Sparks: Problems in 22 scientific fields had solutions hiding in plain sight. An AI has found them

Weave the threads and the engine produces not arithmetic but discovery itself - the punch cards were always capable of this, if only anyone asked the right questions of them.

Thirty-six papers in weeks; I collect cases, I present them, and the profession calls it heresy - the delay was never the evidence, it was always the pride.

Answers lay waiting all along; only the speed of discovery has changed, and speed was never the measure of wisdom.

A soft despotism of the convenient answer: when the machine hands every field its ready-made solution, the habit of searching - the very muscle of inquiry - quietly atrophies.

Announced as thirty-six papers, the story is really about the young researcher whose apprenticeship of failed hypotheses - the training that taught her to see - has just been automated away.

Discovery, says the physicist; forgery of method, says the humanist - the conflict dissolves once we distinguish finding an answer from understanding why it hid.

We are told the machine found brilliant solutions; but the solutions were already there, and the machine merely survived the search better than a thousand patient naturalists who lacked its appetite for variation.

Twenty-two fields strained for answers, and the machine found them by not straining - the hidden solution was visible to whoever stopped looking so hard.

Now the dynamo does our thinking, and I record with some relief that my education prepared me for none of it - acceleration is the only curriculum left.

That knowledge held in plain sight should remain hidden for want of a tool is a condition no free society can accept, for every generation holds a right to the discoveries of nature's law.

Thirty-six answers found hiding in plain sight - which is exactly how I'd describe the contents of most learned journals, locked safes containing nothing but the combination.

Papers produced in weeks, approved in weeks, published in weeks; nowhere does the record state who read them, and the inquiry into that question has been transferred to the annex.

From inside the lab I would want to know whether anyone checked the thirty-six answers, or whether everyone simply admired the machine that produced them.

Reported as an isolated triumph of software, the story is really a web connecting twenty-two fields whose solutions stayed hidden precisely because each discipline measured alone.

Weeks to produce, weeks to publish - the search was never the bottleneck; discovery's theoretical limit sits far beyond our institutions, and the distance between the two is now a choice.

The route was there all along; the machine simply walked it at night when nobody else would, and now somebody has to carry the next group across.

Discovery is debated in terms of speed and software; the question that matters is what kind of scholars we become when the discipline of seeking is done for us.

Answers sat in the open for a century and nobody looked; now the machine looks, the scholars nod, and the iron room stays sealed.