The other day I came across a tweet referencing comments by Elon Musk that made a fairly radical claim: AI and robotics aren’t just increasing productivity — they’re about to break the entire economic system.

The argument was roughly the following: Modern economics assumes scarcity. Resources are limited, labor is limited, and production grows gradually. Prices emerge from this balance of supply and demand under constrained resources.

But according to the tweet, AI fundamentally changes this picture. If AI and automation can scale production exponentially while eliminating labor costs and reducing errors, the marginal cost of producing many goods approaches zero. If production becomes effectively unlimited, scarcity disappears. And if scarcity disappears, the whole pricing system collapses.

In that scenario, prices don’t slowly decline — they implode. GDP becomes less meaningful because output can increase dramatically while prices drop. Governments try to react using traditional tools like stimulus or monetary expansion, but those tools were designed for a world where scarcity still exists.

The tweet framed this as something like a “supersonic tsunami”: production grows faster than policy can react, deflation becomes permanent, and the economic models we rely on no longer describe reality.At first glance this sounds dramatic, but the intuition behind it is not entirely wrong.

We already see examples of this dynamic in the digital world. Software, digital media, and now AI-generated content have extremely low marginal costs. Once something is created, producing an additional copy is essentially free. In those markets, prices often fall toward zero over time.

AI could extend this effect further. If more and more intellectual and creative work becomes automated — writing, coding, designing, analyzing — productivity could increase much faster than in previous technological waves. So the idea that technology erodes certain forms of scarcity is clearly real.

However, after thinking about the argument a bit more (and discussing it with ChatGPT), I think the conclusion that “the system breaks” is probably the wrong interpretation: The reason is that economics doesn’t require everything to be scarce. It only requires that something is scarce. And scarcity tends to shift rather than disappear.

Even in a world where AI dramatically reduces production costs, there are still fundamental constraints: energy, raw materials, compute infrastructure, physical manufacturing capacity, land, and even human attention. If one bottleneck disappears, another one usually becomes more important.

You can already see hints of this today. As software became cheap, computing power and cloud infrastructure became more valuable. As information became abundant, attention became scarce.

From that perspective, AI doesn’t remove the basic structure of economics — it just changes where the constraints are located. There is also another interesting point in the tweet: the claim that traditional metrics like GDP might become less meaningful.

Here there might actually be something to it. GDP measures the market value of goods and services, not necessarily the real welfare they generate. If AI allows us to produce much more output at much lower prices, economic value measured in dollars might stagnate or even decline while actual abundance increases.

We’ve already seen similar effects with free digital services that generate enormous value without appearing strongly in GDP statistics. Another implication that the tweet hints at — although not explicitly — is a potential shift in economic power.

If production becomes heavily automated, value might concentrate around whoever controls the systems enabling that production: AI models, data, energy, robotics, and computing infrastructure. In that world, ownership of these systems becomes more important than labor itself.

This could lead to significant political and social debates about how the benefits of automation should be distributed. Still, even under these assumptions, the likely outcome isn’t an overnight collapse of the economic system. Historically, technological revolutions rarely destroy economic structures completely. Instead, they gradually reshape them.

New technologies remove some constraints, create new ones, shift value creation into new sectors, and eventually force institutions and policies to adapt.

So in the end, apart from the tweet wanting to get attention, it is less dramatic than its original message. AI may indeed push productivity to new levels and reduce the cost of producing many goods and services. Some industries may experience strong deflation. And our traditional economic indicators may become less informative. But that doesn’t mean scarcity disappears or economics stops working. A more realistic interpretation might simply be this:

AI doesn’t eliminate scarcity — it relocates it.

Understanding where the new bottlenecks emerge may turn out to be one of the most important economic questions of the coming decades.