WEBVTT 00:00:00.000 --> 00:00:02.580 Welcome to the Deep Dive! You know, we talk all 00:00:02.580 --> 00:00:05.500 the time about how fast AI is moving. Breakthroughs 00:00:05.500 --> 00:00:09.779 seem to happen, like, every day. But for a lot 00:00:09.779 --> 00:00:12.619 of us, how we actually use AI, it's still kind 00:00:12.619 --> 00:00:14.439 of superficial, right? Just asking for funny 00:00:14.439 --> 00:00:17.019 poems, maybe generating a cool picture. More 00:00:17.019 --> 00:00:19.460 out of curiosity, really. But here's the real 00:00:19.460 --> 00:00:22.179 challenge, the thing hiding underneath. While 00:00:22.179 --> 00:00:24.620 everyone's sort of experimenting, very few people 00:00:24.620 --> 00:00:27.640 genuinely understand how to harness AI's, well, 00:00:27.739 --> 00:00:31.660 its raw power to build a real, sustainable, It's 00:00:31.660 --> 00:00:34.320 this amazing tool, but how do you actually flip 00:00:34.320 --> 00:00:36.520 that switch? How do you make it a wealth generating 00:00:36.520 --> 00:00:38.780 engine for yourself? And that's exactly what 00:00:38.780 --> 00:00:40.240 we're digging into today. This isn't going to 00:00:40.240 --> 00:00:42.600 be another surface level chat about AI buzzwords. 00:00:42.600 --> 00:00:44.740 Forget that. We're doing a proper deep dive into 00:00:44.740 --> 00:00:47.700 seven specific essential AI skills. Skills that 00:00:47.700 --> 00:00:49.159 businesses, well, they're not just interested 00:00:49.159 --> 00:00:50.840 in them. They are desperate for people who have 00:00:50.840 --> 00:00:53.079 these skills and they're willing to pay top dollar 00:00:53.079 --> 00:00:55.479 in 2025 and, you know, looking beyond that too. 00:00:55.679 --> 00:00:57.579 Yeah, that's the key. Whether you've got a business 00:00:57.579 --> 00:00:59.899 idea simmering or you just want to seriously 00:00:59.899 --> 00:01:01.719 level up. up your career, getting good at even 00:01:01.719 --> 00:01:04.879 one or two of these, it's going to put you miles 00:01:04.879 --> 00:01:07.540 ahead. You'll create this huge gap between yourself 00:01:07.540 --> 00:01:10.159 and everyone else who's just using AI for the 00:01:10.159 --> 00:01:13.560 basics. So get ready. We're about to map out 00:01:13.560 --> 00:01:15.519 how you can turn AI from just an interesting 00:01:15.519 --> 00:01:20.099 toy into a powerful financial tool. OK, let's 00:01:20.099 --> 00:01:23.340 kick things off. Skill number one, prompt engineering. 00:01:23.900 --> 00:01:26.359 So. Think of AI like this incredibly powerful 00:01:26.359 --> 00:01:29.620 jet engine. Prompt engineering is the control 00:01:29.620 --> 00:01:31.239 panel. And maybe the pilot, too. It's how you 00:01:31.239 --> 00:01:32.959 talk to it. Exactly. It's how you give it those 00:01:32.959 --> 00:01:34.739 precise instructions to get it to do amazing 00:01:34.739 --> 00:01:36.840 things. And it's absolutely foundational. It 00:01:36.840 --> 00:01:38.200 unlocks pretty much everything else. And get 00:01:38.200 --> 00:01:40.519 this. Here's a pretty wild statistic I saw recently. 00:01:41.040 --> 00:01:43.680 Something like 78 % of all AI projects actually 00:01:43.680 --> 00:01:47.920 fail. 78 %? Really? Why? Is the tech not good 00:01:47.920 --> 00:01:49.959 enough? No, that's the crazy part. It's not usually 00:01:49.959 --> 00:01:53.579 the technology. It's the communication. It's 00:01:53.579 --> 00:01:57.599 the gap between the human asking and the AI understanding. 00:01:58.340 --> 00:02:00.480 Companies are basically throwing away mountains 00:02:00.480 --> 00:02:02.420 of money just because they haven't figured out 00:02:02.420 --> 00:02:04.870 how to. you know, talk to these systems properly. 00:02:04.950 --> 00:02:07.109 Wow, so it's a communication breakdown, essentially. 00:02:07.170 --> 00:02:09.750 At its core, yeah, a very expensive communication 00:02:09.750 --> 00:02:12.669 breakdown. It shows that mastering AI really 00:02:12.669 --> 00:02:15.129 starts with mastering how you ask the questions. 00:02:15.349 --> 00:02:16.949 Right, so if you get good at this, you're not 00:02:16.949 --> 00:02:18.669 just getting slightly better results, you're 00:02:18.669 --> 00:02:21.229 getting results that are, what, faster? More 00:02:21.229 --> 00:02:25.110 creative? And, critically, results that are actually 00:02:25.110 --> 00:02:27.810 aligned with what you intended to get. It's the 00:02:27.810 --> 00:02:30.930 difference between the AI giving you back some 00:02:31.069 --> 00:02:33.969 you know, mediocre first draft and prompting 00:02:33.969 --> 00:02:36.090 it to deliver a strategic masterpiece that just 00:02:36.090 --> 00:02:38.930 nails it. So the trick is you just start treating 00:02:38.930 --> 00:02:41.969 these AI systems less like a search box and more 00:02:41.969 --> 00:02:45.189 like a maybe a genius employee. A genius employee. 00:02:45.409 --> 00:02:47.789 OK, I like that. Yeah, they can do incredible 00:02:47.789 --> 00:02:50.129 things, but they aren't mind readers. They need 00:02:50.129 --> 00:02:52.909 really clear, detailed, structured instructions. 00:02:53.129 --> 00:02:55.270 We actually have a model for this sometimes called 00:02:55.270 --> 00:02:57.530 the professional prompt structure. You can remember 00:02:57.530 --> 00:03:01.310 it as our CTAF. RCTAF, what does that stand for? 00:03:01.590 --> 00:03:05.610 Roll, context, task, audience, and format. Let 00:03:05.610 --> 00:03:07.610 me give you a quick example. Instead of just 00:03:07.610 --> 00:03:09.750 saying, write me an ad for a new coffee shop, 00:03:09.870 --> 00:03:12.210 which is super vague, right? Yeah, very vague. 