199 sats \ 1 reply \ @k00b 25 Feb
This is awesome and expect AI to do a lot more of this kind of thing for us - see patterns in non-human quantities of data. Practically all domains can benefit from this. One wtf happened in 1971 theory is that we solved all human scale problems. (1971 is likely a gold standard thing but perhaps going off the gold standard was driven by running out of human scale problems to solve.)
reply
I was also intrigued by this. It's exciting to think of all the places where ML techniques have a chance of instantly detecting useful order in chaos.
reply
let's do this
reply
The latest success means another significant obstacle has been passed, with the AI capable of recognising plasma instabilities 300 milliseconds before they happen – enough time to make modifications to keep the plasma under control.
Published in Nature.
reply
however harnessing nuclear fusion energy has proved immensely difficult
The plant kingdom (and algae & cyanobacteria) have been harnessing nuclear fusion energy for billions of years, and humanity has developed the technology to do it too (photovoltaic cells) and is building out this technology extremely rapidly now and providing the cheapest source of energy that humanity has ever had.
Nuclear fusion energy generated by the sun is massively abundant. It is also 'safe' (under the Earth's magnetosphere at a distance of 1 AU) and the reactor is entirely self maintaining and will run for billions more years.
This article is putting together buzzwords of fusion and 'AI', two things beloved by mediocre, uncritical intellects. Putting money into fusion research and development (and quite large sums) is a waste when it could go towards solar and batteries, both of which are technologies which promote decentralisation (the opposite of nuclear power plants).
reply