complex 00:01:33.269 --> 00:01:36.489 company culture. And you're losing sleep worrying 00:01:36.489 --> 00:01:39.230 about making payroll every single month. Plus, 00:01:39.530 --> 00:01:42.250 you inevitably get trapped in endless non -productive 00:01:42.250 --> 00:01:44.890 meetings. The new goal for a solo founder is 00:01:44.890 --> 00:01:47.790 entirely different. You want maximum profit with 00:01:47.790 --> 00:01:50.290 the fewest people possible. AI agents handle 00:01:50.290 --> 00:01:52.530 all the heavy lifting for you now. They manage 00:01:52.530 --> 00:01:55.109 customer support, they write software, and they 00:01:55.109 --> 00:01:57.370 drive sales. The profit margin stays right in 00:01:57.370 --> 00:02:00.209 your own pocket. It really echoes Elon Musk's 00:02:00.209 --> 00:02:02.450 core engineering philosophy. Right. Building 00:02:02.450 --> 00:02:04.629 the machine that runs the actual machine. Exactly. 00:02:04.810 --> 00:02:07.049 Your job as a solo founder is simply the designer. 00:02:07.409 --> 00:02:10.930 You create the system, set the parameters and 00:02:10.930 --> 00:02:13.949 let it run itself. It's like shifting from managing 00:02:13.949 --> 00:02:17.830 a massive factory floor beat to just being the 00:02:17.830 --> 00:02:19.930 architect who designs the factory. That's a great 00:02:19.930 --> 00:02:22.090 way to look at it. And the starting costs are 00:02:22.090 --> 00:02:24.669 practically at zero now. You can run a complex 00:02:24.669 --> 00:02:27.030 application for 20 bucks a month. So if your 00:02:27.030 --> 00:02:29.830 idea fails, you only lose a little bit of time. 00:02:29.990 --> 00:02:32.349 Yeah, you aren't risking your entire life savings 00:02:32.349 --> 00:02:35.370 anymore. Beat. But what about the psychological 00:02:35.370 --> 00:02:38.110 toll of working this way? You don't have a human 00:02:38.110 --> 00:02:40.469 team to bounce ideas off of. It's definitely 00:02:40.469 --> 00:02:43.389 a lonely process at first. Without a team, you 00:02:43.389 --> 00:02:45.490 have to become your own sounding board completely. 00:02:45.729 --> 00:02:48.729 You have to rely purely on data instead of daily 00:02:48.729 --> 00:02:51.430 office chatter. Got it. So pure data replaces 00:02:51.430 --> 00:02:54.050 the comfort of office chatter. Precisely. It's 00:02:54.050 --> 00:02:56.509 a huge adjustment. Since it costs almost nothing 00:02:56.509 --> 00:02:59.210 to start nowadays, the traps change entirely. 00:02:59.409 --> 00:03:01.389 Right. The real risk isn't losing your startup 00:03:01.389 --> 00:03:04.090 capital anymore. The real trap is spending months 00:03:04.090 --> 00:03:06.810 building something nobody actually wants. Exactly. 00:03:07.050 --> 00:03:09.449 So how do we find what people actually do want? 00:03:09.770 --> 00:03:12.030 The absolute biggest mistake is falling in love 00:03:12.030 --> 00:03:14.150 with cool tech. Oh, I see this all the time. 00:03:14.530 --> 00:03:17.689 Founders see a flashy new AI tool and get overly 00:03:17.689 --> 00:03:20.729 excited, they rush to find a way to sell it immediately. 