00:03:12.289 --> 00:03:14.669 You use the RCTAF structure. Like for this fictional 00:03:14.669 --> 00:03:18.530 shop, the Saigon Brew, roll. Act like a marketing 00:03:18.530 --> 00:03:20.490 expert, 10 years in high -end food and beverage, 00:03:20.990 --> 00:03:23.009 specialized in coffee shops. See? You give it 00:03:23.009 --> 00:03:25.900 a persona. OK, got it. Roll, then context. I'm 00:03:25.900 --> 00:03:27.979 launching the Saigon Brew in District 1, Ho Chi 00:03:27.979 --> 00:03:30.780 Minh City. It's pour -over, cold brew, quiet 00:03:30.780 --> 00:03:33.300 vibe, targets office workers, expatch. Price 00:03:33.300 --> 00:03:35.939 is about 20 % above average. Lots of detail. 00:03:35.960 --> 00:03:38.680 Ah, setting the scene. Exactly. Then the task. 00:03:39.139 --> 00:03:41.219 Write a Facebook post for the grand opening week. 00:03:41.780 --> 00:03:44.319 Create curiosity. Emphasize the quality beans, 00:03:44.479 --> 00:03:46.860 the workspace. Call to action is visit for a 00:03:46.860 --> 00:03:50.159 15 % discount. Very specific goal. Right. Audience. 00:03:50.729 --> 00:03:53.849 Coffee lovers, office workers, 2540, expats in 00:03:53.849 --> 00:03:56.569 each CMC who like quality. Who are we talking 00:03:56.569 --> 00:03:59.189 to? Audience. Makes sense. And finally, format. 00:03:59.770 --> 00:04:02.449 Catchy headline under 10 words. Body about 150 00:04:02.449 --> 00:04:06.250 words. Two, three paragraphs. Clear CTA. Use 00:04:06.250 --> 00:04:08.770 three, four emojis. How should it look? Format. 00:04:08.889 --> 00:04:10.810 Okay, R -C -T -A -F. I can see how that gets 00:04:10.810 --> 00:04:12.569 you a much better result than just write an ad. 00:04:12.889 --> 00:04:14.629 Night and day difference. And that roll part, 00:04:14.710 --> 00:04:17.629 that's huge. Roll -based prompting. Telling the 00:04:17.629 --> 00:04:20.660 AI who to be. changes everything. It's tone, 00:04:20.879 --> 00:04:22.839 it's knowledge, it's style. Like you could say, 00:04:23.019 --> 00:04:25.759 act as a senior strategy consultant from McKinsey. 00:04:25.939 --> 00:04:28.459 Analyze the top three mobile payment trends in 00:04:28.459 --> 00:04:30.860 Southeast Asia by 2026. Suddenly, you're getting 00:04:30.860 --> 00:04:33.339 high -level strategic thinking. Without the McKinsey 00:04:33.339 --> 00:04:36.000 bill. Exactly. Then there's example -based prompting 00:04:36.000 --> 00:04:37.939 or few -shot training. Instead of explaining 00:04:37.939 --> 00:04:40.540 a style forever, you just show it a couple of 00:04:40.540 --> 00:04:43.420 perfect examples. Like teaching by showing. Precisely. 00:04:43.420 --> 00:04:45.500 Yeah. Give it two or three marketing emails you 00:04:45.500 --> 00:04:48.199 like or articles by someone whose voice you want 00:04:48.199 --> 00:04:51.870 to mimic. It learns fast. And for really complex 00:04:51.870 --> 00:04:53.870 stuff, there's chain of thought prompting. You 00:04:53.870 --> 00:04:55.490 don't ask for the final answer straight away. 00:04:55.709 --> 00:04:58.569 You tell the AI, think step by step. Break it 00:04:58.569 --> 00:05:00.949 down for it. Yeah. Like for a business plan. 00:05:01.410 --> 00:05:04.209 First, analyze market demand. Second, identify 00:05:04.209 --> 00:05:06.949 customer segments. Third, evaluate difficulty. 00:05:07.889 --> 00:05:10.589 And so on. It forces logical thinking. You get 00:05:10.589 --> 00:05:13.420 much more thorough results. Okay, so mastering 00:05:13.420 --> 00:05:16.060 prompt engineering. It really is the matter key, 00:05:16.120 --> 00:05:18.879 isn't it? It unlocks how you use all the other 00:05:18.879 --> 00:05:21.980 AI tools effectively. It lets you be the director, 00:05:22.480 --> 00:05:25.019 which leads us perfectly into skill number two, 00:05:25.339 --> 00:05:27.769 AI automation. Once you can talk to AI, how about 00:05:27.769 --> 00:05:29.769 getting it to do all that boring, repetitive 00:05:29.769 --> 00:05:31.829 stuff for you? There was this Slack statistic 00:05:31.829 --> 00:05:34.050 recently. It said office workers spend something 00:05:34.050 --> 00:05:36.649 like 40 % of their time on low -value, repetitive 00:05:36.649 --> 00:05:39.589 tasks. 40%. Yeah, it's staggering when you think 00:05:39.589 --> 00:05:41.470 about it. It's almost two full workdays every 00:05:41.470 --> 00:05:44.009 single week just gone, wasted on stuff anyone 00:05:44.009 --> 00:05:46.430 could do, or maybe an AI could do. I know I've 00:05:46.430 --> 00:05:48.610 wasted hours on things like that. Absolutely. 00:05:48.930 --> 00:05:51.050 And that's the problem AI automation solves. 00:05:51.399 --> 00:05:54.199 It lets you build these, well, intelligence systems 00:05:54.199 --> 00:05:56.759 that just handle those tasks. Two under four 00:05:56.759 --> 00:05:59.699 seven. No getting tired, no making silly mistakes. 00:06:00.000 --> 00:06:02.319 It frees up human time and brain power for the 00:06:02.319 --> 00:06:04.980 stuff that actually matters, you know? Strategy, 00:06:05.160 --> 00:06:07.459 creativity, talking to customers. You're essentially 00:06:07.459 --> 00:06:09.600 buying back your time. And the good news is you 00:06:09.600 --> 00:06:11.759 don't need to be some coding genius to get started, 00:06:11.899 --> 00:06:13.779 right? There are tools for this. Oh, definitely. 00:06:14.100 --> 00:06:16.620 There are some great beginner -friendly options. 