Web3 Was a Distraction. AI Is Not.

Key Points

Web3 promised solutions to problems that didn’t exist. AI solves real problems you can use today.
AI shows immediate, measurable ROI across writing, coding, analysis, and business operations. Web3 peaked then plateaued.
I’m betting my businesses on AI because it compounds every month. The competitive advantage goes to those building with it now.

I watched the Web3 boom from the sidelines. While everyone else was arguing about decentralization and mining rigs, I was skeptical. Not because I’m anti-technology or afraid of change. I turned down millions in crypto-related opportunities because Web3 failed my core test: does it solve a real problem people actually have?

Flash forward to today, and AI is nothing like Web3. I’m not just watching from the sidelines anymore. I’m all in. Here’s why the comparison matters and what I’ve learned about betting on technology.

Web3 Was a Solution Looking for a Problem

The crypto narrative was seductive. Decentralize everything. Cut out the middleman. Give power back to the people. It sounded revolutionary, and for a moment, everyone believed it might reshape the internet.

Except it didn’t solve anything people actually needed solved.

Yes, there are legitimate use cases for blockchain—escrow, transparent ledgers, certain financial applications. But the Web3 hype cycle wasn’t about solving those specific problems. It was about a technology in search of justification. The hype came first. The practical applications came later, if at all. Marc Andreessen was bullish on crypto, but even VCs struggled to point to products people genuinely wanted versus products they needed to justify their investments in.

I looked at every pitch. Decentralized social networks that nobody used. NFTs that turned out to be speculative assets, not utility. DeFi platforms that mostly served as casinos for people trying to get rich quick. The fundamental issue was that these solutions created complexity, cost friction, and inconvenience without delivering proportional value.

Web3 had hype. It didn’t have staying power because it didn’t have real, everyday utility.

AI Is Different—Radically Different

AI isn’t a solution looking for a problem. It’s a solution finding thousands of problems simultaneously, and it’s solving them right now.

I can sit down today and use AI to write an email in seconds that would take me 15 minutes. I can paste code with bugs and get fixes instantly. I can analyze business data, brainstorm product ideas, draft contracts, research competitors—all in hours instead of weeks. This isn’t theoretical. It’s not “coming soon.” It’s here, and it works.

The difference is almost embarrassing in retrospect. With Web3, the pitch was always about what could be possible in some future state. With AI, the pitch is about what you can do right now. That’s a fundamental gap between hype that’s hollow and hype that’s grounded in reality.

The Hype Cycle Reveals the Truth

Both Web3 and AI had massive hype cycles. Both got written about constantly. Both attracted billions in venture capital and caught the imagination of technologists everywhere.

Here’s what happened next: Web3 peaked, and adoption flatlined. The hype deflated, and the actual use cases turned out to be niche. Blockchain is a legitimate technology, but the Web3 movement was always more about the hype than about the problem-solving.

AI, meanwhile, is still accelerating. Every month, the models get better. Every month, new applications emerge that seemed impossible six months ago. Anthropic released Claude Opus in early 2024, and it’s noticeably better than the models from just months before. Adoption isn’t decelerating—it’s widening across industries, company sizes, and use cases.

Carlota Perez writes about technological revolutions in her book “Technological Revolutions and Financial Capital.” She describes how real transformative technologies move through phases of enthusiasm, then correction, then sustained deployment. AI is still in the enthusiasm phase, sure, but the underlying technology is genuinely becoming more powerful, not just more hyped.

Why I’m Betting My Businesses on This

I run multiple businesses: Rotate (our creative agency), advisory work with founders, this newsletter, side projects like Openmark and Refract. I’m now integrating AI into every single one of them.

At Rotate, we use AI to accelerate our creative process, from brainstorming campaigns to drafting copy to analyzing competitor positioning. In my consulting practice, I use AI to synthesize research and generate frameworks that would take me 10x longer without it. For my writing, I use AI to help me think through ideas, structure arguments, and catch places where my writing gets muddled.

The ROI is measurable. My team moves faster. I can take on more projects. The quality of output hasn’t decreased—it’s actually improved because I can spend less time on grunt work and more time on judgment calls.

With Web3, the ROI was always theoretical. Nobody was building businesses on top of crypto networks that generated real revenue. With AI, I’m literally making money because I’m using these tools better than my competitors are.

What Warren Buffett Would Say

Warren Buffett’s investment approach is famously conservative: only invest in things you deeply understand. He’s been skeptical of technology investing for this reason. But here’s what’s interesting—Buffett’s principle actually supports betting on AI.

AI isn’t hard to understand anymore. You can use it. You can see what it does and doesn’t do well. You don’t need to be a researcher or a PhD to grasp its utility. You can literally try it today for free and make a judgment about whether it creates value for your life or your business.

That’s the opposite of Web3. Web3 required you to understand complex systems, buy into speculative narratives, and trust that the technology would eventually create value. AI requires you to try it and see the value immediately.

The Lesson Is About Real Problems

I’ve spent my career working with entrepreneurs and technologists. I’ve seen hype cycles come and go. The pattern is always the same: technologies that solve problems people actually have stick around. Technologies that create problems in search of solutions eventually fade.

Web3 was elegant. It was visionary. But it was solution-first, problem-second. AI is problem-first, solution-second. The ordering matters enormously.

If you want to invest your time, capital, or career into a technology, ask yourself: can I use this to do something I currently can’t do? Can I do it without friction and without jumping through hoops? Does it work today, or am I betting on it working someday?

For AI, the answers are yes, yes, and it’s already working.

What’s Next

We’re only a couple of years into the AI boom. The models are already getting better at reasoning, at understanding context, at following complex instructions. The next 24-36 months will see AI move from “cool tool I tinker with” to “critical infrastructure for how I work.”

The businesses that will win are the ones building with AI now. Not just using it, but thinking deeply about how it changes their entire operation. The founders who learn to prompt well, who understand the limitations, who build AI into their product and process—those are the people who’ll have an insurmountable advantage over competitors who treat it as a toy.

I don’t know if you should bet on AI. I know I am, because the problems it solves are problems I can see and touch every single day. That’s enough for me.

Further Reading: