Thursday, October 10, 2024

Space & Physics: Is AI taking over science?

October 10 — This week, the Nobel Prize in Physics goes to pioneers in machine learning. Plus, a new European mission launching to a familiar asteroid, a "super-Mars" exoplanet found orbiting a nearby star, the complex darkness of the night sky, and more. Enjoy!

--Lee Billings, Senior Editor, Space and Physics


The 2024 Nobel Prize in Physics was given to John Hopfield and Geoffrey Hinton for development of techniques that laid the foundation for revolutionary advances in artificial intelligence

A Nobel Prize in Physics for... AI?

For science journalists such as myself, covering the Nobel Prizes can be a nerve-wracking experience. In October of each year as we dutifully await the announcement of the latest prizewinners, many of us wonder if the Nobel Committee will throw us a topical curveball. And this year's Physics Nobel for breakthroughs in machine learning was certainly a curveball, as the research involved scarcely concerns physics at all, but rather dwells in the realm of computer science. The tech-heavy trend continued with this year's Chemistry Nobel, which also went to machine-learning approaches—this time for predicting and designing protein structure and function.

Besides inspiring some obvious jokes ("Will next year's Nobel Prize in Literature honor the inventors of Microsoft Word's autocomplete function, perhaps?"), the trend does highlight a very real concern: Just as "one device to rule them all" smartphones have rendered most other general consumer electronic gadgets obsolete, are artificial intelligence and machine learning going to do the same thing for scientific research? Is this trend of tech-heavy Nobels a signal of The End of Science?!

Dear reader, I won't pretend to really know the answer, but my guess is that the much-hyped rise of AI isn't the existential threat to old-fashioned flesh-and-blood ingenuity and intuition that some pundits proclaim. Not yet, anyway. These new technologies are first and foremost tools for human use, much like their predecessors the integrated circuit and the optical fiber (both of which also received Nobels). In most cases, actually making them useful will still require having humans involved. The relationship is symbiotic. Real science demands a lot of simple-but-hard work at which humans excel.

The real imminent threat science now faces from this notional rise of the machines is the prospect of too many scientists and their sponsoring institutions prematurely buying in to the hype, favoring and funding AI-infused projects and proposals at the expense and exclusion of all others. And in this respect, at least, 2024's tech-heavy science Nobels may be an ill portent indeed.

Lee Billings

Top Stories
How Dark Is the Night Sky?

The night sky isn't perfectly dark—instead it glows faintly, and the source isn't exactly local

Astronomers Spot a 'Super-Mars' Exoplanet around Barnard's Star

Claims of worlds orbiting Barnard's star have been made before. But an advanced instrument could provide the breakthrough that finally confirms the star hosts a planetary system

A Century-Old Question Is Still Revealing Answers in Fundamental Math

Mathematicians have made lots of recent progress on a question called the Mordell conjecture, which was posed a century ago

Infinite Tiling Presents a Modern Mathematical Challenge

Today's mathematicians grapple with higher-order mathematical questions and real-world applications.

Europe's Hera Mission Launches to Visit an Asteroid Smacked by NASA

Hera will arrive in 2026 at Dimorphos, an asteroid deliberately struck by NASA's DART spacecraft

Can Scientists Save the World from an Apocalyptic Asteroid Strike?

Sooner or later a doomsday asteroid will wipe out most life on Earth—unless, that is, we prevent threatening space rocks from hitting us in the first place

What We're Reading
  • Both this year's Nobel Prize in Physics and Nobel Prize in Chemistry went to breakthroughs in artificial intelligence rather than in physics or chemistry. Is this a cause for concern—or for celebration?  | Marcus on AI
  • A severe solar storm washing over Earth today and tomorrow could spark spectacular low-latitude auroral displays—and potentially damage power grids. | Space.com
  • Hurricane Milton's terrifying rampage through the Gulf of Mexico was visible from the International Space Station, and the monstrous storm's impacts were felt there, too, as its effects forced a crew of astronauts to delay their homecoming. | NASA

From the Archive
What Is Machine Learning, and How Does It Work? Here's a Short Video Primer

Deep learning, neural networks, imitation games—what does any of this have to do with teaching computers to "learn"?

Scientist Pankaj

Today in Science: Humans think unbelievably slowly

...