00:06:17.579 --> 00:06:20.360 Zapier is a classic example. Think of it like, 00:06:21.280 --> 00:06:23.079 Training wheels for automation. Training wheels, 00:06:23.079 --> 00:06:25.819 okay. Connects thousands of apps, over 6 ,000 00:06:25.819 --> 00:06:27.860 I think, with really simple trigger action logic. 00:06:28.079 --> 00:06:30.279 Like when someone fills out my website contact 00:06:30.279 --> 00:06:33.360 form, that's the trigger. Then automatically 00:06:33.360 --> 00:06:35.819 add them to MailChampagneD, send a Slack message 00:06:35.819 --> 00:06:39.439 to sales. That's the action. Simple, but incredibly 00:06:39.439 --> 00:06:41.439 useful. Yeah, I can see that saving a ton of 00:06:41.439 --> 00:06:43.819 manual work. What about more complex stuff? Well, 00:06:44.220 --> 00:06:46.079 then you might look at something like Make .com. 00:06:46.220 --> 00:06:48.480 It used to be called IntegraMath. It lets you 00:06:48.480 --> 00:06:50.560 build more sophisticated workflows, almost like 00:06:50.560 --> 00:06:52.899 visual decision trees. Decision trees? Yeah, 00:06:52.899 --> 00:06:55.600 with multiple steps and branching logic. Imagine, 00:06:55.819 --> 00:06:59.060 say, a home repair business. Trigger, job mark 00:06:59.060 --> 00:07:01.660 complete. Action one, send a thank you text. 00:07:01.860 --> 00:07:05.379 Action two, wait three days. Then action three 00:07:05.379 --> 00:07:08.240 uses a router. If they leave a five -star review, 00:07:08.699 --> 00:07:11.439 send a 10 % discount code. But TIF, the review 00:07:11.439 --> 00:07:14.100 is less than three stars, automatically create 00:07:14.100 --> 00:07:17.040 a support ticket and alert a manager. Wow, okay. 00:07:17.209 --> 00:07:19.230 That's pretty smart. Very tailored follow up. 00:07:19.490 --> 00:07:21.550 Exactly. And for people who want even more power 00:07:21.550 --> 00:07:24.089 and control, maybe over their data privacy too, 00:07:24.449 --> 00:07:27.990 there's n8n .io. It's open source, super flexible. 00:07:28.209 --> 00:07:30.170 Got it. So how do you turn this into a high paying 00:07:30.170 --> 00:07:32.850 skill? Just learn Zapier. Well, you can, but 00:07:32.850 --> 00:07:35.949 the real money, the profitable approach is specialization. 00:07:36.490 --> 00:07:38.589 Become an automation expert for a specific niche. 00:07:38.769 --> 00:07:41.759 A niche, like dentists. There were lawyers. Exactly 00:07:41.759 --> 00:07:43.480 like that. Pick an industry, dental practices, 00:07:43.660 --> 00:07:46.300 law firms, real estate agents, whatever. Then 00:07:46.300 --> 00:07:48.860 really dive deep, research their world, watch 00:07:48.860 --> 00:07:50.740 their YouTube channels, read their trade magazines, 00:07:50.860 --> 00:07:52.720 hang out in their Facebook groups. Find out what 00:07:52.720 --> 00:07:55.220 makes them tick. And find out their pain points. 00:07:55.879 --> 00:07:58.240 What repetitive tasks are driving them crazy? 00:07:58.519 --> 00:08:01.660 What takes up way too much time? The best way, 00:08:01.860 --> 00:08:04.629 just talk to them. Ask actual business owners 00:08:04.629 --> 00:08:07.329 what's the single most frustrating time -consuming 00:08:07.329 --> 00:08:09.850 task you have to deal with. Okay, so find the 00:08:09.850 --> 00:08:13.189 pain. Find the pain. Then build a specific done 00:08:13.189 --> 00:08:15.629 -for -you automation package that solves that 00:08:15.629 --> 00:08:18.410 specific pain. Not just general automation, but 00:08:18.410 --> 00:08:21.329 say a patient appointment and follow -up automation 00:08:21.329 --> 00:08:23.769 system for dental clinics. Ah, so you're selling 00:08:23.769 --> 00:08:26.990 a solution, not just a tool. Precisely. And because 00:08:26.990 --> 00:08:30.110 it's specialized, solving a known expensive problem 00:08:30.110 --> 00:08:33.019 for them, you can charge way more. Like, three 00:08:33.019 --> 00:08:34.919 to five times what a generalist freelancer might 00:08:34.919 --> 00:08:37.000 charge. You're the expert for their specific 00:08:37.000 --> 00:08:38.879 problem. That makes a lot of sense. You're saving 00:08:38.879 --> 00:08:40.820 them time, probably helping them make more money. 00:08:41.320 --> 00:08:43.759 Tangible value. Huge value. Now, moving up the 00:08:43.759 --> 00:08:46.240 ladder a bit. Skill number three, AI development. 00:08:46.559 --> 00:08:48.460 If prompt engineering and automation can get 00:08:48.460 --> 00:08:50.860 you into six figures, actually learning to build 00:08:50.860 --> 00:08:54.600 AI applications, that opens doors to, well, $200 00:08:54.600 --> 00:08:57.159 ,000, $300 ,000, potentially much more. Okay, 00:08:57.159 --> 00:08:59.580 so this is moving beyond using existing tools. 00:08:59.840 --> 00:09:02.840 Right. This is about creating custom AI solutions 00:09:02.840 --> 00:09:06.120 from maybe not scratch, but closer to it. Building 00:09:06.120 --> 00:09:08.379 applications that solve unique business problems 00:09:08.379 --> 00:09:12.799 that off the shelf tools just can't handle. So 00:09:12.799 --> 00:09:15.379 if someone wants to go down this path, what are 00:09:15.379 --> 00:09:17.379 the core technical skills they need? Where do 00:09:17.379 --> 00:09:19.980 they start? Python seems to come up a lot. Python 00:09:19.980 --> 00:09:22.659 is definitely key. It's kind of the default language 00:09:22.659 --> 00:09:25.419 for AI right now. It's known for being relatively 00:09:25.419 --> 00:09:27.220 easier to learn than some other languages, which 00:09:27.220 --> 00:09:29.779 helps. But the main thing is its ecosystem. It 00:09:29.779 --> 00:09:33.200 has these massive libraries, TensorFlow, PyTorch, 00:09:33.460 --> 00:09:35.879 Scikit -learn, that handle a lot of the really 00:09:35.879 --> 00:09:38.440 complex math and algorithms for you. Makes building 00:09:38.440 --> 00:09:41.000 powerful AI models much more accessible. OK, 00:09:41.220 --> 00:09:43.440 so Python first. What else? You mentioned APIs 00:09:43.440 --> 00:09:46.899 before. Right, AI APIs. API stands for Application 00:09:46.899 --> 00:09:49.200 Programming Interface. Think of it like a messenger. 