AI Doesn’t Remove Scarcity. It Shifts It to Other Places

or why the claim that everyone will lose their job is wrong

there’s this idea that AI is about to get so good it’ll outperform nearly every white collar worker, faster, cheaper, better. and since companies follow efficiency like a compass, it’ll just replace them. even the new jobs it can’t do yet? give it a few weeks. maybe months. then it’s lights out for the humans.

and when we finally get real world robots, the same thing happens to manual labor. blue collar workers aren’t safe either. that “at least you can’t automate plumbers” take is gonna age like milk. so what do we do with all the displaced people? redistribute the AI wealth? drop some UBI so folks don’t starve in a world they’re no longer useful to?

but this whole theory is cracked.

it imagines the economy as frozen. people want a fixed pile of goods and services, and once AI makes that pile without humans, we’re obsolete. but that’s not how markets work.

markets value what’s scarce and in demand. if AI suddenly floods the world with things that used to take human skill, copywriting, illustrations, design, the price crashes. when anything becomes abundant, its market value evaporates. people might pretend it’s still worth something out of habit, but that’s nostalgia, not economics.

if your response to this shift is to cling harder to the old ways, you’ll feel your job slipping like sand through your fingers. but that’s bad thinking. the market moves. your value isn’t in what you used to make. it’s in what becomes valuable next.

Scarcity doesn’t go away

the one thing that doesn’t happen in markets is that all prices go to zero. that’s ridiculous. that wouldn’t even happen if we had energy too cheap to meter (and we’re heading the other way thanks to anti nuclear sentiment) and near zero marginal costs for producing anything, both material and creative. not convinced? let’s think this edge case through so you see the point and maybe catch a bit of enlightenment.

so, imagine the AGI world. cheap robots everywhere. abundant nuclear fusion energy. all material needs handled. isn’t scarcity eliminated?

nope. some things stay scarce.

first of all: status. status is a zero sum game. it's an implicit ranking, a social hierarchy we’re wired to care about. we want to be higher up than our fellow apes. AI and robots don’t change that. they just become another way we signal status. most likely outcome? more and more of life turns into status games. real life video game. everyone playing. not everyone winning. participation trophies don’t count.

then there’s space. not cyberspace, not infinite cosmic space, but actual desirable human space. where you want to be. cities, homes, vibes. AI doesn’t change the fact that some places are just better to live in than others. that means real estate scarcity isn’t going anywhere.

human bonds are scarce too. AI might be the only true Christian, actually loving everyone equally. we can’t. we won’t. bonding with people is selective and always will be. maybe AI gives us artificial friends, but the minute you say all your friends are AIs, you’ve marked yourself as low status. people will still chase real connections. ironically, full abundance might even boost family bonds. make raising kids cheaper, easier, and more fun. could make parenting high status again.

and attention. attention is forever scarce. social media, news companies, entertainment giants, they all run on the same basic truth: people can only look at so much. even if creation becomes nearly free, we won’t magically gain infinite attention spans. even with neuralinks beaming content directly into our heads, all that means is more content created faster. we’ll still need filters. taste. curation. that’s not going away.

so that’s total abundance. and even there, prices will emerge. the scarcities remain, and the market finds them.

but let’s get back to earth. our real near future world. what should we actually expect?

Expecting Dynamism

the core question we should ask ourselves in our new AI dominated world is: what becomes scarce and demanded once something gets abundant? basically, thinking in second order effects. seeing where the puck is going next, not where it currently is.

how?

the basic idea is that if something becomes abundant, people use it much more. once, sending SMS messages had marginal costs and so you tried to use as few as necessary. now you can send infinite instant messages and no longer care about the price since there isn’t any besides your monthly internet flatrate and the energy your phone consumes. but if people use something much more, that usage creates new demands, and those demands are often where new scarcity shows up.

for example, SMS was fine with a few smilies like :-) or :D. but with WhatsApp and nonstop messaging, people wanted more nuanced ways to show emotion and personality. so we got emojis. now imagine you're stuck in the age of expensive SMS and trying to predict something like emojis if SMS suddenly became free. that’s the mindset. that’s how you figure out where to specialize next.

there are a few issues with this strategy though, so let’s go through them before digging deeper into the method.

first, you don’t really know whether the abundance is actually happening. there are tons of rosy AI predictions out there, but then you see hallucinated nonsense or nerfed models and wonder if any of it will really materialize or just stay a fantasy of almost abundance. if abundance doesn’t arrive, the scarcity doesn’t shift, and there won’t be demand for something based on that false abundance.

second, your predictions of what becomes demanded post-abundance can just be wrong. human behavior is weird. desires shift in complex ways. it’s easy to look in the wrong direction and miss the actual new sources of value.

third, it might be that what becomes demanded also quickly becomes abundant itself, and the truly valuable thing is the next layer out. the third order effect, not the second.

the answer to all of these problems is kind of the same: you need better models. more creative, more ambitious, more original ideas about how the market shifts and what people might want. and then you need reps. you make your guesses, see what lands, and learn. over time, your predictions sharpen. the map gets more detailed. your edge compounds. so let’s think this through.

Demand Forecaster

the interesting thing is that the things created in response to shifted demand aren’t just niche sub-markets of their parent category. they can actually end up being way more valuable than the original scarce thing. take Instagram. it only became possible once taking pictures became free and abundant, but it’s worth far more than Kodak ever was.

that alone is a strong counter to the idea that AI will just erase value creation and leave humans with scraps. in reality, abundance opens the door to higher value activities. when something gets cheap, we stop focusing on how to make it and start focusing on what to build on top of it.

and depending on the kind of thing that becomes abundant, the demand shift tends to follow certain patterns. so here’s the plan: we’re going to look at four big areas where future abundance is likely to hit (AI, robots, autonomous transport & energy) and compare them to four areas where abundance already emerged (light, communication, photography & air travel).

i also built a custom GPT you can use to explore demand shifts in whatever space you’re curious about. link at the end.

first, let’s look at the ripple effects from past technologies that made something abundant.