What follows is a scenario, not a prediction. This isn’t bear porn or AI doomer fan-fiction. The sole intent of this piece is modeling a scenario that’s been relatively underexplored… Hopefully, reading this leaves you more prepared for potential left tail risks as AI makes the economy increasingly weird.
It’s an extremely long article as it considers so many of the personal, business, financial, and government interrelationships in an AI-saturated economy but, as a thought experiment, it’s worth considering every word.
For the entirety of modern economic history, human intelligence has been the scarce input. Capital was abundant (or at least, replicable). Natural resources were finite but substitutable. Technology improved slowly enough that humans could adapt. Intelligence, the ability to analyze, decide, create, persuade, and coordinate, was the thing that could not be replicated at scale.
Human intelligence derived its inherent premium from its scarcity. Every institution in our economy, from the labor market to the mortgage market to the tax code, was designed for a world in which that assumption held.
We are now experiencing the unwind of that premium. Machine intelligence is now a competent and rapidly improving substitute for human intelligence across a growing range of tasks. The financial system, optimized over decades for a world of scarce human minds, is repricing. That repricing is painful, disorderly, and far from complete.
But repricing is not the same as collapse.
The economy can find a new equilibrium. Getting there is one of the few tasks left that only humans can do. We need to do it correctly.
This is the first time in history the most productive asset in the economy has produced fewer, not more, jobs. Nobody’s framework fits, because none were designed for a world where the scarce input became abundant. So we have to make new frameworks. Whether we build them in time is the only question that matters.
I fear that human intelligence will be devalued in favor of machines. I have seen what happens when elderly people stop using their brains, and it scares me that this could become the norm for the majority at all ages. I suppose, though, that this is not necessarily a new phenomenon across the course of human history. Archaeology has revealed humans who were ahead of their time, and those civilizations disappeared.
My latest exercise with AI tools made me think my chemistry-related job is secure at least for now. Why? It is a relatively small niche, so even though structures, even the most complex ones, that can look like chickenwire to a non-chemist can be linearized and coded, metabolically unstable sites can be predicted, etc. no one is yet rushing to make the tools that would completely replace us! Yes, there are tools to assist us, but they are far from a replacement. Somehow, we survived the rise of combinatorial chemistry, China outsourcing, and the SCOTUS attacks on patents. We are ancient species still roaming the planet.
Sadly, the chances of AI wiping out humanity seem to be higher than the chances of AI putting chemists out of business.
Yeah, I don’t think structural engineers and detailers are going to be replaced for awhile. I detail precast concrete on buildings. IF IF IF an architect could produce a decent set of construction documents, it MIGHT be possible for AI to do a lot of the detailing. But I hardly ever see drawings that make much sense. My job is to turn the chaos into order by taking my best shot and asking lots of questions.
My S also has a job that can’t be done by AI. The question is, can/will AI change the landscape in a way that negates the need for his job due to radical changes to everything else. Time will tell.
I know the next two to five years are going to be disorienting in ways most people aren’t prepared for. This is already happening in my world. It’s coming to yours.
@shawbridge posted that article upthread as well as a rebuttal. How this all actually plays out, we’ll have to see, but the Citrini Research article I posted above provides a thoughtful piece of modeling.
I think some of the high prestige jobs that are so coveted by the high achievers on CC are going to be impacted. Who is going to need a “quant” when AI can do the work so much more quickly. The same goes for consulting - AI can take in all the data and spit out recommendations at a quicker pace and more cheaply than MBB. Of course, people will still be needed in these jobs/at these companies - just not nearly as many.
I.e. more professions will become more elite-or-bust than before, with fewer (but perhaps better paid) haves and more have-nots. Or, in economic terms, the Gini coefficient will increase.
I’ve read a few interesting bits about AI and many seem to indicate that many of the financial (and other) benefits will accrue to people who are already the “haves”. I don’t know how true that is but it is a frightening prospect (to me) because I think vastly increasing the proportion of people that are scrambling for some semblance of a decent life creates a really unstable society.
This instability is a big concern of our military as I posted upthread:
…this new entrant barges into the widening divide between the haves and have nots with the few positioned to benefit against a mass of those displaced potentially faster than our economy can absorb them meaningfully. This disruption may not be as smooth as in the past with AI and automation quickly disenfranchising an entire population and spawning an unsustainably growing underclass. This is a scenario our military is watching and studying carefully as we know what happens when people who are left out have nothing to lose.
Defining have and have not in terms of “elite or bust”, I think both groups can benefit from AI in different ways. For the less naturally gifted employees, AI can serve as a capacity equalizer. For example, there was an earlier thread about students using AI to write their papers. It’s not the most naturally gifted students who benefit from this. They have no problems getting A’s on their own without support of AI. It’s instead the less naturally gifted students whose performance is most enhanced most by AI.