00:09:49.299 --> 00:09:51.360 A messenger. Yeah, it lets your custom application 00:09:51.360 --> 00:09:55.289 talk to and use the power of those huge pre -trained 00:09:55.289 --> 00:09:59.309 AI models, like OpenAI's GPT -4 or Anthropix 00:09:59.309 --> 00:10:01.529 quad, without you having to build that massive 00:10:01.529 --> 00:10:03.389 model yourself. Ah, so you can plug into their 00:10:03.389 --> 00:10:06.330 brain, basically. Essentially, yes. So instead 00:10:06.330 --> 00:10:08.470 of you going to the chat GPT website to generate 00:10:08.470 --> 00:10:11.730 something, your custom app can call the API and 00:10:11.730 --> 00:10:14.250 get, say, product descriptions generated automatically 00:10:14.250 --> 00:10:16.870 for thousands of items on an e -commerce site. 00:10:16.950 --> 00:10:20.429 OK, Python, APIs, what's the third piece? This 00:10:20.429 --> 00:10:23.070 one's maybe less glamorous, but totally critical, 00:10:23.389 --> 00:10:25.820 data handling. data handling, cleaning things 00:10:25.820 --> 00:10:29.059 up. Exactly. Cleaning, organizing, structuring 00:10:29.059 --> 00:10:32.159 data. Most businesses have tons of data, but 00:10:32.159 --> 00:10:34.679 it's often messy, inconsistent, stuck in different 00:10:34.679 --> 00:10:38.080 places. Being able to prepare that data so an 00:10:38.080 --> 00:10:40.080 AI can actually understand and learn from it 00:10:40.080 --> 00:10:43.019 is a hugely valuable skill. Because you know 00:10:43.019 --> 00:10:45.879 the old saying, garbage in, garbage out. It's 00:10:45.879 --> 00:10:48.500 especially true for AI. Right. Bad data leads 00:10:48.500 --> 00:10:51.100 to bad AI results. Makes sense. How does someone 00:10:51.100 --> 00:10:53.159 learn this stuff practically? Just read books. 00:10:53.340 --> 00:10:56.340 Books help. But the best way is by doing. Start 00:10:56.340 --> 00:10:58.240 with a personal project. Build something simple 00:10:58.240 --> 00:11:01.320 for yourself, like maybe an AI tool to automatically 00:11:01.320 --> 00:11:03.419 categorize your bank statements, or one that 00:11:03.419 --> 00:11:05.559 summarizes news articles you want to read. The 00:11:05.559 --> 00:11:07.840 act of building teaches you way more. OK, build 00:11:07.840 --> 00:11:10.299 something small first. Then maybe look at platforms 00:11:10.299 --> 00:11:13.059 like Kaggle. It's a huge online community for 00:11:13.059 --> 00:11:15.080 data science and machine learning. Companies 00:11:15.080 --> 00:11:17.700 post real business problems, real data sets. 00:11:17.919 --> 00:11:21.279 and people compete to solve them using AI. Yeah. 00:11:21.419 --> 00:11:23.000 You get to practice on real world challenges, 00:11:23.419 --> 00:11:25.539 see how top developers approach problems, learn 00:11:25.539 --> 00:11:28.059 from their code. It's an amazing learning resource. 00:11:28.279 --> 00:11:30.919 So realistically, what's the learning curve like 00:11:30.919 --> 00:11:33.960 here if you're not already a coder? Are we talking 00:11:33.960 --> 00:11:37.039 months, years? It depends on your focus, obviously. 00:11:37.100 --> 00:11:40.759 Yeah. But for Python basics relevant to AI, you 00:11:40.759 --> 00:11:43.779 could get reasonably proficient in, say, a few 00:11:43.779 --> 00:11:46.600 dedicated months. Learning to use APIs comes 00:11:46.600 --> 00:11:49.220 after that. with practice. Honestly, sometimes 00:11:49.220 --> 00:11:51.700 the hardest part isn't the raw coding, is learning 00:11:51.700 --> 00:11:54.340 how to think about a business problem in a way 00:11:54.340 --> 00:11:56.840 AI can tackle it and then wrestling with that 00:11:56.840 --> 00:11:59.379 data. But the payoff is potentially huge. Definitely. 00:11:59.740 --> 00:12:02.259 You're solving complex, unique problems few others 00:12:02.259 --> 00:12:05.080 can. That commands a premium. OK, let's shift 00:12:05.080 --> 00:12:09.279 gears slightly. Skill number four, AI data analysis. 00:12:10.019 --> 00:12:12.419 You mentioned data handling. This sounds related, 00:12:12.500 --> 00:12:14.879 but different. The opportunity here seems massive. 00:12:15.340 --> 00:12:18.360 Most businesses are just like Swimming in data, 00:12:18.879 --> 00:12:21.240 right. Sales figures, website clicks, customer 00:12:21.240 --> 00:12:23.740 info. Exactly. They're sitting on a potential 00:12:23.740 --> 00:12:27.159 gold mine of insights, but most are barely scratching 00:12:27.159 --> 00:12:30.299 the surface. Traditional data analysis, the kind 00:12:30.299 --> 00:12:32.419 businesses have done for years, mostly tells 00:12:32.419 --> 00:12:33.820 you what happened. You know, sales went down 00:12:33.820 --> 00:12:37.740 5 % last month. Useful, but limited. Right, looking 00:12:37.740 --> 00:12:40.019 backwards. AI -powered analysis goes way beyond 00:12:40.019 --> 00:12:42.000 that. It can tell you what will happen. that's 00:12:42.000 --> 00:12:44.539 predictive analysis. And even better, what you 00:12:44.539 --> 00:12:46.639 should do about it that's prescriptive analysis. 