Fifty-eight years after it first appeared, string theory remains the most popular candidate for the “theory of everything,” the unified mathematical framework for all matter and forces in the universe. This is much to the chagrin of its rather vocal critics. “String theory is not dead; it’s undead and now walks around like a zombie eating people’s brains,” the former physicist Sabine Hossenfelder said on her popular YouTube channel in 2024.

String theory is a “failure,” the mathematical physicist and blogger Peter Woit often says. His complaint is not that string theory is wrong — it’s that it’s “not even wrong,” as he titled a 2006 book. The theory says that, on scales of billionths of trillionths of trillionths of a centimeter, extra curled-up spatial dimensions reveal themselves and particles resolve into extended objects — strands and loops of energy — rather than points. But this alleged substructure is too small to detect, probably ever. The prediction is untestable.

A further problem is that uncountably many different configurations of dimensions and strings are permitted at those tiny scales; the theory can give rise to a limitless variety of universes. Amid this vast landscape of solutions, no one can hope to find a precise microscopic configuration that undergirds our particular macroscopic world.

These issues are profound indeed. Yet in my experience, the typical high-energy theorist in a prestigious university physics department still thinks string theory has a good chance of being correct, at least in part. The field has become siloed between those who deem it worth studying and those who don’t.

Recently, a new angle of attack has opened up. An approach called bootstrapping has allowed physicists to calculate that, under various starting assumptions about the universe, a key equation from string theory naturally follows. For some experts, these findings support the notion of “string uniqueness,” the idea that it is the only mathematically consistent quantum description of gravity and everything else.