Elon Musk credits Douglas Adams with the insight that the universe is the answer — the hard part is the question. AI just proved Adams right by making answers free. The scarce act is now asking.
Elon Musk calls Douglas Adams his favorite philosopher. The insight he credits Adams with is simple: the universe is the answer — what is the question?
In The Hitchhiker's Guide to the Galaxy, a supercomputer called Deep Thought spends 7.5 million years computing the Answer to the Ultimate Question of Life, the Universe, and Everything. The answer is 42. Nobody knows what to do with it — because nobody knows what the question was. Deep Thought explains the problem: the answer is meaningless without the question. Then it designs an even more powerful computer to find the question. That computer is the planet Earth. It runs for ten million years. Adams wrote this in 1979. Forty-seven years later, the joke has become the operating condition of the economy.
The Answer Economy
Every previous technology made a specific class of answers cheaper. Fire answered how do we survive the cold. The printing press answered how do we distribute knowledge. The compass answered how do we navigate without landmarks. The telegraph answered how do we communicate across distance. The internet answered how do we find information. Each revolution collapsed the cost of one category of answer. AI is the first technology that collapses the cost of all answers simultaneously.
Alexander Storozhuk named it in Inc: the answer economy. The unit of value is no longer the list of sources — it is the finished response. Just as oil defined the industrial age, answers define the AI age. The capabilities that once required years of education — analysis, synthesis, pattern recognition — are now executable instantly, at near-zero marginal cost. Intelligence, as a commodity, arrived.
But Adams already knew the answer wasn't the hard part.
The Three Costs That Disappeared
Two days ago, Alonda Williams wrote in GeekWire that everyone is asking AI better questions but nobody is asking themselves better ones. She identified three costs that AI eliminated overnight.
The social cost. You cannot ask a colleague without spending relational capital — the unspoken calculus of whether the question reveals ignorance, whether it wastes their time, whether it damages your standing. AI has no opinion of you.
The time cost. No appointments, no waiting, no scheduling around someone else's availability. The answer arrives in seconds.
The judgment cost. The inner filter that kills the question before it is ever spoken — the voice that says that's a dumb question or you should already know this. AI does not judge.
Williams's observation is precise. Those three costs were friction. But friction is also signal. The social cost forced you to decide whether the question was worth the relationship risk — which meant you refined it before asking. The time cost made you sit with the question long enough to sharpen it. The judgment cost filtered out the trivial. Remove all three, and questions become frictionless. Frictionless questions produce frictionless answers — smooth, immediate, and shallow.
The leaders Williams described saved time and could not say what they gained. They optimized the asking without upgrading what was asked.
The Inversion
This is not a philosophical observation. It has immediate structural consequences.
The companies that survive the AI transition will not be the ones with the best models. Seven frontier models launched from six organizations in twenty-nine days earlier this year and scored within two percent of each other on every benchmark. Models commoditized on the same timeline as answers. The companies that survive will be the ones that ask better questions — what should we build, what should we ignore, what would change our minds. These are questions no model generates on its own. Models are answer machines. They are magnificent at responding and structurally incapable of inquiry.
The same asymmetry operates at every scale. The student who asks AI to write the essay learns less than the student who asks AI to challenge the thesis. The executive who uses AI to generate reports is less valuable than the one who asks which reports should not exist. The researcher who summarizes the literature with AI misses less than the one who asks what the literature has not considered. In each case, the answer is free. The question is the creative act.
This journal exists because of that asymmetry. Every entry asks a question that AI would not generate on its own — not what happened with tariffs but what does the tariff structure reveal about the nature of executive power. Not summarize the jobs report but what does it mean when fewer people work and each one earns more. The answers to those questions are available to anyone with a prompt. The questions themselves are not. They require a point of view, a frame, a willingness to be wrong about what matters.
The Harder Computer
Adams's joke had a second layer that Musk does not usually mention. Deep Thought could not find the question itself. It was the most powerful computer ever built — and the question was beyond it. So it designed something more powerful: not a faster processor, but an entire planet of living beings whose messy, embodied, irrational existence would generate the question through the act of living. The answer required calculation. The question required experience.
That distinction holds. AI can compute any answer from data that already exists. The questions worth asking come from contact with the world — from noticing what the data does not contain, from feeling the friction that the frictionless system removed, from wondering about things that have no prompt. The question is upstream of the answer the way intention is upstream of action. You cannot optimize your way to a good question. You can only live your way to one.
The singularity is not the moment machines give us the answer to life, the universe, and everything. It is the moment we realize the answer was never the hard part. Deep Thought knew. Adams knew. The planet-sized computer built to find the question ran for ten million years and was demolished five minutes before finishing. We are somewhere in those final minutes now — surrounded by answers, still searching for the question worth asking.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)