If you attempt to approach someone who looks educated and busy scrolling her phone with a question that goes something like: "do you even know what the word technology means?" — you stand a good chance of either getting whacked by the folded newspaper held in the other hand or simply dunked by the very smartphone being scrolled, especially if it is a pretty girl. It's hard not to engage with the meaning of technology, or a pretty woman, to complete my spin on the analogy.
Wired Contorts of Wireless Technology
But every atom in our planetary airwaves is busy ferrying some kind of data across vast distances. That's technically technology. A wireless device somewhere in New Delhi initiates a one-line text message: "Hi, I am back home." And that piece of data then travels to the home router, goes underground and overground, gets modulated and remodulated several times, and goes through a hoopla of terrestrial wires, goes airborne in between before finding a cable endpoint that dives through choppy waters of the Pacific or Atlantic, swoops out of the water somewhere ashore, and onto an Earth receiving station.
Imagine this: it launches from the cable at the speed of light — miraculously, flawlessly, and perfectly dry. It slices through Earth's atmosphere with precise intent, knowing exactly where it's headed. In an instant, it strikes a targeted satellite, which captures it, remodulates it, and hurls it away like a shard of discarded energy. It ricochets to another satellite, this one unfurling its solar-hungry wings to soak up power and track those scattered atoms. Then, with relentless force, it flings the signal back toward another point on Earth, unstoppable and exact in its journey, before pinging you in another part of the world to read the message.
I went through all of that not to be cute, but to show what a simple attribute of technology looks like. This example is by no means the best way to explain the meaning of technology — I am sure there are simpler, briefer ones that do the same thing in two sentences. But I did that for a reason that should become clear as you stay with me on this atomic ride.
The Hedonic Treadmill
We humans tend to take technology for granted, as if it were inevitable one day to have a microwave or refrigerator. And it doesn't surprise us when we see a more powerful oven, and shortly thereafter a computer, a moon rocket capable of landing itself back on a waiting barge somewhere in the Atlantic to be re-used — a concept never thought possible before. And just like that, you don't bat an eyelid when AI comes along and shoots your productivity to record highs overnight.
Godlike technology — generative AI, instant global communication, genetic editing that may one day create entirely new creatures breaking from all prior evolutionary paths — quickly becomes standard in our minds. Have you ever wondered why that happens? It is due to a combination of rapid psychological adaptation, the brain's hunger for order and efficiency, and a peculiar acceptance that exponential technological growth builds with accumulated experience. This phenomenon is what psychologists call the hedonic treadmill: humans quickly revert to a baseline level of expectation after a major positive change.[1] The miracle normalises. The furniture arrives.
There are four engines driving this, and understanding them is not an academic exercise — it is the prerequisite for seeing clearly in the current moment.
The first engine is psychological normalization. When a revolutionary tool — a smartphone, a generative AI — is released, it causes a spike in excitement and cognitive engagement. That novelty fades quickly, and the tool becomes part of daily unremarkable routine. Research suggests that humans adapt faster to positive experiences than to negative ones, internalising new capabilities so completely that they stop appreciating the wonder and focus only on the utility.[2] Our brains also exhibit what might be called abundance denial: we fail to appreciate the vast depth of improvement baked into the new technology, and shift our expectations upward so rapidly that the revolutionary feels standard before we have finished reading the launch announcement.
The second engine is cognitive ease. Modern technology is specifically designed to eliminate friction — making tasks instantly available, invisible in their workings. The technology becomes opaque precisely when it works perfectly, allowing us to focus on what we want rather than on what is carrying us there. This ease makes advanced technology feel natural rather than alien. When tools take over mundane tasks, they also permit higher-order thinking, making the technology feel like an extension of our cognitive processes rather than a separate entity. We don't pause to think about it, but using smartphones is already a form of brain augmentation — it is just that the hardware is not yet embedded in our skulls. That is coming.
The third engine is technological building-block momentum. Each advancement builds upon the previous one; a new technology is simply the next step in a long trajectory of compounding evolution. Advancements in one field — AI compute, say — accelerate progress in biotechnology, materials science, energy systems. This relentless acceleration makes us treat the impossible of yesterday as the basic of today, because at each step along the chain, the step looked modest from where we stood.
The fourth engine is social proof. Once a technology is adopted by our social networks, it becomes a necessity rather than a novelty. Modern platforms are designed to hold attention and integrate themselves into routine. The advertising economy ensures rapid adoption. We are Paleolithic creatures operating in a high-tech world, and our brains carry a veneer of cognitive flexibility that allows us to manage — though it frequently overwhelms us. In summary: technological advancements do not just change what we do. They rapidly reframe what we expect from life, often before we have consciously registered that our expectations shifted at all.
Such "godlike" capabilities are soon seen not as an existential challenge but as a cognitive partner — making them instantly indispensable. This is the most worrisome thing for those who study AI seriously. It leads us moronically forward without worrying about the consequences.
Thirty Steps
To understand why the hedonic treadmill is more dangerous now than it has ever been, you have to understand the difference between three types of growth. Linear growth adds a fixed amount at each step: one, two, three, four. Exponential growth multiplies by a fixed factor: one, two, four, eight, sixteen. Both of these are comprehensible to human intuition, with some effort.
Recursive growth is different. In recursive growth, each step does not merely add to or multiply the previous result — it improves the very process by which the next step is taken. The system gets better at getting better. The rate of improvement itself improves. This is what AI systems are now doing, and it is what makes the intelligence curve consistently exceed even expert expectations.
