To make sense of the massive amount of money being poured into artificial intelligence right now, the market for end-user AI services will eventually need to hit $920 billion. Annually. For that kind of investment to pay off, AI can't just be a slightly better version of the software we already use. It has to do something much more disruptive. It either has to replace human labor and much of the current (mostly ERP) software stack to cut costs, or it has to create massive demand for entirely new kinds of goods and services.
Tech companies love to talk about infinite growth, but a realistic look at what humans actually need and demand suggests that the potential for true economic expansion is limited. AI will certainly help us discover new materials and better drugs. It will optimize our supply chains and energy grids, allowing us to maintain our current standard of living while putting much less strain on the environment. And it will make personalization cheap—tailoring everything from a product to a meal to our specific tastes. But the actual value of this optimization has a natural ceiling. You can only make a delivery route so efficient before it’s a straight line. Optimization fixes waste, but it doesn't create a brand-new pie; it just cleans up the one we have.
This leaves us with a real question: Are we moving toward a saturated world where humans simply have more and do less? Are we entering an era of human deflation, where each new generation is less skilled and less intellectually savvy than their parent generations?
If technology makes physical goods cheap and takes care of all our daily cognitive friction, our historic struggle against scarcity basically ends. We get a life of total convenience, but we also get a life where we rarely have to exert ourselves. Physically and mentally.
Paradoxically, abundance doesn't mean we stop fighting. It just changes what we fight over. When all our travel, entertainment, and homes are perfectly customized and we have all the baseline comfort we could want, we run right into the problem of relative distribution. Humans don't measure their status by absolute wealth. We measure it by relative access to things that will always be scarce. AI can't build more coastline real estate. It can't mass-produce social status, influence, or attractive mates. Because these things are strictly limited, competing for them remains a zero-sum game.
And this struggle won't just happen between individuals; it will happen between entire societies, countries, and regions. We have seen the conflict between energy-rich and energy-poor countries. We may add new tension between nations that own the computing infrastructure and those that are just paying to use it.
This is the real trap of the AI era. As automated intelligence rises to do our thinking for us, humanity faces two quiet crises: the temptation to do, and thereby become, less and the urge to compete over the remaining scarcity to define our status. If we hand over all our mental effort to machines, we risk sliding into a state of happy passivity, where we only think when we are absolutely forced to. At the same time, if we use our newfound free time just to battle each other for relative status, we will turn a world of structural abundance into a hyper-efficient conflict over the few things that can't be replicated.
If we want to avoid this decline, we have to raise our own standards. Collectively. We need to resist the pull of effortless comfort and find new, meaningful ways to challenge our minds and grow. Working with our bodies and minds must be a collectively esteemed value in and of itself. And we have to figure out how to distribute what is scarce without tearing our societies apart. The machines are learning to optimize the world; our job is to learn how to live in harmony.