I was at the supermarket the other day, looking to buy some bread. I’d made the mistake of actually reading the ingredients on a loaf of what was marketed as “artisanal sourdough.” Sugar. The third ingredient. In sourdough bread. I put it back and reached for another brand. Sugar again. And another. Same story.
It got me thinking about how we got here, to this place where sugar is in everything from bread to salad dressing to cured meats. It’s such an integral part of our food supply that removing it requires genuine effort and vigilance. But it wasn’t always this way. Sugar has a history, and that history has some uncomfortable parallels to what we’re experiencing right now with artificial intelligence.
When Sweet Was Dear #
Let me take you back a few centuries. In medieval Europe, sugar was a luxury reserved for the extraordinarily wealthy. It was so expensive that it was kept locked away, displayed in elaborate sugar sculptures at royal banquets as a demonstration of wealth and power. Common people sweetened their food with honey if they could afford it, or simply went without.
The transformation began with colonialism and the triangular trade. Sugar plantations in the Caribbean and Americas, built on the backs of enslaved people, began producing sugar at scale. By the 18th century, what had once been a luxury was becoming affordable to the middle classes. By the 19th century, it was everywhere.
Here’s what’s fascinating: nobody really understood what sugar was doing to us. It was sweet, it was energy-dense, it was (finally) cheap. Why wouldn’t you add it to everything? Manufacturers discovered that a little sugar made just about any food more palatable, more shelf-stable, more profitable. It became the magic ingredient that could make people come back for more.
The food industry built empires on sugar. They added it to cereals, to bread, to sauces, to drinks. They found ways to hide it in plain sight with names like “high fructose corn syrup,” “dextrose,” “maltodextrin.” By the time the medical establishment began to understand the connection between excessive sugar consumption and obesity, diabetes, heart disease, and a host of other health problems, it was too late. Sugar was everywhere. It was in everything. The infrastructure was built. The supply chains were established. The consumer expectations were set.
You can’t just remove sugar from the food supply. The entire system is built around it.
The New Addiction #
Now, let’s talk about artificial intelligence.
We’re watching, in real time, another luxury become ubiquitous. AI started as something extraordinary, expensive, and exclusive - the domain of research labs and tech giants. Then came the democratization. ChatGPT, Copilot, Gemini, Claude, and countless others. Suddenly, AI was everywhere.
But here’s the thing that keeps me up at night: we’re adding AI to everything without really understanding what it’s doing to us.
Every app now has an AI assistant. Every word processor has AI writing help. Every spreadsheet has AI analysis. Every email client has AI composition. Every search engine has AI summaries. Heck, NOTEPAD has Copilot! We’re reaching a point where it’s harder to not use AI than it is to use it. Just like sugar in bread.
And just like sugar, everyone’s using it not because they necessarily need it, but because… well, because it’s there. Because it’s easy. Because it gives us that quick hit of productivity dopamine. Because our competitors are using it, and we can’t be left behind.
I watch friends use AI to write emails that used to take them two minutes. I see students using AI to write essays they could have written themselves. I observe organizations implementing AI solutions to problems that didn’t require AI to solve, and in most cases aren’t the problems that need solving in the first place. We’re not asking “should we use AI for this?” We’re asking “how can we use AI for this?”
Nobody stopped to question what problems we’re trying to solve. There’s just adoption. Rapid, wholesale, unquestioning adoption.
What We’re Losing #
The medical establishment eventually figured out that sugar was metabolically damaging. It took decades, and by then, the damage was done to both individual health and to the entire food system.
I worry we’re not paying attention to what AI is doing to our cognitive metabolism.
Every time we use AI to write something we could have written, we’re not practicing writing. Every time we use AI to solve a problem we could have solved, we’re not practicing problem-solving. Every time we use AI to research something we could have researched, we’re not practicing research.
Literacy isn’t just about reading and writing. It’s about thinking. It’s about wrestling with ideas until they make sense. It’s about the productive struggle of putting thoughts into words. When AI does that work for us, what happens to our ability to do it ourselves?
Critical thinking isn’t a skill you maintain by watching AI think. It’s a skill you maintain by doing the thinking yourself. And we’re offloading more and more of that thinking to machines that don’t actually think - they pattern-match at scale. There is no understanding in artificial intelligence, only pattern matching.
I’m not talking about using AI as a tool for specific, well-defined tasks. I use tools all the time. I use calculators. I use spell checkers. I use search engines. Tools are fine when they augment our capabilities without replacing them. Keyword being: “augment”.
But we’re not using AI as a tool. We’re using it as a replacement. We’re letting it do the work we should be doing to keep our minds sharp. I mean, you can’t learn how to swim by reading about it.
By Then It Was Already Too Late #
The thing about sugar is that we can’t easily undo it. Yes, you can choose to eat less sugar. You can read labels. You can cook from scratch. But you’re fighting against an entire infrastructure that assumes sugar is fundamental. The recipes assume it. The manufacturing assumes it. The supply chains assume it. The consumer palates expect it.
We’re building the same kind of infrastructure around AI. We’re training a generation that expects AI to be there, that doesn’t know how to function without it. We’re creating dependencies.
The students who use AI to write their essays today will be the professionals who can’t write clear reports tomorrow. The programmers who rely on Copilot to write their code today will be the architects who can’t think through complex system designs tomorrow. The analysts who let AI summarize their data today will be the leaders who can’t spot the faulty logic in an AI-generated strategy tomorrow.
And by the time we fully understand what we’ve done - by the time the cognitive equivalent of the obesity epidemic becomes undeniable - it will be too late to easily reverse course. The tools will be embedded. The workflows will be established. The expectations will be set.
What Now? #
I don’t have easy answers. I don’t think we either can or should go back to a world without AI, just like we can’t or shouldn’t go back to a world without sugar. Both have legitimate uses. Both can be valuable in the right contexts.
But I think we need to be a hell of a lot more thoughtful about when and how we use AI. We need to ask ourselves: am I using this because it genuinely helps me do something I couldn’t otherwise do, or am I using it because it’s easy? Am I preserving my ability to think and write and solve problems, or am I outsourcing those capabilities?
We need to be conscious consumers of AI, the same way some of us have learned to be conscious consumers of sugar. We need to read the labels. We need to understand what we’re putting into our workflows, our organizations, our minds.
Most importantly, we need to maintain the skills that AI can erode. We need to deliberately practice writing, even when AI could write it faster. We need to deliberately practice thinking, even when AI could think it quicker. We need to deliberately practice learning, even when AI could learn it better.
Because unlike sugar, which affects our bodies, AI affects something even more fundamental: our minds.
And we’re only going to get one shot at this. By the time we realize we’ve built a society that can’t think without AI assistance, it will be too late to rebuild those cognitive capabilities. The sweetness of easy answers will have already poisoned the well.
I’m going to keep working with data and AI - it’s what I do. But I’m also going to keep writing my own thoughts, solving my own problems, and doing my own thinking. Not because it’s efficient, but because it’s essential.
The bread without sugar is harder to find, but it’s worth the effort.
The work without AI assistance takes longer, but it’s worth doing.
We still have time to be intentional about this. But not much.
Join the Conversation #
What actual use cases do you have for AI? I’d love to hear of use cases where AI solved problems that could not otherwise be overcome. Please reach out to me or comment on LinkedIn or BlueSky!
Photo by Shay Wood: https://www.pexels.com/photo/pile-of-pancake-with-honey-574111/