MegaWatt Chargers, AI-Physicists, & Fusion Fuels
This week’s science bits from SWTG
China Gets Ultra-High-Speed EV Chargers
Image: BYD
Chinese car maker BYD has revealed a new EV charger that reaches unprecedented speed, replacing battery charge worth 400 km in about five minutes. Their “Super E-Platform” runs at 1000V and 1000A which amounts to 1 MW, about three times more power than the current fastest chargers. BYD says it plans to install 4000 such chargers across China.
But even BYD can’t defy the laws of physics: The faster you force energy through a wire, the more heat will be generated. The super-fast chargers hence need to come with extra cooling systems. It remains to be seen what they will do to battery lifetimes. More here.
AI Comes for Physics
Liam Fedus has left OpenAI and founded a materials science AI startup. In a post on X/Twitter he writes that “OpenAI is planning to invest in and partner with my new company.”
I think that the coming generations of AI models will be able to do most of the current physics research equally well, if not better. Materials science is an obvious starting place, especially in the range of metamaterials that are just about to hit the consumer market, and quantum materials that are about a decade away from applications.
However, seeing that most current physics research is useless, the most obvious monetization strategy for physics-AI tools is to squeeze money out of academia, which seems to be what Google is aiming at with their AI Co-scientist. In related news, people are reporting the first likely AI-written papers on the arXiv. This, I am afraid, is only the beginning.
Nuclear Fusion Fuel Industry Springs Up
Construction plan for the tritium production plant in Oxford. Credit: UKAEA
In the past decade we have seen dozens of nuclear fusion startups spring up (see here for an overview). However, one of the big unanswered questions is where the fuel is supposed to come from, notably tritium (an isotope of hydrogen) and Helium-3, which are used in some of the most commonly pursued nuclear fusion technologies.
We now hear that the United Kingdom Atomic Energy Authority has teamed up with the Italian energy company Eni to construct a tritium production plant near Oxford. The U.S.-based startup Interlune, meanwhile, wants to mine Helium-3 on the moon, with plans to begin operation as early as 2027. The Helium-3 supplies on the moon have an estimated value exceeding 10, 000 trillion US$ at the current price.