The Utter Rubbish You’re Asking AI

Why You’re Still a Dullard with a Supercomputer

You’ve got your generative AI humming away, a miraculous engine of pure knowledge, and what are you doing with it? You’re using it like a marginally faster, slightly smugger version of Google. You’re asking it things you already know you don’t understand, which is, frankly, the easy, lazy bit.

And it’s bloody infuriating.

Everyone, and I mean everyone, is asking the wrong questions. You arrive with your flimsy, pre-packaged queries: “How do I improve conversion rates?” “Should we hire a new bloke for sales?” “What’s the best strategy for something-or-other?” The machine dutifully spits out its polished, statistically-sound answer, and you toddle off feeling vaguely productive.

But underlying this, nothing fundamental has changed. You’ve simply optimised within the utterly flawed, restrictive frame you dragged into the conversation. You’ve polished the brass on the sinking ship and declared it seaworthy.

The real point of AI, the one that makes me want to scream and then take a long nap under a bus, isn’t answering the trivial stuff. The genuine, game-changing capability of these things is their ability to act as a forensic tool, a digital chisel used to expose the rotten foundations of your own thinking. Its true talent lies not in giving answers, but in interrogating the question itself.

The companies truly struggling aren’t the ones asking, “How do we adopt AI faster?” They should be asking, with a cold shiver of terror, “What work are we doing that we only perform because we’ve always done it, like some pointless, depressing office ritual?” The frantic, stressed executive isn’t asking, “How do I get more output from my staff?” They should be asking, “What metric am I religiously measuring that actually means nothing, like tracking the frequency of damp patches in the office canteen?”

The shift needed is simple, brutal, and desperately uncomfortable. Stop asking the AI to solve your problem; start asking it to audit your sodding thinking.

Feed it your sacred company strategy and demand it highlights every unspoken, ridiculous assumption you’ve made. Describe your biggest corporate challenge and ask what crucial, adjacent problem you are deliberately, perhaps subconsciously, ignoring. Shove your ‘decision framework’ in its face and ask what utterly vital consideration you haven’t even thought to factor in.

This isn’t about clever “prompt engineering”. This is a systematic, cynical way of weaponising a machine against your own ingrained cognitive blind spots. You use it to force yourself outside the suffocating conceptual box you brought with you. You use it to discover that the question you thought was important was merely a flimsy, three-questions-deep distraction from the real, terrifying monster lurking beneath.

Most of you are using AI to get smarter about the answers. You’re wrong. The only worthwhile move is using it to get savagely, uncomfortably smarter about the questions. Everything else is just expensive noise.