00:03:21.030 --> 00:03:23.189 They look for a problem to fit their cool new 00:03:23.189 --> 00:03:26.610 solution. I still wrestle with prom drift myself, 00:03:26.830 --> 00:03:29.370 honestly. Yeah. Or just getting seduced by cool 00:03:29.370 --> 00:03:31.849 tech instead of focusing on real problems. It's 00:03:31.849 --> 00:03:35.009 a, well, it's a completely natural trap for smart, 00:03:35.050 --> 00:03:38.629 curious people. But you have to ruthlessly focus 00:03:38.629 --> 00:03:41.770 on painkillers, not vitamins. Vitamins are just 00:03:41.770 --> 00:03:44.150 nice things to have around. Right. People often 00:03:44.150 --> 00:03:47.289 forget to take them every single day. But a painkiller 00:03:47.289 --> 00:03:50.050 solves a burning problem, keeping customers awake 00:03:50.050 --> 00:03:52.530 at night. The strategy here suggests targeting 00:03:52.530 --> 00:03:55.409 old school established industries. We're talking 00:03:55.409 --> 00:03:57.629 about places like real estate, corporate law, 00:03:57.830 --> 00:04:00.719 or health care. Those industries still have very 00:04:00.719 --> 00:04:04.080 old, deeply manual ways of working. They're already 00:04:04.080 --> 00:04:06.199 spending huge amounts of money to solve basic 00:04:06.199 --> 00:04:08.800 administrative issues. You can actually use an 00:04:08.800 --> 00:04:11.020 LLM for your initial market research. Right, 00:04:11.139 --> 00:04:14.340 and LLMs are just large AI models that understand 00:04:14.340 --> 00:04:17.100 and generate text. You ask a tool like Gemini 00:04:17.100 --> 00:04:19.939 to dig deep into an industry. There's a brilliant 00:04:19.939 --> 00:04:22.660 specific example of using Gemini for this. You 00:04:22.660 --> 00:04:25.420 ask the AI for real estate tasks taking over 00:04:25.420 --> 00:04:28.879 two hours, specifically tasks that cost serious 00:04:28.879 --> 00:04:31.139 money if they're done wrong. Yeah. And in that 00:04:31.139 --> 00:04:33.839 example, the AI highlighted lead qualification. 00:04:34.199 --> 00:04:37.180 Real estate agents manually check hundreds of 00:04:37.180 --> 00:04:40.379 emails to find two actual buyers. It's a massive, 00:04:40.720 --> 00:04:43.019 exhausting pain point, perfectly ripe for AI 00:04:43.019 --> 00:04:44.959 automation. Then there's the advice strategy 00:04:44.959 --> 00:04:47.399 to dig even deeper. You find 10 people in that 00:04:47.399 --> 00:04:49.339 industry and ask for 10 minutes. You ask for 00:04:49.339 --> 00:04:51.100 their advice, making sure you weren't selling 00:04:51.100 --> 00:04:53.100 anything at all. You literally just ask them 00:04:53.100 --> 00:04:55.459 what tasks make them go crazy. It's amazing how 00:04:55.459 --> 00:04:57.420 much people love complaining about their jobs. 00:04:57.540 --> 00:04:59.620 They will hand you your exact business model 00:04:59.620 --> 00:05:02.420 if you just listen. But how do you resist the 00:05:02.420 --> 00:05:04.959 urge to start building immediately? The tech 00:05:04.959 --> 00:05:07.259 is just so incredibly accessible and tempting 00:05:07.259 --> 00:05:09.399 right now. You have to force yourself to talk 00:05:09.399 --> 00:05:12.819 to humans first. Writing code is cheap, but finding 00:05:12.819 --> 00:05:16.060 real, painful problems is incredibly hard. Finding 00:05:16.060 --> 00:05:18.639 the actual problem is much harder than writing 00:05:18.639 --> 00:05:21.670 code. Exactly. Don't build until it hurts. So 00:05:21.670 --> 00:05:24.769 you found the pain and you got the advice. The 00:05:24.769 --> 00:05:27.290 next logical step really feels like opening a 00:05:27.290 --> 00:05:30.430 code editor, but the strategy says absolutely 00:05:30.430 --> 00:05:33.069 do not do that yet. Right, you never build software 00:05:33.069 --> 00:05:35.490 until someone actually pays for it. You test 00:05:35.490 --> 00:05:38.389 the concept through a process called manual validation. 