On August 20, the MIT Technology Review declared that artificial intelligence (AI) will help predict space weather.
Yes, space has weather. Serious weather. It comes from eruptions from the sun called Coronal Mass Ejections (CME), which fling a vast quantity of mass into space. It can be dangerous to us if one of these solar expectorations intercepts the Earth. That last big one happened in 1859, known now as the Carrington Event, which disrupted telegraph systems worldwide. Sparks sometimes flew from telegraph keys, shocking the operators. Others were luckier: there were several reports of telegraph operators able to send messages even after disconnecting their batteries, their signals literally powered by solar wind.
Given the total permeation of electronics into our lives, the effects of a Carrington event today would be far more serious than the 1859 event. Lloyds of London has estimated that the cost of a Carrington event to the United States alone would be roughly $3.3 trillion.
Astrophysicists have a pretty good handle on the causes of CMEs. What they don’t have is a good means of predicting where on the sun and when a CME might occur. If one erupts from the poles of the sun, we on Earth will not be affected. If a CME is aimed directly at us … well, Katie, bar the door.
Given the high stakes, space weather is closely monitored and predicted, which you can follow on several sites, both private and public. The problem is predicting where on the sun and when a CME would erupt. This is where Surya: The Foundation Model for Heliophysics comes in. This applies machine learning methods—the same methods that gave us ChatGPT—to NASA’s databases of past solar weather to refine predictions of space weather, including the crucial prediction of where on the sun and when there will be a CME.
As science, Surya is pretty solid. If it works, it will be a helpful improvement in everyone’s lives. But take a look at the MIT Technology Review article: it’s full of weasel words.
“Hope” makes an appearance twice. So do “believe” and “possible.” “Could” shows up six times. The article has “publicist” written all over it. Just as Jennifer Lopez has a publicist whose job is to keep J-Lo’s face and adventures in the public eye, so too do scientists—or, more precisely, the institutions that employ them. J-Lo can afford to pay a publicist because it boosts her revenue, even if the only story that week is about the person who walks her dog.
The same logic applies to “science journalism.” The science itself—like Surya’s work—is solid. It could operate entirely out of the public eye and still accomplish a great deal. So why hire a “science journalist” as a publicity agent? Short answer: it secures a revenue stream for the institution. The scientist gets fed, to be sure, just like J-Lo’s dog is fed from J-Lo’s wealth. J-Lo is the one who gets fat, though.
Hence the perversity of the “science journalism” racket. It’s not about science; it’s about publicity and the revenue that publicity brings. Read with a jaundiced eye.
Follow Scott Turner on X and visit our Minding the Science column for in-depth analysis on topics ranging from wokeism in STEM, scientific ethics, and research funding to climate science, scientific organizations, and much more.
Image: Logo of MIT Technology Review by MIT Technology Review on Wikipedia
Bear in mind one other thing — something similar would happen if someone touched off a nuke 250 miles up in space, and it is similar to lightning causing static on AM radios.
A wire — any wire — acts like a radio antenna, turning electromagnetic radiation into an electrical current. Carrington was 19th Century telegraph wire, probably 1/4 inch copper, and largely able to carry the amperage without melting.
Light bulb filaments are much smaller wire, with higher resistance, so the nuke in space (Starfish Prime) fried some 300 streetlights in distant Hawaii. I once witnessed a lightning strike fry all the (then) incandescent street lights in the area, even though they weren’t turned on yet.
The concern about EMP is that all of the microscopic wires in all of our microchips will also pick up the signal, except these wires are so thin that any unshielded microchip will be fried. Even our toasters won’t work, except we won’t have any electricity because it would also fry the grid.
It’s really quite scary because we’d instantly be back in the 17th Century — without 17th Century technologies such as wells, outhouses, and fireplaces for cooking. There’d be no city water, no city sewer — people can only survive a day without potable water, and human waste has to go somewhere. Gasoline engines and post 2007 Diesels wouldn’t run so there’d be no trucks, no refrigeration — i.e. no food.
Cities would break down FAST…
Bear in mind it’s a combination of two things — a massive CME the size of the Carrington one, and it being sent to where the Earth will be when it arrives. It’s a bullet being blindly fired from 93 million miles away — overwhelming odds are that it will miss.
The sun is an uncontrolled hydrogen bomb, it isn’t exactly predictable. This stuff happens all the time, it’s what (indirectly) causes the Northern Lights.
Estimating percentage of possibility may be possible if you crunch enough numbers. And it is mathematically possible to tell exactly where a CME has to be to hit the earth at any given time. But to know there will be one there, then, I don’t think so…