What is Abundance? — Josh Loredo

Almost every conversation I have about artificial intelligence arrives at the same question: where is this going?

It’s a fair thing to ask. For most people AI showed up looking like a hydra, growing a new head every few months, reaching into parts of life that felt safely human a year ago. Fear is a rational response to that. So is uncertainty.

But the discourse is loud, and a lot of what’s loud is wrong. I want to clear some ground first, then get to the part I actually find interesting.

Clearing Ground

Is AI actually thinking?

No. What we have runs large-scale matrix operations over learned patterns and produces output that resembles reasoning. Resemblance is not the thing itself. There’s no evidence of interiority here, no thoughts, no feelings, no experience of being the system.

Will AI take everyone’s jobs?

Not overnight, and not in the way the question implies. It will automate tasks, dissolve some roles, and reshape most of the rest. Cars did this. Computers did this. The internet did this. Each one deleted categories of work and spawned industries that had no name beforehand.

But I’d hold this answer loosely. It has a shelf life, and I’ll come back to why.

Is AI a bubble?

Market value does not measure technical value. Both things can be true at once. Investors can badly overprice the near-term economics while the underlying technology remains real, useful, and load-bearing. The dot-com crash did not make the internet fake.

Is AI stealing human work?

Unclear, and anyone confident in either direction is skipping steps. Training data is not stored as a lossless copy, but its patterns are encoded across the model in a deep and distributed form that we don’t have good language for yet. Every framework we have for ownership assumes copies exist. This isn’t a copy. It also isn’t nothing. That gap is where the real questions live: learning, consent, attribution, payment.

Can we just control it, or turn it off?

Simple rules cannot anticipate every behavior of a complex system. That’s the technical problem, and it’s hard. The political problem is worse. Whoever holds the off switch on a sufficiently capable system holds leverage with no historical equal. “Just control it” quietly assumes we have already settled who “we” refers to.

Is AGI impossible, or about to arrive?

It is not impossible unless physics makes it impossible, and nothing we currently know about physics does. Timing is the open question, and I’ll give you a real number: 12 to 16 months.

Here’s how I break that down:

Fully autonomous self-improvement loops, meaning zero human in the loop, are weeks to months out. Call it 2 to 3 months conservatively. Once that lands, the improvement rate we’ve watched over the last few years begins to look like slow motion. Then add 3 to 6 months for hardware to scale into the demand that creates, since the bottleneck moves to silicon and power almost immediately. Then another 3 to 6 months for expansion into genuinely emergent behavior, the capabilities nobody trained for and nobody predicted.

The (really) Old Economy

Now the part I actually want to talk about, because this is where the economics gets more interesting to me.

Definition first, since the word gets thrown around carelessly:

Abundance: having so much more than enough of something that the concept of scarcity rounds to zero.

Everything we call “the economy” is built on artificial scarcity. Currency, supply, demand, price discovery, all of it is machinery for allocating things there aren’t enough of. It works. It has worked for millennia. It is also a mirage, in the sense that it’s a local convention stacked on top of something way older and infinitely less negotiable.

The real economy has been running since roughly 13.8 billion years before we minted the first coin, and it does not accept dollars. It’s the economy of energy.

The rules are simple, and they don’t care about your politics or your passport. Energy cannot be created or destroyed, which makes the supply fixed. Entropy converts usable energy into heat, permanently, in one direction only. Every structure in the universe, you included, is a temporary local slowdown of that conversion.

That’s the job. Slow the conversion. Stay coherent slightly longer than the gradient wants you to.

It’s the only economy that really matters. (pun intended)

Why Intelligence Is the Bridge

The claim that AGI moves us onto that economy deserves more than an assertion, so heres how I see it playing out:

Every barrier between us and the energy economy is an intelligence problem wearing a different costume.

Fusion is a plasma control and materials problem.
Orbital solar is a launch cost and thermal management problem.
Grid-scale storage is a chemistry problem.
Efficient computation is an architecture problem.

None of these are blocked by insufficient energy. They’re blocked by insufficient understanding of how to manipulate the energy that’s already there, in quantities that make our current consumption look like a rounding error.

I want to be precise about the claim though, because the strong version is wrong. Intelligence is not the only constraint. Capital is real. Materials are real. Political will is real, and it’s usually the slowest of the three.

What I’m claiming is narrower. Of everything on that list, intelligence is the one sitting closest to a discontinuity.

The other constraints move linearly. Raise more capital, mine more lithium, win more elections, and each one buys a roughly proportional amount of progress. Intelligence is the only constraint where a breakthrough compounds, where the thing you built shortens the time to build the next thing. Once that delta gets large enough, the others stop behaving like independent variables. Capital chases demonstrated capability. Materials science is itself an intelligence problem. Political will bends around a fait accompli much faster than it bends around an argument.

Remove that constraint and the cost of producing almost anything converges on the cost of the energy required to produce it. Not labor. Not scarcity rent. Not distribution. Energy.

At that point the mirage stops being useful. You can keep pricing things in dollars, but every dollar is doing nothing except standing in for joules, badly.

Work as Something We Used To Do

Which brings me back to the jobs answer I told you to hold loosely.

Every previous technology moved labor between scarcities. The plow moved people out of subsistence farming and into cities. The assembly line moved them out of craft and into process. The computer moved them out of process and into knowledge work. In each case the scarcity relocated and human effort followed it.

This one is different because it attacks scarcity itself.

I don’t think anything happens overnight. There’s a plateau in front of us, probably a long and uncomfortable one, where the technology is capable enough to displace but the economics haven’t reorganized to absorb it. That period will be genuinely hard for a lot of people and I don’t want to hand-wave it.

But on the other side of that plateau, work becomes something we used to do. Not something we were freed from, and not something taken from us. Something that stopped being necessary, the way hauling water stopped being necessary.

Here’s the part that convinced me this isn’t fantasy: the entropy required to keep every human alive and comfortable is trivially small on a Kardashev II scale. All of it. Food, shelter, climate control, medicine, transportation, entertainment, the entire material footprint of eight billion people. Against the full output of a star, it rounds to zero.

The Mirage Goes Deeper Than You Think

The standard objection here is positional scarcity. Land, originals, proximity, status. The argument goes that even in material abundance these stay scarce because they’re relational rather than physical, so someone always has to lose.

I think this objection proves the opposite of what people think it proves.

Status and ego are not physical constraints. They exist entirely inside human minds and they’re downstream of the scarcity economy that produced them. A hierarchy that exists because people agree it exists can stop existing when the conditions that made it useful go away.

The physical version doesn’t hold up either. Land is not fixed. It’s fixed on this planet, at this technology level, which is a very different claim. Moons, orbitals, and eventually terraforming all expand the supply. If you want beachfront property, build a beach.

The Actual Question

So where are we headed?

Toward a version of the question our current vocabulary handles badly. We spend most of our time arguing about how to divide a pie we invented, while sitting inside a universe that has been running a completely different accounting system the entire time, one that doesn’t care what we’ve agreed things are worth.

Abundance isn’t the end of economics. It’s the wake up call for our civilization to stop trading in monopoly money and start trading in the real currency of the universe: joules.

Scarcity was never the problem. Our units were.