00:05:38.589 --> 00:05:40.709 This is also known as the spreadsheet method. 00:05:41.029 --> 00:05:44.389 You essentially clean and process raw data yourself 00:05:44.389 --> 00:05:48.589 in Excel or Sheets. You just use a simple AI 00:05:48.589 --> 00:05:51.110 chat interface to help you move faster. It's 00:05:51.110 --> 00:05:53.170 brilliant because you're literally getting paid 00:05:53.170 --> 00:05:55.649 to learn the problem. There's a great example 00:05:55.649 --> 00:05:58.230 about organizing messy customer support emails. 00:05:58.829 --> 00:06:01.310 A founder did this manually in a Google Doc for 00:06:01.310 --> 00:06:03.610 two whole weeks. Just doing it entirely by hand. 00:06:03.810 --> 00:06:06.410 Yeah. And by doing it by hand, they discovered 00:06:06.410 --> 00:06:09.550 a crucial pattern. 80 % of the emails were the 00:06:09.550 --> 00:06:11.569 exact same three questions. And that specific 00:06:11.569 --> 00:06:14.170 data perfectly maps out the bot you'll build 00:06:14.170 --> 00:06:16.449 later. Exactly. You're building the blueprint 00:06:16.449 --> 00:06:19.360 by doing the dirty work yourself. I'm struggling 00:06:19.360 --> 00:06:20.759 with this part. I have to push back a little. 00:06:21.620 --> 00:06:24.399 If I'm a solo founder trying to move at lightning 00:06:24.399 --> 00:06:29.160 speed, isn't manually sorting emails in a Google 00:06:29.160 --> 00:06:32.720 Doc incredibly exhausting and unscalable? It's 00:06:32.720 --> 00:06:35.120 definitely exhausting, but it's a strictly temporary 00:06:35.120 --> 00:06:38.800 phase. You are deeply learning the exact details 00:06:38.800 --> 00:06:41.980 the AI must eventually handle. You find out if 00:06:41.980 --> 00:06:44.639 your customers actually care about speed or just 00:06:44.639 --> 00:06:48.949 accuracy. automate a broken process, you just 00:06:48.949 --> 00:06:51.509 scale the chaos. Two sec silence. Do customers 00:06:51.509 --> 00:06:54.189 mind the manual delay before the software is 00:06:54.189 --> 00:06:56.509 actually built? Not at all. They're paying for 00:06:56.509 --> 00:06:58.750 their massive headache to go away, not the interface. 00:06:59.250 --> 00:07:00.629 Makes sense. They just want the raw solution, 00:07:00.810 --> 00:07:02.569 not a slick interface. Right. The interface comes 00:07:02.569 --> 00:07:05.040 later. Once you know your manual process actually 00:07:05.040 --> 00:07:07.480 works and solves the pain, you have to package 00:07:07.480 --> 00:07:09.660 it so people believe in the future product. You 00:07:09.660 --> 00:07:12.639 use a proven five -part formula to close those 00:07:12.639 --> 00:07:15.319 early customers. You clearly define the problem, 00:07:15.480 --> 00:07:18.160 the promise, the timeline, and the price. And 00:07:18.160 --> 00:07:20.819 you back it all up with a massive risk -free 00:07:20.819 --> 00:07:23.670 guarantee. This moves us right into the Wizard 00:07:23.670 --> 00:07:27.029 of Oz stage. You make it look like fully functioning, 00:07:27.329 --> 00:07:29.930 polished software, even if you're just pulling 00:07:29.930 --> 00:07:32.189 levers manually in the background. It's just 00:07:32.189 --> 00:07:34.129 like building a classic Hollywood movie set. 00:07:34.310 --> 00:07:36.930 Yes. It looks perfectly like a functioning building 00:07:36.930 --> 00:07:39.389 from the outside, but it's really just a painted 00:07:39.389 --> 00:07:41.889 