The human brain has one uncanny property in the face of all this: it is wired for linear prediction. Consider the analogy Peter Diamandis draws in his thinking about moonshots. If you are tall and take thirty linear steps, each one metre, you end up thirty metres away — a calculation human intuition performs effortlessly. But if you take thirty exponential steps, doubling each time, you end up a billion metres away — enough to circle the Earth twenty-six times.[3] This is not intuitive because you can never experience it in real life. You can only imagine it. And imagining it is not the same as believing it. The treadmill is designed, at the neurological level, to make exponential change masquerade as ordinary until you reach a tipping point and suddenly feel overwhelmed.
ChatGPT, when it burst onto the scene in late 2022, appeared to be an indelible inflection point in human evolution. In barely three years, it already looks much like Web 1.0 — TCP/IP connections to static websites, the opening protocol of a vast new order. What has followed is so far beyond that Web 1.0 analogy that we have no living parallel for it. And yet, here we are: normalised, adapted, scrolling, using it to write shopping lists. The treadmill ran faster than the technology did.
In the backdrop of exponential change, the smooth linear flow of time begins to feel like a distant memory — until we realise that exponential change is a different beast altogether, and that the universe does not slow down for our disorientation. We'd love to make a million dollars in a day. But when you make so much money in seconds, the brain loves the dopamine rush and forgets that the conditions creating that windfall are themselves changing at machine speed. The moral: if and when you do make a million dollars in one minute, do not take it as a blueprint for the future. You will be sorry if you do. The same applies to AI.
The gap between the pace of AI development and the pace of human institutional response — and why that gap is not closing.
Recursive growth isn't abstract — it's happening in public, benchmark by benchmark. A data-driven look at one frontier model and how fast the "impossible" became routine.
When the Tool Stops Being a Tool
Every previous technology in human history has been a tool — something that amplifies the capacity of the person wielding it. The printing press amplified the writer. The telephone amplified the conversationalist. The computer amplified the analyst. Even the internet, the most disruptive technology since Gutenberg, was still fundamentally an amplifier: it expanded the reach of anyone with something to say or sell, but it did not replace the person doing the saying or the selling.
Artificial intelligence is not a tool in this sense. Or rather, it is a tool, but it is the first tool in human history that can also replace the person holding it. And this is precisely the distinction the hedonic treadmill is most perfectly designed to conceal.
We are treating AI the way we treated the microwave — as a clever convenience that will grow invisible with familiarity, as every technology does. We are wrong. The world is not entering a technological revolution. It is entering a discontinuity. Revolutions disrupt industries; discontinuities disrupt assumptions. They redraw the boundaries of what is possible, predictable, and survivable. The period between now and 2028 marks the first time when the rate of improvement in machine intelligence will exceed the rate at which human institutions can comprehend, regulate, and adapt to it.
A 390-fold cost drop in AI reasoning over two years[4] is not a product cycle. It is a phase transition. And the appropriate response to a phase transition is not to go faster on the treadmill. It is to recognise that the treadmill itself has been replaced.
When you watch a signal travel from New Delhi, bounce off satellites, slice through atmosphere, and arrive perfectly dry at a phone on the other side of the planet, you are watching what humans built. When AI designs the next version of that system — optimises its routing architecture, writes the firmware, files the patents, and does it faster and cheaper than the human engineering team that built the previous version — you are watching something categorically different. You are watching the tool become the craftsman. This is not a metaphor. It is already happening.
AI inference costs have collapsed 99.9% in under four years — the arithmetic of the most consequential economic event of our lifetimes, explained in full.
The same argument, applied directly to labor: AI is not replacing the tasks professionals do. It is replacing the professionals — with identities that don't eat, sleep, or make billing errors.
Getting Off the Treadmill
I am not interested in alarmism. Alarmism is a different kind of treadmill — it generates heat rather than light, and keeps you running in place just as effectively as complacency does. I am interested in clarity, because clarity is the only prerequisite for making sound decisions in conditions of rapid change.
The clarity I am after is this: we have built the knee of the curve — the point after which biological evolution comes to a halt and intelligence shoots vertically. What lies at the top of that curve is a different level of civilisation. I believe we are living the singularity. The question is only whether we know it. And the hedonic treadmill is specifically the mechanism by which we avoid knowing it, one comfortable adaptation at a time.
The miracle of that wireless signal I walked you through at the start of this essay did not arrive by accident. It arrived because at every point in that chain — the router, the undersea cables, the satellites, the earth stations — someone refused to take the previous miracle for granted. Someone kept asking questions. Someone stayed uncomfortable with the comfortable explanation. They got off the treadmill long enough to see what was actually in front of them.
That is what I am asking you to do with AI. Not to fear it. Not to celebrate it uncritically. But to refuse the comfortable numbness of the hedonic treadmill for long enough to understand what you are actually looking at. Because what you are looking at is not just another impressive appliance destined to become furniture. It is the most consequential technological transition in human history — and unlike every prior transition, this one does not merely change what we can do. It changes what it means to be the one doing it.
The signal is in the air. It launched at the speed of light. It is perfectly dry. It knows exactly where it is headed.
The only question is whether you are paying attention.
For the first time in history, the quality of a question is the only bottleneck that matters — and the autodidact has been training for this moment their entire life.