We’ve Seen This Movie Before: What the Dot-Com Era Can Teach Us About AI

Author: Javier de la Sierra, Director of Software Development, DecisivEdge

Opening Scene: The Dot-Com Boom

I was about a year into my career when the dot-com bubble began to burst.

At the time, I was working for a consulting company. I was still relatively junior, and I remember watching the effects work their way through the technology industry. Companies disappeared. Investments dried up. People lost jobs.

More than 25 years later, when I look at what is happening with AI, some of it feels very familiar – especially with junior resources bearing the brunt of the impact.

There is enormous excitement around a new technology. There is enormous investment. And there is a fear among companies that if they don’t move quickly enough, they are going to miss out.

But there is an important distinction to make: the dot-com bubble burst. The internet did not.

That matters greatly when we talk about AI today.

Act I: The Internet Wasn’t the Problem

During the dot-com boom, putting “.com” after a company name could suddenly make it cool to investors.

There was a belief that everything needed to go online. Money flowed into companies because they were associated with the internet, sometimes without enough attention being paid to a much more basic question:

What problem does this solve, and what value does it create?

A lot of the investment wasn’t rooted in a clear cost-benefit analysis. It was rooted in FOMO.

Of course you can say – but Javier, some companies were solving real problems. They understood how the internet could allow them to provide something useful in a way that hadn’t been possible before. Many of those businesses survived and became enormously successful.

To that I say yes, but a lot of others didn’t.

The lesson wasn’t that the internet had been overhyped. The internet fundamentally changed how we communicate, shop, work and run businesses.

The lesson was that a transformational technology can be incredibly valuable without every application of it being valuable.

Act II: AI and the Return of FOMO

There is a similar sense of urgency around AI.

Companies are being told they need an AI strategy. Just open a newspaper, your Instagram, a simple Google search – it is everywhere. Investors are looking for AI companies.

On paper, AI and business looks like a marriage made in heaven.

And look, some of that excitement is justified. AI can already do things that would have been extremely difficult, expensive or simply impossible with previous technology.

But “We need to do something with AI!!” is NOT a business case.

The questions to ask before diving into AI are:

1. What pain are you trying to address?
2. What opportunity are you trying to capture?
3. Can AI help you do something better than you could before?

The opportunity comes when you have a real problem to solve and the technology gives you a better way to solve it – one that wasn’t available before.

That requires a serious amount of discipline, especially when everyone around you seems to be adopting AI in some shape or form.

FOMO can tell you where to look, certainly. But, that feeling shouldn’t dictate where to invest your time and money.

Act III: When Knowing AI Isn’t Enough

There are two kinds of knowledge that need to come together for AI to create meaningful value.

You need to understand the technology. And you need to understand the business.

We saw a version of this during the early internet era. The people who knew how to build websites understood the new technology, but they didn’t necessarily understand the intricate challenges of a particular industry. Meanwhile, the people who knew those industries well didn’t necessarily understand what the new technology could do.

The real opportunity comes when those worlds collide.

I think about something as simple as a hammer. (Yes, I use home renovation metaphors a lot. Check out my other articles if you don’t believe me. 😊)

A general-purpose hammer works perfectly well for a lot of jobs. But someone who frames houses all day understands problems with that hammer that someone outside the trade may never notice: The weight matters. The size of the head matters. A smooth surface can slip when it hits a nail. Even having to hold the nail while swinging the hammer presents a problem.

Those observations eventually lead to a tool designed around the job: a framing hammer. One with a heavier head, serrated face and (sometimes) a magnet.

None of those improvements comes from simply asking how to make a more advanced hammer; they come from understanding framing.

AI is obviously far more complex, but the gist remains the same – that the most useful applications of AI will come from one, understanding what the technology can do and two, people who understand the work well enough to know where the technology can actually help.

Act IV: Maybe We’re Still Recording Theatre?

I recently watched a talk by Jack Conte, CEO of Patreon, about AI and the arts. He drew a comparison to the early days of the movie industry that stuck with me.

When movies first appeared, one of the earliest ways people used the technology was essentially to record theater. That made sense. Theater was what people knew, so they used the new technology to reproduce something that already existed.

Then people began experimenting. Film became its own medium, creating new forms of entertainment, businesses and jobs that would have been difficult to imagine at the beginning.

I think we’re still in that stage with AI.

Much of what we’re doing today is asking AI to do existing work faster: write this document, summarize this meeting, generate this image, write this code.

Those are useful applications. But as Conte’s example made me think, they may also be our version of recording theater.

With foundational technologies, it takes time to move beyond improving what we already do and discover what could only exist because of the new technology.

Final Scene: Keep Tinkering with It

So how do you avoid getting swept up in the hype without falling behind?

Experiment.

Get familiar with the technology. Try different tools against problems you understand well. See where AI can save time, improve quality or help you do something you couldn’t do before.

But keep applying critical thinking.

If something takes ten hours and costs $100 today, and AI can do it in five hours but costs $200 in tokens, think twice. Maybe the time savings justify it. Maybe they don’t. That’s the analysis businesses need to do.

And if the business case isn’t there today, keep learning. AI is changing quickly. Try the tools, understand what they can do, and come back to them as they evolve.

—

We’ve seen this movie before. The dot-com crash didn’t mark the end of the internet. In many ways, it was only the beginning of figuring out what the internet was actually good for.

AI may turn out to be the same. We’re still in the opening scenes.

So keep watching. Keep experimenting. Just make sure you understand why you’re buying a ticket.