Extending to the job market, a less naturally gifted teacher might benefit more from AI than a more naturally gifted teacher. The less naturally gifted teacher might using AI for lesson planning, creating and grading effective quizzes, etc. When planning the lesson, the teacher might use prompts like, “Explain photosynthesis for 8th graders using an analogy.” The AI could support the areas in which their teaching is lacking, making them more productive and effective. However, the more gifted teacher doesn’t need this assistance. Their teaching is fine as is, and may decline if trying to use AI in this manner.
Or a less gifted nurse practitioner might benefit more from AI recalling the long patient history and suggesting procedures + diagnosis than a more gifted nurse practitioner.
Or a less gifted salesperson might benefit more from AI suggesting talking points, creating customer proposals, recalling customer history… more than a more gifted salesperson.
I think the negative effects of AI are going to go far beyond impacting “elite or bust” type careers but wanted to note that those careers are not immune despite attracting the best and the brightest. And while I agree that AI has benefits, I’m not sure equalizing student outcomes, so that talented students are indiscernible from those who are using AI to boost the quality of their papers, is one of those. At what point do students - including younger students - become so reliant on AI that they never develop writing or analytical skills of their own?
It is concerning and for good reason. In the past technological upheavals have primarily impacted the working class - but this time AI is coming for the knowledge class (the economic engine of our country). The potential fallout is scary because AI has the potential to disrupt such a wide swath of careers. This whole site is dedicated to college - how many college grads will we actually need if there aren’t going to be white collar jobs for them to go into after they graduate? Of course there are good reasons to attend college beyond getting a job afterwards but given the immense cost, I can’t see how people will continue to justify the expense if the payoff isn’t there.
Limiting education is a tool used by the powerful to remain in power. What a fabulous bonus AI has provided for the already rich and powerful. They will be able to stay on top until the inevitable collapse of our civilization. History provides examples of what our future holds, even if the technology itself is new.
When we were writing an “amends” letter to our estranged son, I input notes and a sample letter to ChatGPT and asked it to do the job. Wow, it came out great. The therapist said, “This is really good!” And our son liked it, too. I know AI did a much better job than I would have.
I know it’s long, but the Citrini Research article I posted upthread is really worth reading to get an idea of just how far-reaching the effects of AI might be. We worry about education and jobs here, but this article goes beyond this myopia to help us think about what our economy/world might look like when AI is firmly embedded in every corner. As it will be.
All past major historical technological/industrial advances I am aware of have led to improvements in economic productivity and increased total employment in long run, when measured at national/world level. However, the benefit is not equal for all groups. For example, in previous decades modern computers have led to various new fields of employment that involve the new computer technology. That may look good in terms of national stats, but telling an assembly line worker who lost their job to automation/robotics/tech/… that they should become skilled with computers and get a new career that involves using the new tech is not practical advice . We still see fallout from this type of job displacement with technological changes decades later. For example, I doubt Trump would have won either election had not there been a significant number of displaced workers in manufacturing sector who want to revert back to previous times and “make America great again”.
In past history, the groups that benefit most from major technological/industrial advancements are not necessarily the rich and powerful, but there is a correlation. It’s more often skilled workers who can effectively use the new technology to amplify their productivity, working along side the new tech rather than against it. Entrepreneurs and capital owners also tend to reap large benefits, which can lead to worsening inequality between “have and have nots”. General consumers also tend to benefit in the form of reduced prices for corresponding quality, with the increased productivity.
It’s unclear how AI is different from all previous past historical examples, but I would not assume “the inevitable collapse of our civilization.” My guess is more likely to lead to increased productivity, more resembling past historical examples. The market seems to largely agree, as do a large portion of economists. Current prices reflect future expectations. Expectations of inevitable collapse of our civilization would manifest differently.
Shock! Shock! I learned yesterday that an open problem I’d been working on for several weeks had just been solved by Claude Opus 4.6 — Anthropic’s hybrid reasoning model that had been released three weeks earlier! It seems that I’ll have to revise my opinions about “generative AI” one of these days.
Why? It is a relatively small niche, so even though structures, even the most complex ones, that can look like chickenwire to a non-chemist can be linearized and coded, metabolically unstable sites can be predicted, etc. no one is yet rushing to make the tools that would completely replace us! Yes, there are tools to assist us, but they are far from a replacement. Somehow, we survived the rise of combinatorial chemistry, China outsourcing, and the SCOTUS attacks on patents. We are ancient species still roaming the planet. 