front until the customer buys in. That's the 00:07:41.889 --> 00:07:44.329 perfect analogy. And you must follow the three 00:07:44.329 --> 00:07:47.889 screen rule strictly for this prototype. Login, 00:07:47.889 --> 00:07:50.670 input, and output. Exactly. You keep the entire 00:07:50.670 --> 00:07:53.490 experience incredibly simple. You can use AI 00:07:53.490 --> 00:07:56.769 design tools like Visily or UX Pilot. You design 00:07:56.769 --> 00:07:59.470 these screens using very simple conversational 00:07:59.470 --> 00:08:01.449 text prompts. Then you just link them all together 00:08:01.449 --> 00:08:03.870 inside an app like Figma. Then you absolutely 00:08:03.870 --> 00:08:06.930 must test it out with five real people. There's 00:08:06.930 --> 00:08:09.370 a crucial anecdote about testing early prototypes 00:08:09.370 --> 00:08:12.410 with live users. A founder was watching customers 00:08:12.410 --> 00:08:14.689 try to use his new app. They kept clicking the 00:08:14.689 --> 00:08:16.910 company logo instead of the submit button. Because 00:08:16.910 --> 00:08:18.790 the submit button was designed way too small 00:08:18.790 --> 00:08:21.470 to see. Exactly. And because it was just a prototype, 00:08:21.769 --> 00:08:24.310 they fixed it in two minutes. Imagine spending 00:08:24.310 --> 00:08:27.149 weeks coding that whole backend perfectly, only 00:08:27.149 --> 00:08:28.949 to find out users can't even find the button 00:08:28.949 --> 00:08:32.730 to start it. Meek. Why is three screens the strict 00:08:32.730 --> 00:08:35.809 maximum for a prototype? Anything more just confuses 00:08:35.809 --> 00:08:38.490 the user and hides the core value you're providing. 00:08:38.830 --> 00:08:41.490 So extra screens just completely dilute your 00:08:41.490 --> 00:08:43.750 product's core value? Keep it shockingly simple. 00:08:44.080 --> 00:08:46.639 We're going to take a quick break right here. 00:08:46.840 --> 00:08:49.059 Sponsor. All right. We are back. All right. So 00:08:49.059 --> 00:08:51.860 the prototype works and your early customers 00:08:51.860 --> 00:08:55.639 are happily paying. Now we finally let the AI 00:08:55.639 --> 00:08:59.120 build the real scalable machine. You can use 00:08:59.120 --> 00:09:01.980 tools like Manus AI to build the actual app. 00:09:02.240 --> 00:09:04.799 It literally codes the application from a single 00:09:04.799 --> 00:09:07.080 detailed text prompt. You just treat the AI like 00:09:07.080 --> 00:09:09.879 a very eager, very fast intern. Yeah. You tell 00:09:09.879 --> 00:09:11.940 it to move buttons, change colors, or adjust 00:09:11.940 --> 00:09:14.049 the layout. This officially begins the learner 00:09:14.049 --> 00:09:17.289 phase, taking you from 0 to 100k. You're still 00:09:17.289 --> 00:09:19.590 operating entirely as a solo worker here, but 00:09:19.590 --> 00:09:22.149 you're speeding yourself up immensely using smart 00:09:22.149 --> 00:09:25.190 tools. You use Claude to draft your emails and 00:09:25.190 --> 00:09:27.909 Gemini for deep research. You use tools like 00:09:27.909 --> 00:09:30.350 SocialSweep to manage your entire social media 00:09:30.350 --> 00:09:33.110 presence. The golden rule here is learning the 00:09:33.110 --> 00:09:35.750 rules before automating the rules. Next is the 00:09:35.750 --> 00:09:38.690 automation phase, scaling you up to $1 million. 