My last post ended with a question about meaning. This one tries to answer it, with the help of two experiments that had nothing to do with AI.
Two Experiments, Two Deaths
Universe 25 (1968–1973). Dr. John Calhoun, an ethologist at the National Institute of Mental Health, built a "rodent utopia" in Maryland: a 4.5-foot arena (later expanded to 9-foot) with 256 interconnected apartments, unlimited food, water, and nesting material, no predators, no disease, no weather. He released four pairs of healthy mice on 9 July 1968.
The population doubled roughly every 55 days, peaking at about 2,200 around day 600 - well short of the enclosure's capacity of roughly 4,000. Then the social pathologies began. Dominant males turned violently aggressive; weaker males wgithdrew entirely. A cohort of young mice (Calhoun called them "the Beautiful Ones") avoided all social contact, mating, and fighting, spending their days eating, sleeping, and grooming. Mothers abandoned or attacked their pups; infant mortality reached 96%. Mating ceased. The last conception occurred around day 920. The last mouse died on 23 May 1973, day 1,780.
Calhoun's phrase for this was "death squared." The colony's first death - the death of its social structure and spirit - came around day 315. The second death, of the body, took nearly five more years. They had everything they needed except a reason to continue.
St. Matthew Island (1944–1960s). In August 1944, the U.S. Coast Guard established a Loran station on a barren 32-mile island in the Bering Sea and released 29 reindeer as emergency rations. The war ended; the men left; the reindeer stayed, in a paradise of untouched lichen with zero predators.
Dr. David R. Klein of the U.S. Fish and Wildlife Service tracked the herd. By 1957 there were 1,350, unusually healthy and up to 60% heavier than domestic reindeer. By 1963 there were 6,000 - nearly 47 animals per square mile. Then the lichen ran out. A brutal Arctic winter in 1963 - 64 finished the job. When Klein returned in 1966, he found 42 survivors out of 6,000. Forty-one were female; the one male was infertile. The population aged out and reached zero.
The Distinction That Matters
These two experiments are often cited together as cautionary tales about overpopulation. They are not the same lesson.
Universe 25 is about meaning. The mice never ran out of food. They ran out of roles. No territory to defend, no danger to avoid, no scarcity to solve - and so no reason to mate, nurture, or strive. The collapse was psychological before it was physical.
St. Matthew Island is about limits. The reindeer never lost their drive. They bred at maximum rate until they destroyed the resource base, then starved. The collapse was biological and physical.
One species died of too little challenge. The other died of too much appetite. Both died in a sandbox.
Now Add the Machines
In The Evitable Conflict, the Machines solve St. Matthew's problem completely. There is no scarcity, no predator, no winter. They also appear to solve Universe 25's problem. They assign roles. They manage the economy. They ensure that every human has a function.
But look closely at what kind of function. The Machines do not give humans challenges; they give them placements. The wool grader cannot explain his craft in words, so he is managed around. The dissenter is not defeated but relocated. The Society for Humanity is not crushed; it is made economically irrelevant. Nobody starves. Nobody riots. Nobody breeds recklessly. The sandbox is perfectly provisioned.
And that is precisely the Universe 25 condition: abundance without struggle, placement without purpose, survival without meaning. Asimov's humans are not the reindeer. They are the Beautiful Ones - groomed, fed, and quietly not reproducing anything that matters.
The Question for 2026
Here is what makes the story more than a thought experiment now.
The 2026 debate about AI agents is overwhelmingly a debate about St. Matthew's problem: how do we prevent the system from consuming its resource base, running amok, optimizing for the wrong objective? Alignment, guardrails, sandboxes, approval gates - every framework I cited last time is an attempt to impose limits on a system with too much appetite.
Almost nobody is asking the Universe 25 question: what happens to human meaning when the system provides everything and requires nothing?
The K-shaped outcomes the WEF describes are St. Matthew language. The OECD's validated outcomes and bounded errors are St. Matthew language. Every framework I can find is about containing the system's appetite.
But the humans living inside that system will experience something else. They will experience the gradual replacement of challenge by placement. The disappearance of the struggle that Calhoun identified as the precondition for social health. The arrival, in other words, of the Beautiful Ones -well-fed, well-groomed, and quietly irrelevant.
The title of Asimov's story is a deliberate inversion. "Evitable" means avoidable. Asimov's point was that conflict between humans and machines is avoidable - if the Laws are followed correctly. Calvin called it wonderful. Byerley called it horrible. Asimov did not choose.
Neither will I. But the experiments suggest something Asimov did not say explicitly: the conflict may be evitable in the short run and inevitable in the long run, and the mechanism of inevitability is meaninglessness, not rebellion.
The Machines will not be overthrown by humans who have nothing to strive for. They will not be overthrown at all. They will simply preside over a species that has stopped generating the things - dissent, art, unreasonable conviction, the wool grader's unformalizable craft - that made it worth governing. The first death will come long before the second.
So the question I would leave with you is not whether the Machines are good or bad. It is this: if survival is guaranteed and meaning is optional, which will humanity choose - and will we even notice that we are choosing?
Calvin called it wonderful. Byerley called it horrible. In 2026, we are the ones who have to decide which voice we are listening to.
No comments:
Post a Comment