AI Misconceptions: What People Get Wrong About Artificial Intelligence

AI: love it or hate it, it’s here to stay. This is the new world, and it isn’t slowing down. Let’s look at some of the biggest misconceptions floating around about it.

Misconception #1: “AI Uses Loads of Water”

This one gets repeated constantly, and it’s not made up. Training and running AI models does use water, mostly to cool the data centres behind them. But the numbers people repeat almost never come with any context.

A typical day of AI chatbot use, say 10 questions, works out to roughly 5 litres of water once you count the electricity behind it. That’s about a bottle of water for every couple of questions.

Now compare that to two things probably sitting in your kitchen right now:

Item	Approximate water footprint
A full day of AI chatbot use (~10 questions)	~5 litres
One avocado	~140–320 litres to grow
One 8oz steak	~3,500 litres
A full year of daily AI use	~1,825 litres, still less than one steak

Growing the avocado on your toast this morning likely used 30 to 60 times more water than a full day of chatbot conversations. And a single steak dinner has a bigger water footprint than an entire year of daily AI use.

Sources: UC Riverside’s 2025 study “Making AI Less Thirsty”; water footprint data via Cranfield University and Wikifarmer.

Misconception #2: “AI Understands and Reasons the Way We Do”

Modern AI can hold a fluent conversation, explain complex topics, and sound completely sure of itself. It’s easy to assume there’s real understanding and objectivity sitting behind the answer. There isn’t, not in the way you’d expect.

The reality: these systems are pattern-matching engines, predicting the most statistically likely next word based on everything they were trained on. There’s no “someone” in there weighing what’s true, and there’s no neutral observer either. The training data carries the imbalances and assumptions of whoever produced it, and whoever fine-tuned the model, which means AI answers can be just as biased as the humans and text they came from. Fluent and confident isn’t the same as correct or objective — a persuasive answer can still be wrong. Treat AI output like a well-read colleague who’s occasionally full of it: useful, but worth checking.

Misconception #3: “AI Is Coming to Replace Everyone’s Job”

This is the one that generates the most anxiety, usually pitched as all-or-nothing: either every job vanishes, or nothing changes.

The reality: we’ve been here before. The Industrial Revolution didn’t end work, it moved it. The dot-com boom didn’t wipe out jobs either, it created entirely new categories of them while quietly retiring others. AI looks a lot more like that: a new way of working, not a mass extinction event.

That doesn’t mean nothing changes, though. Researchers are already mapping out which jobs are most exposed. Microsoft’s 2025 study of real Copilot usage found the highest-exposure roles are knowledge and desk-based work: translators, writers, historians, customer service reps, and sales reps. The World Economic Forum’s Future of Jobs 2025 report points to a similar pattern, with cashiers, ticket clerks, and administrative or data-entry roles among the most at risk. On the flip side, hands-on physical roles like equipment operators, water treatment plant staff, and skilled trades barely register on the AI-exposure scale.

The WEF also expects this shift to be a net job creator, not destroyer: an estimated 92 million roles displaced by 2030, but 78 million new ones created alongside them — a net gain of around 7%. That’s the historical pattern repeating: the jobs change shape more than they disappear outright.

The Upside

Cut through the noise, and the real story of AI at work isn’t replacement, it’s reallocation. The tasks getting automated are overwhelmingly the repetitive, low-judgment ones: first drafts, data lookups, routine replies. That frees people up to spend more of their day on the stuff that actually needs a human: judgment calls, relationships, creativity, strategy.

That’s the shift worth paying attention to, and it’s exactly the kind of shift we help people navigate at Day Eight: less time on the busywork, more time on the work only you can do.