00:12:46.840 --> 00:12:49.100 OK, give me an example. So instead of just sales 00:12:49.100 --> 00:12:52.700 dropped 5%, AI might say, OK, based on all these 00:12:52.700 --> 00:12:54.299 trends we're seeing, sales are likely to drop 00:12:54.299 --> 00:12:57.759 another 7 % next month. But we recommend running 00:12:57.759 --> 00:13:00.340 a targeted discount campaign aimed specifically 00:13:00.340 --> 00:13:02.879 at customer group X, because our analysis shows 00:13:02.879 --> 00:13:04.679 they have the highest probability of churning 00:13:04.679 --> 00:13:07.820 soon based on their recent behavior. Whoa. That's 00:13:07.820 --> 00:13:09.620 actionable. That's telling you exactly what lever 00:13:09.620 --> 00:13:12.809 to pull. Precisely. And the reason AI can do 00:13:12.809 --> 00:13:16.269 this is scale. The human brain just can't compute 00:13:16.269 --> 00:13:18.629 correlations across thousands of data points 00:13:18.629 --> 00:13:21.950 coming in daily from like 10 different systems. 00:13:22.289 --> 00:13:25.590 It's impossible. Too much noise. Right. But AI 00:13:25.590 --> 00:13:28.149 can analyze every single point. It finds these 00:13:28.149 --> 00:13:31.649 incredibly complex, subtle patterns that no human 00:13:31.649 --> 00:13:34.919 would ever spot, like Maybe it discovers that 00:13:34.919 --> 00:13:37.759 blog posts about topic Y only when published 00:13:37.759 --> 00:13:40.240 on Tuesday mornings consistently lead to the 00:13:40.240 --> 00:13:42.759 highest conversion rate for product Z two weeks 00:13:42.759 --> 00:13:45.519 later, but only for customers in a specific city. 00:13:45.980 --> 00:13:48.440 No way a person finds that manually. Absolutely 00:13:48.440 --> 00:13:50.899 not. Or it can predict which customer is about 00:13:50.899 --> 00:13:53.100 to leave before they even consciously decide, 00:13:53.440 --> 00:13:55.179 giving you a chance to step in and save them. 00:13:55.480 --> 00:13:57.460 So how do you get started with this skill? Do 00:13:57.460 --> 00:14:00.240 you need a PhD in statistics? Not necessarily. 00:14:00.409 --> 00:14:02.809 You definitely need to learn SQL structured query 00:14:02.809 --> 00:14:05.169 language. That's the fundamental language for 00:14:05.169 --> 00:14:07.149 talking to databases where most business data 00:14:07.149 --> 00:14:09.429 lives. It's how you ask questions of the data. 00:14:09.649 --> 00:14:11.970 Surprisingly learnable. Then you can leverage 00:14:11.970 --> 00:14:15.679 tools that already have AI built in. Platforms 00:14:15.679 --> 00:14:19.000 like Tableau or Microsoft Power BI are adding 00:14:19.000 --> 00:14:22.159 more and more AI features for forecasting, anomaly 00:14:22.159 --> 00:14:24.740 detection, deep dives, often without you needing 00:14:24.740 --> 00:14:26.940 to write tons of code. So the tools are getting 00:14:26.940 --> 00:14:29.059 smarter, too. Definitely. And you can even use 00:14:29.059 --> 00:14:32.059 the large language models themselves, like ChatGPT. 00:14:32.440 --> 00:14:34.539 Its advanced data analysis mode lets you just 00:14:34.539 --> 00:14:36.840 upload a file, like an Excel spreadsheet, and 00:14:36.840 --> 00:14:39.620 then ask questions in plain English. Analyze 00:14:39.620 --> 00:14:42.019 these sales figures and tell me the top three 00:14:42.019 --> 00:14:45.659 insights for boosting revenue. It can do a surprising 00:14:45.659 --> 00:14:47.620 amount of analysis just from that. That's pretty 00:14:47.620 --> 00:14:50.340 amazing. So an AI data analyst isn't just reporting 00:14:50.340 --> 00:14:53.100 numbers. They're actively finding ways to grow 00:14:53.100 --> 00:14:56.100 the business. Exactly. Optimizing pricing, forecasting 00:14:56.100 --> 00:14:58.600 inventory, finding profitable customer segments, 00:14:59.080 --> 00:15:01.179 personalizing marketing. It all leads directly 00:15:01.179 --> 00:15:03.519 to revenue. That's why businesses need this skill 00:15:03.519 --> 00:15:07.580 badly. OK. Moving on to still number five, AI 00:15:07.580 --> 00:15:10.600 copywriting. This feels very practical. Every 00:15:10.600 --> 00:15:13.960 single business needs words, right? Website text, 00:15:14.360 --> 00:15:17.200 emails, social media ads, video scripts. The 00:15:17.200 --> 00:15:20.000 list is endless. Copywriting is that skill of 00:15:20.000 --> 00:15:22.820 using words to get someone to do something. It's 00:15:22.820 --> 00:15:24.399 absolutely universal. It's the art and science 00:15:24.399 --> 00:15:28.250 of persuasion through text. And AI has just completely 00:15:28.250 --> 00:15:31.889 stormed into this field. Tools from the everyday 00:15:31.889 --> 00:15:35.210 chat GPT up to specialized platforms like Jasper 00:15:35.210 --> 00:15:38.490 or Copy .ai, they can generate really high quality 00:15:38.490 --> 00:15:40.929 content now and at incredible speed. And the 00:15:40.929 --> 00:15:42.629 impact on the bottom line can be huge, right? 00:15:42.669 --> 00:15:44.490 You mentioned an example. Yeah, think about a 00:15:44.490 --> 00:15:46.090 sales webinar. Let's say your current script, 00:15:46.169 --> 00:15:48.250 maybe one you wrote yourself, keeps about 20 00:15:48.250 --> 00:15:50.009 % of people engaged right through to the pitch. 