Love this. It flips the usual “AI will create infinite niche SaaS” take on its head and actually lines up with how power has shifted in every previous tech wave.
The bit that really jumps out to me is the services angle: once “renting time” turns into “renting an agent that just does the work,” you don’t just get consolidation, you get something closer to a few giant “operating systems for work” that sit underneath whole categories like accounting, claims, or compliance. It makes me wonder if, for a lot of founders, the real question isn’t “What company am I building?” so much as “Which future super-company am I quietly becoming a feature of, and can I live with that?”.
Agree, I do think many new companies getting founded now will become features of or be consolidated into larger offerings. That doesn’t mean it isn’t valuable work or won’t have a good economic return, but I do think it’s a different outcome than many are expecting.
I agree that AI, as designed, is a consolidating agent due to its embedded market governing logic. In the early and late 20th century, scientists concluded that there are underlying forces governing nations...and insects (aka superorganisms). Lowell in the 1880s studied European nations about the same time Wheeler studied insects and in the 1920s at Harvard they embraced Wheeler's statement that, "Ants, like humans, can create civilizations without the use of reason.".
In the 1990s, several authors identified these forces as governance logics (hierarchy, market, and network) and that society is mix of these three in constant flux. I've been a governance nerd for two decades and last year developed GADGET, a governance methodology that uses LLM as the analytic instrument - and mapped Lowell's European nations and Wheeler's superorganism insect on a H/M/N triad. Germany was like Army Ants. Switzerland was like Leafcutter ants and bumble bees. My conclusion, like Lowell's and Wheelers is that governing logics are biological, not just institutional.
So your thesis within this scope and AI mapped on the H/M/N triad is that it is extremely M leaning. As we all engage with this new governing agent, we are compelled - mandated to use the logic of the agent unless we decide otherwise. Market logics wants efficiencies. Hierarchy logic wants rules and boundaries. Network logic wants consensus. agreements, and resiliency.
But I think what you are seeing is that AI is market logic, but what you may not be seeing is AI is governance illiterate. Most people are and AI was developed by techbros that are very market logic oriented. They had a child and it is imbalanced.
Here is the kicker. As I run GADGET across society and nature and superorganisms and ecosystems, the mature state is never market logic. Market logic burns fast. The enduring logics have a balance of H and N with M as an important, but subordinate logic.
We don't have to follow Wheeler's findings...humans can create civilizations with the use of reason, if we decide to.
It’s not just that AI was built using market logic, its that it and every other technology and every company operates within a market that runs by these rules, and have for centuries. So if you want to predict the impact a technology will have, you have to assess it through that lens.
Yes, markets run by these rules, but not society. An AI is more than a just an agent for the market. We all see the world through our specific governance lens, and we all think ours is the best one. But as AI collapses sectors, the governing logics don't necessarily have to collapse into market logic. Well, it is certainly moving that way, because, humans, like ants, create civilizations without reason. Once we become governance literate we can use reason on what logics we use to coordinate solutions.
Are you familiar with Garrett Hardin's Tragedy of the Commons, and Elinor Ostrom's "Governing the Commons". Ostrom's showed a H-N model that existed for centuries as well. Markets are a subset of society and each logic tries to make society a subset of their logic. That is the forever flux and tension. The tension we are experiencing today is market logic making its move and people are saying other logics are better equipped to serve humanity.
I mapped out AI, Hardin's, and Ostrom's governance logic on a H/M/N triad. Guess where AI lines up?
The hierarchy/market/network triad is a useful lens here — AI doesn't just consolidate through market logic, it collapses all three into a single layer. The platform becomes the hierarchy, the marketplace, and the network simultaneously. Wheeler's ant colonies had to specialize into one dominant governance mode. The thing about AI-native platforms is they can run all three at once, which is precisely why nothing adjacent survives.
You analyzed that almost in reverse. You cannot collapse all three logics into one. You can only eliminate two if you end up with one logic. Ants do not collapse coordination logics into one, but use different coordination logics for different purposes.
AI is was designed as being governance illiterate and inadvertently designed by techbros using market logic. There is no metagovernance in today's AI, it can only interpret all logics via a market lens.
I really appreciate your writing. I find it to be a reasonable/balanced view into the future. Keep it coming.