00:09:39.669 --> 00:09:42.090 This is where you automate repeated tasks with 00:09:42.090 --> 00:09:45.370 a tool called N8n. Right, N8n is a tool that 00:09:45.370 --> 00:09:47.210 connects different apps automatically without 00:09:47.210 --> 00:09:49.549 any clicking. It essentially acts as a digital 00:09:49.549 --> 00:09:52.590 nervous system for your business. AI chatbots 00:09:52.590 --> 00:09:55.750 take over 90 % of your routine customer support. 00:09:56.169 --> 00:09:58.750 AI sales agents start handling all your cold 00:09:58.750 --> 00:10:01.610 outreach emails completely autonomously. Then 00:10:01.610 --> 00:10:04.090 we finally arrive at the ultimate stage. the 00:10:04.090 --> 00:10:06.370 agentic phase. This is the architecture that 00:10:06.370 --> 00:10:08.909 scales you from one to ten million dollars. You're 00:10:08.909 --> 00:10:11.269 building complex workflows. Right, which are 00:10:11.269 --> 00:10:14.049 essentially chains of specialized AI agents working 00:10:14.049 --> 00:10:16.809 together. Let's walk through a concrete example 00:10:16.809 --> 00:10:19.370 of how this actually looks. Think of Agent A 00:10:19.370 --> 00:10:22.230 as your relentless digital researcher. It scours 00:10:22.230 --> 00:10:24.649 the internet daily, pulling hundreds of fresh, 00:10:24.789 --> 00:10:27.210 targeted leads. Then it instantly hands that 00:10:27.210 --> 00:10:30.330 data off to Agent B. Agent B checks every single 00:10:30.330 --> 00:10:33.350 leads website to ensure a perfect fit. Then Agent 00:10:33.350 --> 00:10:36.049 C steps in to write highly personalized emails 00:10:36.049 --> 00:10:39.090 to those leads. Exactly. Agent D monitors your 00:10:39.090 --> 00:10:41.789 inbox, checks your calendar, and books the actual 00:10:41.789 --> 00:10:45.409 meetings. And Agent E? Agent E sends them a customized 00:10:45.409 --> 00:10:48.450 demo video before the call starts. The only time 00:10:48.450 --> 00:10:51.509 a human ever touches this process is at the very 00:10:51.509 --> 00:10:53.789 end. You just show up to the Zoom call and close 00:10:53.789 --> 00:10:56.889 the deal. That's it. Two sec silence. Whoa! Imagine 00:10:56.889 --> 00:10:59.730 an entire... bustling corporate sales floor, 00:11:00.549 --> 00:11:03.429 just entirely replaced by five silent scripts 00:11:03.429 --> 00:11:06.029 running flawlessly in the background. It is staggering 00:11:06.029 --> 00:11:08.450 when you see it working in real time. What typically 00:11:08.450 --> 00:11:10.789 breaks first when transitioning to this complex, 00:11:10.950 --> 00:11:13.590 agentic phase? The complex handoffs between agents 00:11:13.590 --> 00:11:16.210 fail if your underlying data gets messy. Right. 00:11:16.210 --> 00:11:18.570 Messy data breaks the crucial handoffs between 00:11:18.570 --> 00:11:21.610 your AI agents. Garbage in, garbage out still 00:11:21.610 --> 00:11:24.169 applies to AI. Scaling like this really sounds 00:11:24.169 --> 00:11:26.590 like a straight, glorious lineup. It really is, 00:11:26.690 --> 00:11:29.730 though. Right. This roadmap warns of three massive 00:11:29.730 --> 00:11:32.350 potholes along the way. These specific mistakes 00:11:32.350 --> 00:11:35.590 completely kill these promising solo AI businesses. 