Elon Musk has said there is no point in saving money for the future, because artificial intelligence and robots will raise production so far that money loses its conventional importance.
The remark came as a reply on X. A user wrote that he was justifying some questionable financial decisions on the expectation of future abundance — that one should not be reckless, but that in a few years most of those decisions would not matter. Musk replied that in an age of “superintelligent abundance”, saving money for the future makes no sense.
The reasoning
It is not a throwaway line; he has made the argument before, including in an interview with Peter Diamandis of the XPRIZE Foundation earlier this year. The chain runs like this: robots can work cheaply and without stopping, so output rises enormously and production costs fall. Supply far exceeds demand, prices fall towards zero, human labour is needed less, wages are needed less — and an economy built to allocate scarce goods has less and less to do.
Where the argument is strong
On the production side it is not absurd. The cost of a manufactured good has fallen dramatically over two centuries of mechanisation, and some digital goods already cost effectively nothing to reproduce. Anyone who has watched the price of a solar panel or a flat-screen television over twenty years has seen a version of the mechanism.
And where it has a hole you can walk through
The argument is entirely about production and says nothing about distribution, which is where the whole question lives.
Cheap goods do not reach you because they are cheap. They reach you because you have money, and in Musk’s scenario the mechanism that puts money in most people’s hands — wages for labour — is the thing being removed. Abundance that arrives alongside the disappearance of income is not abundance for the person whose income disappeared; it is abundance for whoever owns the robots.
What would bridge that gap is a transfer: a universal income, public ownership of productive capital, taxation of automated output. Each is a political fight, not a technical consequence, and none of them follows automatically from better robots. Musk, who owns a great deal of the relevant capital, does not say which he favours.
And the advice itself
Separate the forecast from the instruction, because the instruction is the dangerous part. Even if the forecast is right about 2050, savings are not for 2050. They are for the illness next year, the lost contract next month, the daughter’s admission fee.
For a household in Bangladesh, where there is no state pension to fall back on and a medical emergency is paid out of pocket, “don’t save, abundance is coming” is advice whose downside is borne entirely by the person who takes it — and whose upside, if it arrives at all, arrives for everyone whether they saved or not.
One historical note, because the idea is older than the technology. Keynes predicted in 1930 that his grandchildren would work fifteen-hour weeks, on much the same reasoning: output per worker would rise so far that the problem would become what to do with the leisure. Output per worker did rise roughly as he expected. The fifteen-hour week did not arrive, because the gains were taken as higher consumption and higher returns to capital rather than as free time. The forecast about production was right and the forecast about life was wrong, for reasons that had nothing to do with machines.
