The AI Shortcut: Reflections on Work, Research and What Might Be Lost

When I started building websites, the equation was beautifully simple. People needed websites. I could make them. They paid me. Sorted. Now there’s a AI shortcut being dangled at almost every turn. Is it a golden era of sustainable web design or are we edging into the abyss?

Let me be clear, I love AI. But over time, I began to clock the bigger picture—that every site I built was adding to the internet’s growing environmental footprint. The internet is now the world’s seventh-largest polluter, which is both impressive and deeply depressing. So I decided to change tack. These days, the digital agency I run is focused on sustainable web design and AI-powered SEO for purpose-driven businesses. We work lean, stay nimble, and over the last year or two, we’ve let AI loose on almost every part of our process.

From writing ad copy and analysing reports to building custom internal tools, AI has given me something approaching superpowers. It’s allowed me—a solo operator with a handful of trusted freelancers—to punch well above our weight. Tools like ChatGPT, Claude, SEMrush, and Swydo help me move faster, dig deeper for clients, and reclaim the time I used to lose to soul-crushing production bottlenecks.

Need a report restructured? Done in minutes. Want eight new ad variations tested by tomorrow? Already running, mate.

AI as the Great Leveller

At its best, AI is properly democratising. Like any game-changing technology, it gives people access to knowledge they might never have had before. Qatar recently offered all its citizens free ChatGPT subscriptions—and fair play to them, that’s brilliant. If knowledge is power, then shared access is equity. Whether you want to build a better website, work out how to get your kid to sleep, or start a business, you’ve now got a world-class assistant at your beck and call.

But—and there’s always a but—there’s a catch.

It Is Full of Shit (Sometimes)

AI is only as good as the questions you ask, and it’s trained to give you what you want to hear. Which can be properly dangerous. People grew sceptical of Google’s top results over time but eventually got a bit casual about it. We learned the algorythems and started to trust what we saw as it became ubiqitous with search. We risk doing the same with AI, except there’s a crucial difference.

Unlike Google, which historically ranked truth through a kind of consensus—weighing backlinks, domain authority, and reputation like a digital democracy—AI doesn’t bother with those trust signals. It just mimics what it’s seen before, which means it can be spectacularly, confidently wrong. And when anyone can churn out thousands of convincing but dodgy articles in seconds, the mirror we’re checking might already be warped beyond recognition.

In my world, this means being constantly on guard. It’s not uncommon for a client to screenshot an ad and ask, “What the bloody hell is this?”—only for me to realise it was an AI-generated variant that’s gone completely rogue. There might be typos, factual howlers, or it might just sound like it was written by a socially awkward robot. My own ChatGTP has, by it’s own admission been particularly “spicy with the hypotheticals” on occasion.

As clever as it is, AI doesn’t actually “know” anything. It mimics, repurposes, reshuffles. And we—the humans—are still the ones responsible for making sure the reflection isn’t completely warped.

Sustainable Tools, Unsustainable Energy?

There’s a philosophical tension I wrestle with daily around sustainability, and it’s doing my head in a bit. AI’s environmental cost is enormous—it guzzles computing power like a Tesla guzzles electricity. But here’s what helps me sleep at night: it saves human energy. It cuts out waste from inefficiency. And in my small business, it lets me do more with less.

So while the tech itself might be a proper energy hog, its impact—when used thoughtfully—can actually support more mindful operations. It’s a bit like using a dishwasher instead of washing up by hand: technically uses more energy, but saves water and sanity. We wrote more on this topic here, if you’re curious about AI’s Environmental Impact.

What It Can’t Replace (Yet)

This came up in a chat I had recently with a qualitative researcher. We were discussing whether AI might eventually muscle in on what they do: interviews, fieldwork, behavioural analysis. And while AI can transcribe interviews in seconds and sort responses into neat little themes, it can’t build proper rapport. It can’t nod empathetically or clock the motivational quotes blu-tacked to someone’s fridge. It can’t read between the lines or interpret what isn’t being said.

Good human research is fundamentally a practice of care. It requires presence, curiosity, emotional intelligence—the stuff that makes us properly human. These are things machines can simulate but not genuinely feel. Not yet, anyway.

That said, AI is a brilliant sidekick for researchers. It can help them move faster, generate stimulus materials, run variations, transcribe lengthy sessions. It’s like having the world’s most efficient intern, minus the coffee runs and existential crises.

The Bigger Risk

The real risk isn’t that AI will nick our jobs—it’s that it’ll make us intellectually bone idle. If we outsource too much thinking, we shortcut the learning. Simon Sinek spoke about this on a podcast with Steven Bartlett, and he’s spot on: experience, nuance, and growth don’t come from just having the answer handed to you. They come from the proper slog of getting there yourself. Just because you’ve read: Think like a Monk, doesn’t make you one.

AI lets us skip that slog entirely. And that’s a risk worth taking seriously, particularly for researchers, analysts, and anyone whose job involves making sense of complexity and chaos.

The Way Forward

If you’re in research—or any field getting a proper shake-up from AI—my advice is dead simple:

Try everything. But get better at your questions. Ask it what you’re missing, or what even as your AI what it would ask an AI to get the answer… If that makes senes. The AI Shortcut can provide valuable gains, but always pair it with a pinch of salt.

The quality of what comes out still depends entirely on what you put in. And if you’re someone who knows how to ask the right questions (which most researchers are, thankfully), you’ve got a massive head start.

The age of AI isn’t just about what’s suddenly possible—it’s about staying sharp, staying human, and using the shortcuts without losing your way.

You got this!