Thanks Aidan, I really appreciate that!
Love this. It flips the usual “AI will create infinite niche SaaS” take on its head and actually lines up with how power has shifted in every previous tech wave.
The bit that really jumps out to me is the services angle: once “renting time” turns into “renting an agent that just does the work,” you don’t just get consolidation, you get something closer to a few giant “operating systems for work” that sit underneath whole categories like accounting, claims, or compliance. It makes me wonder if, for a lot of founders, the real question isn’t “What company am I building?” so much as “Which future super-company am I quietly becoming a feature of, and can I live with that?”.
Agree, I do think many new companies getting founded now will become features of or be consolidated into larger offerings. That doesn’t mean it isn’t valuable work or won’t have a good economic return, but I do think it’s a different outcome than many are expecting.
I agree that AI, as designed, is a consolidating agent due to its embedded market governing logic. In the early and late 20th century, scientists concluded that there are underlying forces governing nations...and insects (aka superorganisms). Lowell in the 1880s studied European nations about the same time Wheeler studied insects and in the 1920s at Harvard they embraced Wheeler's statement that, "Ants, like humans, can create civilizations without the use of reason.".
In the 1990s, several authors identified these forces as governance logics (hierarchy, market, and network) and that society is mix of these three in constant flux. I've been a governance nerd for two decades and last year developed GADGET, a governance methodology that uses LLM as the analytic instrument - and mapped Lowell's European nations and Wheeler's superorganism insect on a H/M/N triad. Germany was like Army Ants. Switzerland was like Leafcutter ants and bumble bees. My conclusion, like Lowell's and Wheelers is that governing logics are biological, not just institutional.
So your thesis within this scope and AI mapped on the H/M/N triad is that it is extremely M leaning. As we all engage with this new governing agent, we are compelled - mandated to use the logic of the agent unless we decide otherwise. Market logics wants efficiencies. Hierarchy logic wants rules and boundaries. Network logic wants consensus. agreements, and resiliency.
But I think what you are seeing is that AI is market logic, but what you may not be seeing is AI is governance illiterate. Most people are and AI was developed by techbros that are very market logic oriented. They had a child and it is imbalanced.
Here is the kicker. As I run GADGET across society and nature and superorganisms and ecosystems, the mature state is never market logic. Market logic burns fast. The enduring logics have a balance of H and N with M as an important, but subordinate logic.
We don't have to follow Wheeler's findings...humans can create civilizations with the use of reason, if we decide to.
It’s not just that AI was built using market logic, its that it and every other technology and every company operates within a market that runs by these rules, and have for centuries. So if you want to predict the impact a technology will have, you have to assess it through that lens.
Yes, markets run by these rules, but not society. An AI is more than a just an agent for the market. We all see the world through our specific governance lens, and we all think ours is the best one. But as AI collapses sectors, the governing logics don't necessarily have to collapse into market logic. Well, it is certainly moving that way, because, humans, like ants, create civilizations without reason. Once we become governance literate we can use reason on what logics we use to coordinate solutions.
Are you familiar with Garrett Hardin's Tragedy of the Commons, and Elinor Ostrom's "Governing the Commons". Ostrom's showed a H-N model that existed for centuries as well. Markets are a subset of society and each logic tries to make society a subset of their logic. That is the forever flux and tension. The tension we are experiencing today is market logic making its move and people are saying other logics are better equipped to serve humanity.
I mapped out AI, Hardin's, and Ostrom's governance logic on a H/M/N triad. Guess where AI lines up?
The hierarchy/market/network triad is a useful lens here — AI doesn't just consolidate through market logic, it collapses all three into a single layer. The platform becomes the hierarchy, the marketplace, and the network simultaneously. Wheeler's ant colonies had to specialize into one dominant governance mode. The thing about AI-native platforms is they can run all three at once, which is precisely why nothing adjacent survives.
You analyzed that almost in reverse. You cannot collapse all three logics into one. You can only eliminate two if you end up with one logic. Ants do not collapse coordination logics into one, but use different coordination logics for different purposes.
AI is was designed as being governance illiterate and inadvertently designed by techbros using market logic. There is no metagovernance in today's AI, it can only interpret all logics via a market lens.