00:15:50.110 --> 00:15:53.450 OK, 20 % stay. Now imagine using AI to optimize 00:15:53.450 --> 00:15:56.159 that script. testing different hooks, refining 00:15:56.159 --> 00:15:58.240 the emotional journey, tailoring the language 00:15:58.240 --> 00:16:00.659 to the specific audience you invited, maybe that 00:16:00.659 --> 00:16:03.159 AI -optimized script keeps 80 % engaged. Wow, 00:16:03.320 --> 00:16:06.019 20 % finance, 80%. That's a massive difference. 00:16:06.519 --> 00:16:10.120 Huge. Now, assume your pitch converts 10 % of 00:16:10.120 --> 00:16:13.539 engaged viewers. With the AI script, you could 00:16:13.539 --> 00:16:16.080 literally make four times the revenue from the 00:16:16.080 --> 00:16:17.980 exact same number of people who initially showed 00:16:17.980 --> 00:16:21.059 up. Four times the revenue. Just from better 00:16:21.059 --> 00:16:24.159 words. Just from better AI enhanced words. That's 00:16:24.159 --> 00:16:26.960 the power. It's directly measurable impact. But 00:16:26.960 --> 00:16:29.460 the advanced technique isn't just asking, write 00:16:29.460 --> 00:16:32.539 me an ad. It's about properly training the AI 00:16:32.539 --> 00:16:35.679 on a specific brand's voice and strategy. How 00:16:35.679 --> 00:16:37.600 do you do that? You create what's called a brand 00:16:37.600 --> 00:16:40.179 voice document. This is super detailed. It outlines 00:16:40.179 --> 00:16:42.600 the brand's mission, target customer personas, 00:16:42.799 --> 00:16:46.080 the exact tone, like are they playful, formal, 00:16:46.659 --> 00:16:49.169 edgy? It includes specific jargon they use or 00:16:49.169 --> 00:16:51.649 avoid, and critically give it examples of good 00:16:51.649 --> 00:16:53.750 and bad copy they've used before. So you're building 00:16:53.750 --> 00:16:56.549 a style guide for the AI. Exactly. A comprehensive 00:16:56.549 --> 00:16:59.549 one. Then you literally paste the entire document 00:16:59.549 --> 00:17:02.049 into the AI's context window when you give it 00:17:02.049 --> 00:17:03.929 a prompt. I see it at the whole rulebook. Yep. 00:17:04.430 --> 00:17:06.509 And then you give it a specific command, like, 00:17:06.869 --> 00:17:09.210 OK, you've memorized this whole brand voice document. 00:17:09.430 --> 00:17:11.950 Now act as our professional copywriter. Write 00:17:11.950 --> 00:17:14.329 five email subject lines for our summer sale, 00:17:14.750 --> 00:17:17.190 specifically targeting persona group A. from 00:17:17.190 --> 00:17:19.670 the document. Ah, so it's not just writing generic 00:17:19.670 --> 00:17:22.329 copy anymore. It's writing your copy. Precisely. 00:17:22.650 --> 00:17:25.589 It becomes like your own in -house copywriter, 00:17:25.990 --> 00:17:29.069 steeped in your brand. It allows for incredibly 00:17:29.069 --> 00:17:32.470 fast, consistent on -brand content creation. 00:17:32.670 --> 00:17:34.829 That's a serious advantage. You're basically 00:17:34.829 --> 00:17:37.089 equipping yourself to drive sales and engagement 00:17:37.089 --> 00:17:40.569 directly. Huge value. Absolutely. Now let's talk 00:17:40.569 --> 00:17:43.470 about skill number six, AI -assisted software 00:17:43.470 --> 00:17:45.769 development. This one feels like a game changer. 00:17:45.960 --> 00:17:48.660 For so long, if you had a great idea for an app 00:17:48.660 --> 00:17:50.779 or a tool, you either had to spend years learning 00:17:50.779 --> 00:17:53.140 to code yourself, or you had to find and pay 00:17:53.140 --> 00:17:55.460 expensive developers. That was a massive barrier. 00:17:55.599 --> 00:17:57.700 Totally. So many ideas probably died just because 00:17:57.700 --> 00:18:00.660 of that hurdle. Exactly. But now the revolution 00:18:00.660 --> 00:18:02.759 is happening. Because AI can help you build stuff 00:18:02.759 --> 00:18:04.779 now, even if you're not a hardcore coder. Yeah, 00:18:04.779 --> 00:18:07.460 that's the exciting part. Tools like GitHub Copilot. 00:18:08.359 --> 00:18:11.259 It's like having an AI programmer sitting next 00:18:11.259 --> 00:18:14.509 to you, suggesting code. sometimes whole functions 00:18:14.509 --> 00:18:17.470 as you type. It learns your style, anticipates 00:18:17.470 --> 00:18:19.809 what you need. Like an intelligent autocomplete 00:18:19.809 --> 00:18:22.589 for code? A very intelligent one. Yeah. Then 00:18:22.589 --> 00:18:24.589 you have platforms like Replet, where you can 00:18:24.589 --> 00:18:26.630 literally describe the software you want to build 00:18:26.630 --> 00:18:29.190 in plain English, and it starts generating the 00:18:29.190 --> 00:18:31.349 code for you. You just tell it what to build. 00:18:31.529 --> 00:18:34.029 Serious getting closer and closer to that and 00:18:34.029 --> 00:18:36.609 even the no code platforms like bubble which 00:18:36.609 --> 00:18:39.529 already let you build apps visually They're integrating 00:18:39.529 --> 00:18:42.549 AI now to help you design databases build complex 00:18:42.549 --> 00:18:45.289 workflows essentially making it faster and easier 00:18:45.289 --> 00:18:47.349 than ever Okay, so the barrier to building is 00:18:47.349 --> 00:18:50.490 coming way down. How does this become a high 00:18:50.490 --> 00:18:53.410 value skill? It becomes incredibly valuable when 00:18:53.410 --> 00:18:56.670 you combine this AI assistance with niche expertise 00:18:56.670 --> 00:18:59.539 Just like with automation It lets you create 00:18:59.539 --> 00:19:01.960 exclusive custom solutions. Okay, give me an 00:19:01.960 --> 00:19:05.019 example. Alright, imagine you know a lot about, 00:19:05.019 --> 00:19:07.740 say, local construction businesses. You know 00:19:07.740 --> 00:19:10.019 their biggest headache is creating accurate quotes 00:19:10.019 --> 00:19:13.359 quickly. It's manual, error -prone, custom jobs. 00:19:13.640 --> 00:19:16.140 Right, I can see that being a pain point. Using 00:19:16.140 --> 00:19:18.339 these AI -assisted tools, even if you're not 00:19:18.339 --> 00:19:20.779 a master coder, you could build them a custom 00:19:20.779 --> 00:19:23.799 quoting tool. Something simple, maybe, but tailored 00:19:23.799 --> 00:19:26.779 exactly to them. It factors in their specific 00:19:26.779 --> 00:19:29.380 material suppliers, their local labor rates, 00:19:29.559 --> 00:19:31.740 their typical profit margins. Your tool nobody 00:19:31.740 --> 00:19:34.640 else has built just for their workflow. Exactly. 