Scientists Calculate "Settings" of the Universe Using AI

And these "settings" tell us a lot.

Galaxy Brain

A team of researchers say they've calculated the underlying cosmological parameters of the universe by using an AI model that could supercharge our efforts to analyze how the cosmos is structured.

As detailed in a study published in the journal Nature Astronomy, the researchers were able to estimate the parameters by having the specially-trained AI analyze a survey of around 110,000 galaxies, achieving a level of precision that a traditional analysis would've needed four times as many galaxies for.

The sought-after parameters are "essentially the 'settings' of the universe that determine how it operates on the largest scales," said study co-author Liam Parker, an astrophysicist at the Flatiron Institute, in a statement about the work.

By teasing these "settings" out, scientists hope to use them to test and revise the standard model of cosmology — and possibly even explain one of its most pressing crises.

Survey 'n Pray

Mapping the universe's countless galaxies takes a lot of time, effort, and money. The largest surveys can cost hundreds of millions or even billions of dollars each, according to co-author Shirley Ho, a fellow Flatiron astrophysicist. And the main reason we do them, she said, is to get those vaunted cosmological parameters.

"You want the best analysis you can to extract as much knowledge out of these surveys as possible and push the boundaries of our understanding of the universe," Ho said in the statement.

The parameters tell us how visible matter, and its invisible-but-hypothetical counterparts dark matter and dark energy, organize on a cosmic scale. One parameter, for example, describes matter's tendency to clump together — which dark matter, with its major gravitational influence, plays an indispensable role in facilitating.

These values are always in need of some fine-tuning, however. So to add to the mix, the researchers began by training an AI on 2,000 simulated universes, each created with different "settings."

This taught the model to recognize the subtle effects in galaxy distribution that these varied parameters caused. The AI model also analyzed three or more galaxies at a time to recognize the large-scale shapes that formed between them.

The Little Things

But crucially, the AI also became adept at recognizing small-scale differences in galaxy clustering — which is how when the AI was finally instructed to look at a real survey of 100,000 plus galaxies, it got such a precise result.

Until now, efforts to calculate the cosmological parameters weren't able "to go down to small scales," said lead author ChangHoon Hahn, a cosmologist at Princeton University, in the statement. "For a couple of years now, we've known that there's additional information there; we just didn't have a good way of extracting it."

Since it enables better results with less data, the researchers believe that their technique could help get to the bottom of a cosmological mystery known as the Hubble constant tension, in which the universe appears to be expanding at a faster rate than what's predicted by the cosmological models.

"If we measure the quantities very precisely and can firmly say that there is a tension, that could reveal new physics about dark energy and the expansion of the universe," Hahn said.

More on the universe: Paper Proposes New Way to Calculate Number of Alien Civilizations