YOU'RE USING THE WORST AI YOU'LL EVER USE

Stolen directly from Joel Comm... if you're not familiar with Joel, he's been an early adopter of just about anything and everything that 's techie and "scary". But he's just a "Joe Schmo" who has made a boatload and influenced many simply because he started speaking about the "new stuff" first.

READ THIS ARTICLE... it may change how you view AI and why you should see the greater opportunity, while seeing the threat, too. Both can be true.

_ _ _ _

In 2004, I was carrying a Palm Treo 650 in my pocket. Let me tell you, I felt like I had space-age technology. Email on a phone. A tiny web browser. A camera that took pictures you could almost identify. I showed it to my friends still carrying a flip-phone and they looked at me like I’d brought fire down from the mountain.

That phone now sits in a dump somewhere, and my iPhone could run circles around it while simultaneously ordering dinner and translating a conversation in Mandarin. The Treo wasn’t just outdated. It became unrecognizable as a “smartphone” by any modern standard.

I’m thrilled to welcome new subscribers every day! Welcome to those of you who recently joined as paid subscribers. And greetings to those of you who enjoy my writing and want to be the next to support!

I keep thinking about that Treo. I did love that device.

We Never See It While We’re In It

I’ve been building things on the internet since 1995. That’s six technology revolutions: the web, search, social media, mobile, blockchain, and now AI. And every single one of them had what I call a “Treo moment.” A point where the technology felt so advanced, so clearly the pinnacle, that you couldn’t imagine what could possibly come next.

The first time a web page loaded over a dial-up connection, I thought I was witnessing the future of communication. The first time Google actually found what I was looking for, it felt like someone had organized the entire internet just for me. The first time I posted something online and a complete stranger on the other side of the world responded, it was electric.

Every one of those moments felt like a ceiling.

Not a single one of them was.

Welcome to the Treo Era of AI

Anthropic recently released Fable 5. People are losing their minds. And they should be. This thing can sustain complex tasks for hours. It one-shots applications that would have taken a team of developers weeks to build. It reasons through problems in ways that feel less like software and more like working with a brilliant colleague.

It’s astonishing. I won’t pretend otherwise.

But I can’t stop thinking about something: we are in the Treo 650 era of artificial intelligence.

Right now, in June 2026, we’re marveling at the equivalent of a 320x320 pixel screen and calling it the future. We’re typing carefully worded prompts into text boxes and celebrating when the AI understands what we mean. We’re copying and pasting outputs from one tool into another. We’re “prompt engineering” like it’s a professional discipline.

And from the perspective of 2036? All of that is going to look adorable.

The Things That Will Make Them Laugh

Close your eyes for a second and imagine someone ten years from now looking back at how we use AI today.

We type our requests. With our own hands.

We craft multi-paragraph prompts because the AI needs us to be really specific or it’ll go sideways. We sit and watch a loading indicator while the model “thinks.” We get a wall of text back and then manually figure out what to do with it.

We celebrate when AI writes a decent email. We post screenshots when it generates a working game on the first try. We write entire LinkedIn posts about how we got ChatGPT to help us with a spreadsheet.

People in 2036 aren’t going to be impressed by any of that. They’ll look at it the way we look at someone excitedly showing off their Treo’s ability to display a pixelated image of their dog.

My mother is 86 years old. She’s not a technologist. But within the next few years, she’ll be able to sit on her couch and say, “I want to watch an original sitcom starring Lucille Ball and Jerry Seinfeld,” and the AI will create it. Not a janky deepfake. A full show. Custom for her.

Granted, there are still copyright issues to be overcome. But I’m sure actors will be thrilled to get paid to do nothing but allow their likeness.

Mom will ask for an app that tracks her medication and reminds her to call her grandchildren, and it’ll just... exist. No prompting skills required. No understanding of how any of it works under the hood.

That’s not science fiction. That’s the trajectory we’re already on.

What We Always Get Wrong

Six revolutions have taught me something about technological leaps: we don’t just underestimate the improvement. We can’t even imagine the categories.

In 2004, you couldn’t have described TikTok to someone holding a Treo. Not because the technology was too advanced to explain. Because the behavior was. Short-form vertical video shot on phones and algorithmically distributed to millions? The concept didn’t exist. There was no framework to even hang it on.

That same blind spot exists right now with AI.

The most transformative use cases of 2036 probably don’t have names yet. We’re not just going to get better chatbots and faster image generators. We’re going to get entirely new categories of interaction that we don’t have words for today.

I’ve watched this happen over and over. The web didn’t just replace the Yellow Pages. Search didn’t just organize websites. Social media didn’t just let you email more people at once. Mobile didn’t just shrink the laptop. Each revolution created behaviors and entire industries that were inconceivable in the previous era.

AI will be no different.

Why This Should Excite You, Not Scare You

As an AI keynote speaker, I talk to audiences all over the world about this. The number one emotion I encounter isn’t excitement. It’s anxiety. People are overwhelmed. They feel behind. They think the AI train left the station and they weren’t on it.

I want you to understand something: if the best AI we have right now is the worst AI we’ll ever have, then there is no “behind.” The platform hasn’t even stabilized yet. We’re all still figuring out what this thing is for.

This is the Doubt phase of what I call the Disruption Confidence Cycle. Every technology revolution follows the same five stages: Disruption, Doubt, Clarity, Confidence, and Momentum. Right now, most people and most organizations are sitting squarely in Doubt. They see the disruption. They know it’s real. But they can’t see the path forward clearly enough to move with confidence.

That’s normal. That’s the pattern. And it’s temporary.

The people who leaned into the web in 1996 weren’t smarter than everyone else. They were just willing to be bad at something new while it was still clunky. The people who built real businesses on social media in 2008 didn’t have a crystal ball. They just didn’t wait for the technology to be perfect before they started experimenting.

The same window is open right now. The AI you’re using today, the one that feels simultaneously magical and frustrating, is your Treo 650. It’s the starting line, not the finish line.

The Worst Is Also the Best

This idea has a flip side that I love. Yes, the AI we have today is the worst we’ll ever have. But it’s also the best AI that has ever existed. Both are true at the same time.

You’re standing at a very specific point in history. The technology is good enough to be transformative but early enough that most people haven’t figured out how to use it yet. That’s the sweet spot. That’s always been the sweet spot.

Twenty years from now, someone will read this article and smile at how quaint it all sounds. They’ll chuckle at the idea that we typed prompts. They’ll find it charming that we celebrated one-shotting a basic game. They’ll wonder how we got anything done with tools this primitive.

And that’s exactly the point.

You’re using the worst AI you’ll ever use. That’s not an insult.

It’s a promise.

3 comments

YOU'RE USING THE WORST AI YOU'LL EVER USE