00:19:34.640 --> 00:19:36.819 And that's where the value explodes. You're no 00:19:36.819 --> 00:19:38.940 longer just a freelancer charging, I don't know, 00:19:39.079 --> 00:19:42.819 $3 ,000 for a generic website tweak. You're a 00:19:42.819 --> 00:19:45.599 solutions provider offering an exclusive custom 00:19:45.599 --> 00:19:48.640 tool that solves a major business problem. Maybe 00:19:48.640 --> 00:19:51.720 you charge $15 ,000 for that tool. because they 00:19:51.720 --> 00:19:54.619 can't get it anywhere else. 15 grand, wow. Because 00:19:54.619 --> 00:19:57.460 it's unique. Unique and solves a costly problem. 00:19:57.960 --> 00:19:59.700 And here's the real kicker. There's more. Once 00:19:59.700 --> 00:20:02.720 you've built that tool for one contractor, what's 00:20:02.720 --> 00:20:04.799 stopping you from turning it into a sauce product? 00:20:05.339 --> 00:20:07.420 Software as a service. You mean sell it to other 00:20:07.420 --> 00:20:10.390 contractors too. Oh, exactly. package it up, 00:20:10.730 --> 00:20:13.029 maybe charge a monthly subscription fee and sell 00:20:13.029 --> 00:20:15.789 it to hundreds, maybe thousands of similar businesses. 00:20:16.329 --> 00:20:18.470 That's how you scale from solving one problem 00:20:18.470 --> 00:20:21.069 to building a potentially massive business. Okay, 00:20:21.150 --> 00:20:23.089 that's powerful. You're not just trading time 00:20:23.089 --> 00:20:24.809 for money anymore. You're building an asset. 00:20:24.970 --> 00:20:27.210 That's the potential. It lets you scale your 00:20:27.210 --> 00:20:30.049 impact and your income exponentially. All right. 00:20:30.089 --> 00:20:32.750 Last one on our list. Skill number seven, AI 00:20:32.750 --> 00:20:35.750 design. This feels similar to copywriting. Everyone 00:20:35.750 --> 00:20:38.990 needs visuals, right? Logos, websites, social 00:20:38.990 --> 00:20:42.150 media graphics, ad images. Businesses constantly 00:20:42.150 --> 00:20:44.569 need design work. And traditionally, that's been 00:20:44.569 --> 00:20:46.970 expensive and slow. Oh, totally. Getting a professional 00:20:46.970 --> 00:20:49.650 logo and brand identity could easily cause thousands. 00:20:50.309 --> 00:20:53.130 Getting custom graphics for ads or posts took 00:20:53.130 --> 00:20:55.849 time. Back and forth with designers, it was a 00:20:55.849 --> 00:20:58.130 bottleneck. AI has just smashed through that 00:20:58.130 --> 00:21:00.789 barrier. You can now create really professional 00:21:00.789 --> 00:21:03.150 -looking visuals, often in just minutes, even 00:21:03.150 --> 00:21:05.349 with zero design background. OK, so what kind 00:21:05.349 --> 00:21:06.809 of tools are we talking about here? How does 00:21:06.809 --> 00:21:08.529 it work? I've heard of things like Mid Journey. 00:21:09.009 --> 00:21:11.049 Mid Journey is a big one, yeah. Tools like Mid 00:21:11.049 --> 00:21:13.410 Journey, Stable Diffusion, DLE3, which is actually 00:21:13.410 --> 00:21:16.150 built into Chat Sheet BT +, now they can generate 00:21:16.150 --> 00:21:18.809 incredible images from just a text description. 00:21:19.049 --> 00:21:21.609 Need a photorealistic picture of an astronaut 00:21:21.609 --> 00:21:23.730 riding a horse on the moon? Just ask for it. 00:21:23.910 --> 00:21:26.170 Need a unique illustration for your blog post 00:21:26.170 --> 00:21:29.789 in a watercolor style? Describe it. It's amazing. 00:21:30.130 --> 00:21:32.170 You often don't need stock photos or expensive 00:21:32.170 --> 00:21:34.490 photo shoots anymore? Mind -blowing. What about 00:21:34.490 --> 00:21:37.609 more standard marketing stuff, like social media 00:21:37.609 --> 00:21:40.240 posts? For that, you've got tools like Canva, 00:21:40.440 --> 00:21:42.519 which many people already use. But Canva now 00:21:42.519 --> 00:21:45.519 has this integrated AI suite called Magic Studio. 00:21:45.759 --> 00:21:47.619 It can help you instantly generate thumbnails, 00:21:48.119 --> 00:21:50.440 resize designs for different platforms, create 00:21:50.440 --> 00:21:52.759 social media posts from a simple prompt, even 00:21:52.759 --> 00:21:55.480 whip up short promotional videos. It speeds up 00:21:55.480 --> 00:21:57.799 the workflow massively. And for more complex 00:21:57.799 --> 00:22:00.299 design like app interfaces or presentations. 00:22:00.599 --> 00:22:02.400 Yeah, there are tools there too. UEZR is one 00:22:02.400 --> 00:22:04.160 you can literally sketch an app interface on 00:22:04.160 --> 00:22:07.099 paper. take a photo, and it turns it into a clickable 00:22:07.099 --> 00:22:09.819 digital mockup. Seriously, hand -drawn sketch 00:22:09.819 --> 00:22:12.299 -to -app mockup. It's pretty wild. And tools 00:22:12.299 --> 00:22:14.259 like Gamma can take a few bullet points about 00:22:14.259 --> 00:22:16.960 your topic and generate a complete beautifully 00:22:16.960 --> 00:22:19.920 designed, multi -slide presentation in seconds. 