Back in 2019, I was in Menlo Park, staying in an apartment right next to Facebook HQ. My flight got in late and I was starving. The only option was a small snack shop inside the apartment complex, open 24/7.
I walked in, grabbed a bag of almonds and some Pringles, and went looking for the checkout. There wasn’t one. No cashier, no kiosk, no self-checkout. Just a door.
I remember standing there for a second, trying to figure out what I was missing. I had scanned my credit card to get in, but now I was just supposed to… leave? Coming from North Carolina, this felt like I was doing something wrong. I half expected someone to knock on my door later that night asking about stolen chips.
They didn’t. My card was charged. I ate my Pringles at 10PM Pacific and moved on.
At the time, it felt like something out of the future.
What I had walked into was an early version of what became “just walk out” retail. These stores let you scan a card or app to enter, pick up whatever you want, and leave. No checkout line, no scanning items, no waiting. Behind the scenes, a mix of cameras, sensors, and software track what you take and charge you automatically.
You see versions of this now with Amazon’s Just Walk Out technology and pilots across chains like Aldi. The experience feels simple, almost invisible, even though the underlying system is anything but. And as this Guardian piece points out, sometimes there are more humans involved behind the scenes than you would expect. That’s a separate conversation.
The point is that in 2019, this felt experimental. Slightly uncomfortable. Not fully trusted. Today, it feels like a glimpse of where retail is going.
That’s what AI feels like right now.
We are still in the phase where it is impressive but inconsistent. You open a tool like ChatGPT or Claude, ask a question, and get something that almost nails it. Sometimes it saves you time. Sometimes it creates more work. Sometimes it feels like magic, and other times it feels like noise.
It is uneven, but it is not imaginary.
There is a tendency right now to compare AI to past hype cycles like crypto or NFTs. Those were driven heavily by speculation and attention. AI is different, not because it is perfect, but because people are actually using it in their daily lives.
According to Pew Research Center, awareness of AI is nearly universal, and a large share of Americans are already incorporating it into everyday tasks. Most people say they are open to using it to assist with day-to-day activities, even if they want more control over how it shows up in their lives. That combination of usage and openness is what separates a trend from a habit.
Hype can create interest, but it cannot sustain behavior. When people start building routines around a technology, you are no longer looking at a bubble. You are watching adoption.
What we are seeing in 2026 feels a lot like that late-night store in Menlo Park. The experience is not fully polished. The edges are still rough. Some of it is held together behind the scenes in ways people do not expect. But it works, and people are starting to trust that it works.
The bigger shift has not even happened yet.
Right now, most people are still using AI as something they ask questions. That is the equivalent of cautiously grabbing one item in that store to see what happens. What comes next is AI working inside real systems. Not just answering questions about your data, but operating on it. Not just suggesting content, but generating and deploying it. Not just identifying patterns, but triggering actions.
You can already see early signals of this. Nonprofits are using AI to analyze program data faster than manual reporting cycles. Design teams are using generative systems to explore directions that would have taken weeks. Developers are removing entire layers of friction from their workflows. None of this is perfect, but it is moving in a clear direction.
Forward.
Every major technology shift has an “aha” moment. The internet had it when information became instantly accessible. Mobile had it when everything moved into your pocket. AI’s moment will be quieter. It will be the point where people stop asking what it can do and start relying on it without thinking.
Is AI overhyped? Yes.
Is it a bubble? Probably not.
Because underneath the noise, something real is happening. The systems are improving, the interfaces are getting better, and people are changing how they work, even if they do not fully realize it yet.
It may not feel finished.
But neither did that store in 2019.