The More We Use AI, The More We Trust It
I recently took a family trip to Paris. It was everything you would expect. Sunlight pouring into cafes, croissants every morning, museums packed with centuries of human creativity. I feel incredibly fortunate to have had the opportunity to go.
The flight over from North Carolina was about seven hours. The flight back from Paris was closer to nine. I never sleep well on planes, so I had a lot of time to sit there, stare into the darkness somewhere over the Atlantic, and observe the people around me.
A few passengers had laptops open and were working for most of the flight. One thing I noticed almost immediately was how many of them had ChatGPT open in their browser tabs. Not one or two people. A lot of them.
One person was planning an itinerary for their trip. Another was comparing a collection of photos to decide which ones to post on Instagram. Someone else appeared to be rewriting work emails. Every few rows, there it was again. The now familiar white chat window.
Like it or not, the common use cases for AI are already solidified.
For many people, ChatGPT has quietly become the new Google. Not in the sense that it replaces search entirely, but in the sense that it has become the first place people go when they want an answer. Questions, recommendations, summaries, comparisons, travel ideas, writing help. The interaction feels faster and more natural than traditional search, so people keep coming back to it.
I saw this firsthand during the trip itself.
A friend traveling with us used ChatGPT to find restaurants in Paris. One recommendation in particular stood out to them: an Asian dim sum spot in the Republique district. They were absolutely convinced it was going to be incredible. Not just decent. Top tier. Worth building dinner plans around.
So we went.
We ordered takeaway and brought it back to eat.
The food was… fine.
Not bad. Not amazing. Just OK.
What surprised me was not that the recommendation missed. Restaurant recommendations are subjective and Paris has thousands of options. What surprised me was how certain my friend was beforehand. There was no hesitation, no second guessing, no cross-checking reviews or asking locals. ChatGPT had recommended it, so in their mind it was already validated.
That stuck with me.
There is an interesting tension happening right now with AI systems. The more people use them, the more they trust them. Familiarity creates confidence. If a system helps you draft emails, summarize documents, plan vacations, and answer random questions every single day, your brain slowly starts categorizing it as reliable.
But reliability and correctness are not the same thing.
At the end of the day, these systems are still probabilistic. They are generating responses based on patterns, training data, and context windows. Sometimes the outputs are incredibly useful. Sometimes they are wrong in subtle ways that are difficult to detect unless you already know the answer yourself.
That is where things get complicated.
For years, people approached search engines with a degree of skepticism. You searched, scanned sources, compared links, and made a judgment call. AI changes the interaction completely because it collapses everything into a single confident answer. The interface itself feels authoritative.
That confidence is powerful.
And honestly, most of the time, the systems are useful enough to reinforce the trust. That reinforcement loop is what matters. The AI helps you enough times that eventually you stop verifying as often.
I do not think this is necessarily a bad thing. Trust is part of how humans interact with every tool and every system. We trust GPS directions without memorizing roads. We trust calculators without redoing the math by hand. Eventually, AI will settle into a similar place in society.
But we are still very early in that transition, and I think people underestimate how much critical thinking still matters when interacting with these systems.
Especially now.
Because AI does not just answer questions anymore. It recommends. It persuades. It prioritizes. It frames decisions in ways that subtly shape human behavior. Sometimes correctly. Sometimes not.
The dim sum was not life changing, but the experience felt oddly symbolic.
A confident recommendation from a machine was enough to override uncertainty entirely.
That is probably the bigger story unfolding right now.
Not whether AI is intelligent, but how quickly humans are learning to trust it.