Everything so far has assumed the plot twist resolves the way both Marx and Musk expect: productive forces mature, costs collapse, money wilts, curtain. This part is the loyal opposition. There is a second ending available, and on current trends it is the better-supported one.
Who owns the machines?
Thomas Piketty's Capital in the Twenty-First Century (2014) rests on a brutally simple inequality: when the return on capital exceeds the growth of the economy — r > g — wealth concentrates. Automation raises capital's share of income almost by definition: the robots' output flows to the robots' owners. AI-driven productivity could therefore push concentration to unprecedented heights rather than deliver shared plenty. Altman, remember, conceded the premise in his own essay — AI shifts power "from labor to capital" — before proposing a tax to patch it. The patch requires taxing precisely the people best equipped to resist taxation, which is where forecasts of political ease go to die. Source
Abundance with a subscription fee
The second objection cuts deeper than distribution. Capitalism has a long record of manufacturing scarcity where none needs to exist — intellectual property, licensing, rentiership — because profit requires a meter. Cheaper production has never automatically meant free goods; it has often meant wider margins. Critics of Bastani make the same point from the left: Fully Automated Luxury Communism borrows its futurism from the billionaires and supplies no theory of class power to explain why the billionaires would hand over the luxury. The Solve Everything authors, to their credit, wrote the objection into their own investment advice: own the rails. A world of Universal Basic Capability running on privately owned rails is not the Gotha Programme. It is company scrip with excellent UX.
Money is stubborn
Even granting the robots, "money becomes irrelevant" overstates what abundance can do. Money prices more than food and transport: it prices time, risk, land, status, and position in every queue that matters. Necessities can be abundant while beachfront property, energy, and attention stay scarce — and people compete for scarce things with whatever token is available. Keynes's own forecast is the standing warning: the productivity arrived roughly on schedule, and the fifteen-hour week never came, because the gains were captured rather than distributed. A prediction that was half right in exactly this way is now being remade, at higher stakes, by people with larger positions in the outcome.
The two endings
So the argument closes on a fork. In one ending, Marx's sequence completes under new management: capitalism develops the productive forces to quasi-infinity, prices deflate toward zero, and something structurally like the higher phase of communism arrives wearing a SpaceX jacket — with the reformist left, in a final irony, having delayed the arrival by cushioning the system that had to exhaust itself. In the other ending, the machines work, the springs flow, and the flow has a valve, owned by the people who built it. Abundance becomes the most stable configuration capitalism has ever found: a universal income keeps the displaced quiet, capability is distributed while ownership concentrates, and UBI turns out to be Bismarck's insurance scheme at planetary scale — the stabilizer, not the gateway.
There is even a test between them, and it is measurable. If labour's share of income rises and the cost of living falls broadly as automation spreads, the first ending gains ground. If capital's share keeps climbing and the gains keep concentrating — the current trend — the second ending is winning. Marx, at least, would recognize the method: watch the material conditions, ignore the slogans.
Where everybody stands
The cast, placed by two questions: does abundance end capitalism or entrench it, and should the transition be accelerated or restrained? The crowded corner is the joke.