00:11:36.149 --> 00:11:39.009 The first massive mistake is overcomplicating 00:11:39.009 --> 00:11:41.450 the underlying tech stack. You really don't need 00:11:41.450 --> 00:11:44.490 to train your own custom AI models. Just use 00:11:44.490 --> 00:11:47.490 existing off -the -shelf models like GPT -4 or 00:11:47.490 --> 00:11:49.570 Claude 3. Yeah, they are already smarter than 00:11:49.570 --> 00:11:52.149 you need them to be. The second mistake is completely 00:11:52.149 --> 00:11:55.250 ignoring the human customer. Yes. You cannot 00:11:55.250 --> 00:11:57.950 sit in a cave coding for six months without talking 00:11:57.950 --> 00:12:00.629 to humans. You must talk to your actual users 00:12:00.629 --> 00:12:03.830 every single week. The third big mistake is trying 00:12:03.830 --> 00:12:06.330 to be everything to everyone. Don't try to serve 00:12:06.330 --> 00:12:08.610 real estate corporate law and plumbing all at 00:12:08.610 --> 00:12:12.049 once. Pick one single specific niche and master 00:12:12.049 --> 00:12:14.870 it deeply. You need to grow to one million dollars 00:12:14.870 --> 00:12:17.649 in revenue before expanding further. Absolutely. 00:12:17.830 --> 00:12:20.450 Focus is everything. Beat. This brings up a really 00:12:20.450 --> 00:12:22.409 important question about business defensibility. 00:12:23.049 --> 00:12:25.710 If we're just using off -the -shelf tech and 00:12:25.710 --> 00:12:28.529 simple text prompts, how do we defend ourselves? 00:12:28.990 --> 00:12:31.750 Does this simple tech get easily cloned by competitors? 00:12:31.870 --> 00:12:34.850 Competitors can copy your code, but they can't 00:12:34.850 --> 00:12:37.570 copy your deep customer workflow knowledge. Deep 00:12:37.570 --> 00:12:40.590 customer knowledge becomes your only true, uncopyable 00:12:40.590 --> 00:12:43.690 defensive moat. Absolutely. Knowing exactly why 00:12:43.690 --> 00:12:46.029 a real estate agent clicks a button is priceless. 00:12:46.250 --> 00:12:48.850 Let's summarize the core philosophy of this entire 00:12:48.850 --> 00:12:51.129 roadmap. You need to start building your business 00:12:51.129 --> 00:12:53.960 today. Don't wait until you feel like an absolute 00:12:53.960 --> 00:12:56.360 AI expert. The so -called experts just started 00:12:56.360 --> 00:12:58.679 six months ago anyway. Find the pain in an old 00:12:58.679 --> 00:13:01.139 industry and validate it completely manually. 00:13:01.360 --> 00:13:04.799 Use these new AI tools as your incredibly fast 00:13:04.799 --> 00:13:08.120 tireless interns. And slowly scale from a solo 00:13:08.120 --> 00:13:10.779 operator to a synchronized agentic workflow. 00:13:11.320 --> 00:13:14.279 Exactly. To sex silence. This leaves me with 00:13:14.279 --> 00:13:17.679 a final provocative thought flow. If a single 00:13:17.679 --> 00:13:20.539 person with just a basic laptop and an internet 00:13:20.539 --> 00:13:23.759 connection. can build a highly customized $10 00:13:23.759 --> 00:13:26.879 million enterprise using just a synchronized 00:13:26.879 --> 00:13:29.879 chain of off -the -shelf AI agents. What happens 00:13:29.879 --> 00:13:31.799 to the fundamental concept of the traditional 00:13:31.799 --> 00:13:33.899 corporation? It's a crazy question. Think about 00:13:33.899 --> 00:13:35.659 this specifically over the course of the next 00:13:35.659 --> 00:13:38.259 10 years. Are we looking at a radically different 00:13:38.259 --> 00:13:41.460 economic future, a future completely dominated 00:13:41.460 --> 00:13:44.679 by single -person, billion -dollar digital empires? 00:13:44.720 --> 00:13:46.879 I think we just might be. Thank you for joining 00:13:46.879 --> 00:13:49.840 us on this Deep Dive. Go right down your one 00:13:49.840 --> 00:13:51.690 -page offer today. We'll catch you on the next 00:13:51.690 --> 00:13:53.009 one. Out to your own music.