00:22:20.519 --> 00:22:23.200 Saves hours of fiddling with layouts and colors. 00:22:23.599 --> 00:22:26.000 This is, it's democratizing design, isn't it? 00:22:26.039 --> 00:22:28.039 That's a perfect way to put it. Yeah. It really 00:22:28.039 --> 00:22:30.920 is. Mastering these tools means you can offer 00:22:30.920 --> 00:22:34.079 design services that are way faster, way cheaper, 00:22:34.339 --> 00:22:36.339 and sometimes even more creative than traditional 00:22:36.339 --> 00:22:38.680 approaches. Or, you know, you could just handle 00:22:38.680 --> 00:22:40.299 all the design for your own business yourself, 00:22:40.819 --> 00:22:42.619 saving a ton of money and bringing your vision 00:22:42.619 --> 00:22:44.720 to life directly. Incredible. So you're putting 00:22:44.720 --> 00:22:47.950 professional level visual creation into everyone's 00:22:47.950 --> 00:22:50.190 hands. Pretty much. Okay, let's just step back 00:22:50.190 --> 00:22:52.130 for a second and think about these seven skills 00:22:52.130 --> 00:22:55.490 together. Tromped engineering, automation, AI 00:22:55.490 --> 00:22:58.890 development, data analysis, copywriting, AI -assisted 00:22:58.890 --> 00:23:01.710 software development, and AI design. Not that 00:23:01.710 --> 00:23:03.789 long ago, maybe 10 years. These were essentially 00:23:03.789 --> 00:23:05.690 seven different jobs, seven different specialists 00:23:05.690 --> 00:23:08.390 you'd need to hire. Right. A developer, a designer, 00:23:08.690 --> 00:23:12.079 a writer, a data person, an ops person. Yet each 00:23:12.079 --> 00:23:14.619 requiring years of training, often demanding 00:23:14.619 --> 00:23:17.779 high salaries or expensive freelance fees. To 00:23:17.779 --> 00:23:19.680 launch a simple business idea, you might have 00:23:19.680 --> 00:23:22.460 easily been looking at, what, $200 ,000 in first 00:23:22.460 --> 00:23:25.960 -year salaries? Or maybe $50 ,000 just in initial 00:23:25.960 --> 00:23:28.539 freelance costs to get the basics built. Easily. 00:23:28.819 --> 00:23:31.099 Probably more. And here's the big shift, the 00:23:31.099 --> 00:23:35.559 real aha moment from all this. Today, one person. 00:23:35.799 --> 00:23:39.799 One single person armed with the right AI skills 00:23:39.799 --> 00:23:42.599 can realistically perform all of those roles 00:23:42.599 --> 00:23:44.900 effectively. Maybe not at the absolute highest 00:23:44.900 --> 00:23:47.140 level in every single one, but certainly well 00:23:47.140 --> 00:23:49.599 enough to build, launch, and grow a successful 00:23:49.599 --> 00:23:51.960 business. One person doing the work of seven 00:23:51.960 --> 00:23:54.809 specialists. It's unprecedented. the barrier 00:23:54.809 --> 00:23:57.170 between having a great idea and actually making 00:23:57.170 --> 00:23:59.670 it real. It has never, ever been lower than it 00:23:59.670 --> 00:24:01.569 is right now. You don't need a huge chunk of 00:24:01.569 --> 00:24:03.789 savings anymore. You don't necessarily need angel 00:24:03.789 --> 00:24:05.509 investors before you even make your first sale. 00:24:05.869 --> 00:24:07.670 You don't need to build a big team from day one. 00:24:07.829 --> 00:24:09.609 The tools are here. But there's a but, isn't 00:24:09.609 --> 00:24:11.730 there? This won't last forever, this window of 00:24:11.730 --> 00:24:14.009 opportunity. It feels like it's closing, slowly 00:24:14.009 --> 00:24:16.650 maybe, but closing. I think that's exactly right. 00:24:16.970 --> 00:24:19.069 Right now, most businesses, most people, are 00:24:19.069 --> 00:24:21.210 still just trying to figure this stuff out. They're 00:24:21.210 --> 00:24:23.150 playing catch up. They're still thinking in the 00:24:23.150 --> 00:24:26.099 old ways. The people, like you listening now, 00:24:26.440 --> 00:24:29.460 who decide to really learn and master these skills 00:24:29.460 --> 00:24:32.960 today, while everyone else is still confused, 00:24:33.519 --> 00:24:37.400 you gain this massive, almost unfair, competitive 00:24:37.400 --> 00:24:39.680 advantage. You get ahead while the getting is 00:24:39.680 --> 00:24:41.880 good, basically. Essentially, yes. And look, 00:24:41.960 --> 00:24:43.779 you don't have to become a world -class expert 00:24:43.779 --> 00:24:46.259 in all seven skills tomorrow. That's overwhelming. 00:24:46.460 --> 00:24:48.960 even just getting really good at one or two of 00:24:48.960 --> 00:24:51.339 these, especially if you combine it smartly with 00:24:51.339 --> 00:24:54.319 a niche or a specific business model, that alone 00:24:54.319 --> 00:24:57.359 can take you incredibly far. Focus is key. So 00:24:57.359 --> 00:25:00.400 the message is the tools are ready. The opportunities 00:25:00.400 --> 00:25:02.339 are literally everywhere. It just comes down 00:25:02.339 --> 00:25:03.559 to are you going to grab them? Are you going 00:25:03.559 --> 00:25:05.299 to figure this out before everyone else does? 00:25:05.559 --> 00:25:07.940 Because these skills, they're not just tech toys. 00:25:08.019 --> 00:25:10.000 They're about solving real problems, creating 00:25:10.000 --> 00:25:13.200 real value for businesses and customers. And 00:25:13.200 --> 00:25:15.740 when you can consistently do that. When you can 00:25:15.740 --> 00:25:19.480 create that value, the money just naturally follows. 00:25:19.660 --> 00:25:20.460 It really does.