The Great AI Divorce: Microsoft's In-House Models Signal the End of the OpenAI Romance
The honeymoon is officially over. Microsoft just dropped two in-house AI models—MAI-Voice-1 and MAI-1-preview—in what can only be described as the...
Elon Musk told The Economist's editor-in-chief Zanny Minton Beddoes that "money won't matter in 2036," arguing that once AI and robotics produce more goods and services than any person could consume, the underlying need for currency fades. The interview, recorded at Tesla's Texas Gigafactory and published in late July as part of the outlet's Insider series, has circulated widely since.
Key Points
Musk's logic is internally consistent: if AI and robotics genuinely produce more than people can consume, the case for money weakening holds. The open question, which Beddoes raised directly in the interview, is sequencing. Job displacement from automation tends to show up well before the abundance that's supposed to offset it. That gap is where most of the political friction actually lives, and Musk didn't dispute the point so much as acknowledge it and move past it, saying the transition would be "bumpy" and that income transfers were likely the next major political fight.
The more substantive critique isn't political, it's economic. Writers at The Daily Economy and researchers at the American Institute for Economic Research argue that cheap manufactured goods don't eliminate scarcity, they relocate it. A house with a desirable view, a seat at a top university, or access to in-demand expertise stays scarce no matter how abundant robots make physical production. Money, or some equivalent rationing mechanism, still allocates access to those things. That's a meaningful gap in the "money won't matter" framing, since most of what people actually compete for isn't mass-produced goods.
Sam Altman has funded basic income experiments while building the systems that could make large-scale transfers necessary. Mustafa Suleyman has discussed a similar universal provision built on abundant intelligence. Ray Kurzweil's singularity forecasts have assumed comparable abundance for years. The idea has real traction among people building the technology in question, which is worth noting alongside the fact that they also have a direct financial interest in the premise that AI abundance is coming.
None of this changes near-term planning. Whether or not Musk's 2036 timeline holds, the more immediate signal is that the people building frontier AI are actively war-gaming what happens to markets, labor, and government policy once automation output scales. That's a useful lens for any growth strategy work that assumes today's economic structure holds steady over the next decade. It probably won't hold exactly as is, whichever direction the disruption actually breaks.
For teams trying to separate genuine signal from founder speculation in AI predictions like this one, the full interview is worth reading in full rather than the headline alone. That kind of source-level diligence is part of how our AI marketing services team evaluates which AI narratives are worth building a message around.
The honeymoon is officially over. Microsoft just dropped two in-house AI models—MAI-Voice-1 and MAI-1-preview—in what can only be described as the...
Oracle posted quarterly earnings this week that beat expectations but delivered revenue below Wall Street estimates—$16.06 billion compared to the...
1 min read
The Browser Company just crossed a line that most browser makers have avoided for decades: asking users to pay for their web surfing experience. The...