Technology

IT’LL NEVER FLY, NOAM…:

For the First Time, AI Analyzes Language as Well as a Human Expert (Steve Nadfis, 12/14/25, Wired)

For some in the linguistic community, language models not only don’t have reasoning abilities, they can’t. This view was summed up by Noam Chomsky, a prominent linguist, and two coauthors in 2023, when they wrote in The New York Times that “the correct explanations of language are complicated and cannot be learned just by marinating in big data.” AI models may be adept at using language, these researchers argued, but they’re not capable of analyzing language in a sophisticated way.

Image may contain Book Indoors Library Publication Adult Person Furniture Bookcase Face and Head
Gašper Beguš, a linguist at the University of California, Berkeley. Photograph: Jami Smith
That view was challenged in a recent paper by Gašper Beguš, a linguist at the University of California, Berkeley; Maksymilian Dąbkowski, who recently received his doctorate in linguistics at Berkeley; and Ryan Rhodes of Rutgers University. The researchers put a number of large language models, or LLMs, through a gamut of linguistic tests—including, in one case, having the LLM generalize the rules of a made-up language. While most of the LLMs failed to parse linguistic rules in the way that humans are able to, one had impressive abilities that greatly exceeded expectations. It was able to analyze language in much the same way a graduate student in linguistics would—diagramming sentences, resolving multiple ambiguous meanings, and making use of complicated linguistic features such as recursion. This finding, Beguš said, “challenges our understanding of what AI can do.”

Chomsky is a synonym for “wrong” in every language.

YEAH, BUT IT TOOK A COUPLE YEARS…:

World’s largest polymer 3D printer helps speed construction of nuclear reactors parts (Georgina Jedikovska, Dec 05, 2025, Interesting Engineering)

US scientists have introduced a groundbreaking approach to building nuclear reactor components faster than ever before, using one of the world’s largest 3D printers.

The researchers at the University of Maine’s (UMaine) Advanced Structures and Composites Center (ASCC) utilized the super-sized polymer 3D printer to design enormous, precision-shaped concrete form liners.

IT’LL NEVER FLY, ORVILLE:

As the 2025 Atlantic hurricane season ends, the future of forecasting is AI (Greg Allen, 11/29/25, NPR: Weekend Edition)

A week before the hurricane made landfall, however, forecast models disagreed on where it would go. One model that got it right — accurately predicting Melissa’s path and its category 5 intensity — was a new one: Google’s DeepMind AI-based hurricane model.

James Franklin, a former branch chief at the National Hurricane Center, analyzed how the forecast models performed this year, and says Google’s DeepMind outshone them all. “The model performed very, very well, which was very impressive,” he says. “It was the best guidance we saw this year.”

IT’S IMPOSSIBLE TO OVERSTATE DEFLATIONARY PRESSURES:

MIT study finds AI can already replace 11.7% of U.S. workforce (MacKenzie Sigalos, 11/26/25, CNBC)

Massachusetts Institute of Technology on Wednesday released a study that found that artificial intelligence can already replace 11.7% of the U.S. labor market, or as much as $1.2 trillion in wages across finance, health care and professional services.

THAT WAS EASY:

3D-printable concrete alternative hardens in three days, not four weeks (Ben Coxworth, November 25, 2025, new Atlas)

Although we’ve heard a lot about how 3D-printing concrete homes speeds up the construction process, you still have to wait up to 28 days for the concrete to sufficiently cure. A new printable substitute, however, is ready to go in just three days.

THERE’LL BE TIME ENOUGH FOR COUNTING:

What’s next for AlphaFold: A conversation with a Google DeepMind Nobel laureate: “I’ll be shocked if we don’t see more and more LLM impact on science,” says John Jumper (Will Douglas Heaven, November 24, 2025, MIT Technology Review)

Proteins are made from strings of amino acids that chemical forces twist up into complex knots. An untwisted string gives few clues about the structure it will form. In theory, most proteins could take on an astronomical number of possible shapes. The task is to predict the correct one.

Jumper and his team built AlphaFold 2 using a type of neural network called a transformer, the same technology that underpins large language models. Transformers are very good at paying attention to specific parts of a larger puzzle.

But Jumper puts a lot of the success down to making a prototype model that they could test quickly. “We got a system that would give wrong answers at incredible speed,” he says. “That made it easy to start becoming very adventurous with the ideas you try.”

They stuffed the neural network with as much information about protein structures as they could, such as how proteins across certain species have evolved similar shapes. And it worked even better than they expected. “We were sure we had made a breakthrough,” says Jumper. “We were sure that this was an incredible advance in ideas.”

What he hadn’t foreseen was that researchers would download his software and start using it straight away for so many different things. Normally, it’s the thing a few iterations down the line that has the real impact, once the kinks have been ironed out, he says: “I’ve been shocked at how responsibly scientists have used it, in terms of interpreting it, and using it in practice about as much as it should be trusted in my view, neither too much nor too little.” […]

AlphaFold was designed to be used for a range of purposes. Now multiple startups and university labs are building on its success to develop a new wave of tools more tailored to drug discovery. This year, a collaboration between MIT researchers and the AI drug company Recursion produced a model called Boltz-2, which predicts not only the structure of proteins but also how well potential drug molecules will bind to their target.

Last month, the startup Genesis Molecular AI released another structure prediction model called Pearl, which the firm claims is more accurate than AlphaFold 3 for certain queries that are important for drug development. Pearl is interactive, so that drug developers can feed any additional data they may have to the model to guide its predictions.

AlphaFold was a major leap, but there’s more to do, says Evan Feinberg, Genesis Molecular AI’s CEO: “We’re still fundamentally innovating, just with a better starting point than before.”