There's a desire we all share as humans – the desire for answers. We're conscious, we live, we search for meaning, and that meaning often comes down to specific answers to specific questions: what is going on, and why do things happen? A lot of our time and effort on this planet goes into seeking those answers. And throughout history, there has been an enormous and recurring appeal to find an Answer Machine – a device that reduces the effort required to get there. This innate desire goes some way toward explaining how completely we can become enraptured with whatever the Answer Machine of the moment happens to be. Once people believe they've found it, it gets embedded into every aspect of life. It becomes the mission, the thing to believe from which springs both hope and desire. Something worth fighting for, worth arguing about, worth losing for. The promise is the great payoff at the end - The Answer. Right now, the Answer Machine is AI. AI is perceived as the ultimate version – the one we've been building toward. The hype distils it down to exactly that: whatever problem you throw at it, AI is the solution. Climate change. Obesity. Maths homework. TPS reports. Government dysfunction. It just needs more AI. And I think this explains, at least in part, why AI feels the way it does. It taps into a millennia-long desire. A mania that feels familiar because we've been here before. A Short History of Answer Machines Search was the Answer Machine of the 1990s. As the internet boomed, we needed a way to find things, and Google became that way. It wasn't overly sophisticated – you still had to click through, read the pages, find and validate the information yourself. AI feels like a better solution partly because it removes that last step. Google already scraped the data. Now it just gives you the answer, with full confidence that it's correct. Which is, of course, where the problems start. Every Answer Machine in history has been flawed. They don't quite understand. They have the same kinds of fallibilities – hallucinations, indeterminacies, false confidence. Yet we keep going back, keep placing our faith, keep pouring money and resources into the machine and what it tells us to do. Go back further and universities played this role. For a long time, academia was the powerhouse of answers – slower, clunkier, accessible to fewer people, but genuinely pushing innovation and making new things happen. Before that, and more controversially, religion was the primary Answer Machine. Religion gave us the rules for life, distilled lessons for what to do in certain situations, who to be, how to behave. The prophets and oracles were its more individualised expressions, but the institutional roots of religion provided the infrastructure. It's worth sitting with how well that maps onto what we're doing now. And if we're willing to look at the fringes – at cults and conspiracy theories – we can start to understand the failure modes. These too are Answer Machines, and the reason they tend not to last is that their flaws are more obviously embedded in their structure. But the dynamics are identical: absolute fervour, total devotion, a willingness to disrupt relationships and distort your view of the world in service of the answers the machine provides. The difference between a cult and a religion is often just time and institutional legitimacy. The difference between a conspiracy theory and accepted wisdom is often just whether the Answer Machine behind it survived long enough to become popular and embedded in culture. We're treating AI with the same fervour. Adoption without evidence. Certainty without validation. The language of hallucination is telling – it implies the problem is in our perception, not in the machine. Which is exactly how the oracles worked: if the answer didn't fit, you'd interpreted the prophecy incorrectly. AI confidently presents answers regardless of whether those answers are correct, and we've built an entire discourse around how to ask the right questions rather than whether the machine actually knows. The Sacrifice Every Answer Machine demands sacrifice. That's perhaps the most important thing to understand about them. Religion has always demanded it. Cults demand it. With Google, the sacrifice was our data – our data became their oil. We submitted to being tracked and surveilled and having everything we did sold back to us as advertising. The world didn't improve because of it. The data didn't benefit humanity. It became profit for a handful of companies. We got free search in return. AI is the same transaction at a larger scale. The capital expenditure going into AI infrastructure right now is staggering – comparable, in raw terms, to what was spent running a world war. And what is it building? Data centres. Server farms the size of hundreds of football fields, consuming power and water at rates that actively worsen the problems we're supposedly using AI to solve. Resources spent not growing food, not housing people and not maintaining the biosphere in any kind of manageable state. We are sacrificing the planet's resources on the altar of the Answer Machine, and the answers we're getting back are mostly being used to build competitive moats for a handful of corporations positioning themselves for whatever the AI wars look like in ten years. This isn't AI for humanity. It isn't AI for the planet or the future. It's AI for profit. And we are the sacrifice. What the Machine Can't Tell You The cruel irony of the Answer Machine is that the answers we most need aren't the ones it can provide. Ask anyone who has earned genuine wisdom over a long life, and they'll tell you it wasn't found in a machine, a doctrine, or a search result. It was found in relationships. In community. In paying attention to the world immediately around them. Truth of that kind isn't something you retrieve – it's something you build, slowly, through connection and presence and time. We live in a society so oriented toward seeking alternative answers that we've lost the habit of sitting with what's actually in front of us. The Answer Machine, whatever form it takes, always promises to shortcut that process. And it always asks more than it gives. The answers to what we seek are closer than the machine would have us believe. They're in our families and communities, in the land, in the connections we form and maintain. Not in what's ephemeral. Not in what some model can confidently generate. In what's real, and present, and ours.