Meanwhile, demand is rising for magnets embedded in the tools of decarbonization, such as cars and wind turbines. Currently, 12 percent of rare earths go into EVs, according to Adamas Intelligence, a market that’s just now taking off. At the same time, rare earth prices have recently whiplashed due to internal Chinese markets and political interventions that outside companies cannot always predict. 

All in all, if you’re in a business where you can make an alternative work, it probably makes sense to do so, says Jim Chelikowsky, a physicist who studies magnetic materials at the University of Texas, Austin. But there are all kinds of reasons, he says, to look for better alternatives to rare earth magnets than ferrite. The challenge is finding materials with three essential qualities: They need to be magnetic, to hold that magnetism in the presence of other magnetic fields, and to tolerate high temperatures. Hot magnets cease to be magnets.

Researchers have a pretty good sense of what chemical elements can make good magnets, but there are millions of potential atomic arrangements. Some magnet hunters have taken the approach of starting with hundreds of thousands of possible materials, tossing out those with drawbacks like containing rare earths, and then using machine learning to predict the magnetic qualities of those that remain. Late last year, Chelikowsky published results from using the system to create a new highly magnetic material containing cobalt. That’s not ideal, geopolitically speaking, but it’s a starting point, he says.

Often the greatest challenge is finding new magnets that are easy to make. Some newly developed magnets, such as those containing manganese, are promising, Vishina at Uppsala University explains, but also unstable. In other cases, scientists know that a material is extraordinarily magnetic but can’t be create in bulk. That includes tetrataenite, a nickel-iron compound known only from meteorites that must slowly cool over thousands of years to precisely arrange its atoms into the correct state. Attempts to make it more speedily in the lab are ongoing but have yet to bear fruit.

Niron, the magnet startup, is a little further along, with an iron nitride magnet that the company says is theoretically more magnetic than neodymium. But it too is a fickle material, hard to make and preserve in a desirable form. Blackburn says the company is making progress but won’t be producing magnets powerful enough to transform EVs in time for Tesla’s next generation of vehicles. First step, he says, is to put the new magnets in smaller devices like sound systems.

It’s unclear whether other automakers will follow Tesla’s rare earth trade-off, Kruemmer says. Some might stick with the baggage-laden materials, while others go with induction motors or try something new. Even Tesla, he says, probably will have a few grams of rare earths sprinkled in its future vehicles, spread across things like the automatic windows, power steering, and windshield wipers. (In a possible sleight of hand, the slides contrasting rare earth content at Tesla’s investor event actually compared an entire current-generation car to a future motor.) Despite workarounds like those in the works at Tesla, rare earth magnets sourced from China will remain with us—including Elon Musk—especially as the world pushes to decarbonize. It might be nice to replace everything, but as Kruemmer says, “we simply do not have the time.”