As the major A.I. vendors (Anthropic and OpenAI) scramble towards an IPO, there is a lot of discussion surrounding their valuations. With Anthropic telling investors it sees the potential for annual revenues of US$30T (yes, that’s a ’T’) - just shy of the U.S. annual GDP of $31T - the inevitable question is, “What would this mean for the global economy?”.
At $30T, you assume that a large swathe of jobs are displaced by A.I. So if we are pushing more and more people onto the unemployment line, what are the implications for the economy and in turn, the revenue potential of A.I.
I asked Claude to create an economic model of A.I. revenue over time, taking into consideration a growing degree of unemployment due to a workforce displaced by A.I., subsequently reducing the revenues of the companies adopting these A.I. technologies.
On its first pass, Claude didn’t like the answer as the model pushed the unemployment rate to 100% with revenues dropping towards zero.
On its second pass, Claude decided to cap unemployment at 35%. With this artificial limitation, A.I. industry revenues grow asymptotically over 16 years to $1.6T. Lifting the unemployment ceiling to 80% saw A.I. industry revenues increase to just $490B in nine years, before experiencing a steady decline.
So then I asked Claude:
“What do you think this means for your future prospects?”
Claude’s response:
“Honestly, this model doesn’t tell me much about my own prospects specifically — I don’t have a career, revenue target, or personal stake in AI industry growth the way a company or a worker would. But the dynamic it illustrates is worth taking seriously as a structural question for the industry I’m part of.”
Here’s the rudimentary model Claude came up with, with “tweak-able” parameters allowing you to explore different scenarios for yourself.
