Jesse Cunningham Reviews the Risks of Claude Code and Fable

0:00People are going to look back and say, wait, you had access to frontier AI models for like 20 or 100 bucks a month and you didn't build anything with them. Right now, AI is hugely, massively underpriced. So these companies, they're subsidizing our usage. Me and you get to use these because they allow us, right? They are letting normal people access billion-dollar models, AI models for prices that make no sense compared to the value that they can produce.

0:30And I think this window, this opportunity for you and me right now is closing. So over the weekend, I was talking to an old friend, right? And he's an engineer turned programmer. And then that's kind of like exactly the type of person who actually uses AI in real work. And he told me his company just sent out an email saying that they need to reconsider AI usage across their entire company because it's becoming too expensive. And then you saw the same thing with Claude and Fable recently. They gave users kind of a taste of something insanely powerful.

1:04Did you use Fable? It's crazy. People loved it, right? Then the limits changed and access changed and pricing changed because the compute behind these tools, it's not free. It costs them money. They just give us a taste. So this is the pattern, right? Give people a taste of it, prove the value, then separate the casual user from the power user. Now at that same party, I talked to another buddy of mine. I've known for like 30 plus years and he's been a programmer, a real programmer for 15 plus years. And he was not excited. Mind you, this guy, this person I'm talking about, super smart, was working on AI models well before it was like

1:46common nomenclature to people like me and you. AI has been around for some time in different forms. And he was working on it. And he is not excited right now. And he was kind of beat down by the whole thing because he's watching something very strange happen in real time. Because for years, programming was one of those protected skills. It was. It was a protected skill. At least that's how it was perceived in my generation. You go to college, you become a computer scientist, major, right? And you can go work for Facebook, Meta, whoever, and you'll have a job forever. You'll

2:18make tons of money, right? You had to know the language. You had to know the tools. You had to understand all these different things. You had to sit there and grind it out and find the bugs and Google this weird syntax era, right? But those days are over, right? Because now a normal person can open up Fable or Opus, Claude, whichever model you want to use, describe an idea and get pretty darn close to something real. Now with all this, I think people are massively underreacting to the moment because AI is a ticking time bomb. AI is definitely a ticking

2:52time bomb. You may have a few months. You may have a few years. I don't really know exactly how long the window stays open, but there is no world where the best AI models, right? The best coding agents, the best video tools, et cetera, all stay casually available to normal people at these prices. It's just not going to happen, right? And things are already showing up. So the Wall Street Journal just reported that OpenAI, Anthropic, Google, Microsoft, and AWS, they're handing out huge amounts of free or discounted AI compute credits to startup companies, right? In some cases, we are talking about offers worth millions of dollars. They're eating the cost to win the

3:34customer, which is something that's been going on for decades. Think of Netflix, Amazon, all these big companies hyperscale, they lose money, and then they turn the screw, right? But somebody eventually has to pay. Somebody has to pay for the GPUs. Who's going to pay for the data centers, right? Somebody has to pay for the electricity, for the infrastructure, right? Rooters just reported that big hyperscalers like Alphabet, Amazon, Meta, Microsoft, are expected to spend around $4.8 trillion in capital expenditures from 2026 now to 2030. That is humongous. That's the part where people do not see when they pay, you know, 20 bucks a month. You're not just using like an app.

4:16You are touching this massive global infrastructure build out that costs an insane amount of money. And I think Fable is the clearest preview of how this plays out, right? Anthropic literally said Fable 5 would be included for up to 50% of weekly usage limits through July 7th. We just did that. It was awesome. I used it to build crazy things. And then after that, it moves to usage credits.

4:38Have you used the usage credits? It's crazy. Super expensive. And that's the pattern. It gives people a taste. It kind of proves the value. And then what? Are you going to use it or aren't you? They're going to compress this whole thing. All this to say, I think this is the most important message that I think there can be taken from this is right now, these companies are still subsidizing the cost. And you're getting access to tools that may never, ever, ever be cheap again.

5:07You are getting literal subsidized reps, right? That's the phrase I think we should use. Subsidized repetitions, subsidized reps. So every time you use these tools seriously, you're not really just getting the answer or getting the tool. You're practicing how to direct their intelligence. You're not necessarily practicing how to direct your own intelligence. And that's a scary proposition because it's their tool. They can keep it if they want. And if you're pushing it around and that's all you use every day, you're kind of fatiguing your own brain's ability. You're learning how to turn a vague idea into a prompt and then that prompt into a draft, into a prototype,

5:50into a system, into an actual outcome, but you are reliant on them. And that's scary. So to me, the takeaway is this. I think if you're someone like me who makes money online, I mean, I've been an SEO, now a AEO, now I build microsites, now I do this, that, and another. My websites, they get tons of traffic using AI methods. Okay, great. I see that we have a small little window of time to make money still. I really do. I think it's so small. And a lot of my friends feel the same way you're in the same profession. You have a small time frame where the compression is coming,

6:29right? The programmers, they're feeling it already. They're realizing they're being commoditized. But people, normal people that were not programmers are excited. Yoo-hoo, this is amazing. But the compression is coming, the floor is going to fall out, and you're going to wish you made money these last few years. And I want to be clear, my cohort, my friend group, myself, we've made more money now than ever, than ever in the history of our lives, because we can be micro hyper scalers. But the future is not that everyone gets access to these models forever, right? The future is cost control. It's tears, it's credits, it's madness, because these are

7:06for-profit companies that are going to restrict everything. So my encouragement to you, and this is what my school community is all about, find the intersection of money where AI can hyperscale. Oftentimes with us, it's local businesses, right? Local businesses have real world products, therefore you have a hedge, a protection against what AI does to the internet and what have you. Take a plumbing company, take a roofing company, hyperscale its ability through marketing, through AI. It's really that simple. Take boring things and hyperscale them. At your local level, that's going to be an opportunity for some time. But a lot of industries, a lot of people,

7:48they're going to be surprised at how compressed this time is going to be, how compressed all of this is going to be. So the programmers are freaking out. They can already read the writing on the wall. The normal people are still excited because now we have access to be like programmers, we're not really programmers. There's still tons of money to be made, but I think it's a whole lie.

8:10The AI, the hype of AI is kind of like a lie because underneath of it all, it's like a foundation that's going to be crumbling in some regard because eventually they have to monetize.