Is it any easier to design AI in space versus doing it on Earth?
It’s plausible that it’s helpful to have computers that are stored at a place that can easily be cooled down to very cold temperatures. Maybe someone finds a way to build quantum computer that needs superconductivity.
Just a reminder that the LHC is pretty big, and cooled to 3K. Superconductivity produces very little waste heat. Refrigeration isn’t that hard. We aren’t going to see space based quantum compute until keeping it cool on earth is impractical. By the time we are dealing in quantum computers larger than the LHC, we are looking at more than enough compute to brute force superintelligence. (Probably, there is some chance of giant space computers but no superintelligence if people are really squandering the compute.)
Cooling the LHC does cost massive amounts of electricity:
CERN uses 1.3 terawatt hours of electricity annually. That’s enough power to fuel 300,000 homes for a year in the United Kingdom.
It’s plausible that shipping a kilo to the moon with SpaceX starship will soon cost ~50$. That’s a low cost for shipping computer chips.
In addition iron/alluminion is very plenty on the moon so you can easily create things that are to be used on the moon but that isn’t complex like casing out of them.
The LHC seemed to cost around $4.75 billion to build and weights around 14,000-tonne. That suggest you need something like 140 starship trips to bring it up. At $5 million per trip that’s $1 billion.
If a LHC that that would operate at −240 C would cost half of the amount to build you might save a billion by shipping it up to the moon.
Just a reminder that the LHC is pretty big, and cooled to 3K. Superconductivity produces very little waste heat. Refrigeration isn’t that hard. We aren’t going to see space based quantum compute until keeping it cool on earth is impractical. By the time we are dealing in quantum computers larger than the LHC, we are looking at more than enough compute to brute force superintelligence. (Probably, there is some chance of giant space computers but no superintelligence if people are really squandering the compute.)
Cooling the LHC does cost massive amounts of electricity:
It’s plausible that shipping a kilo to the moon with SpaceX starship will soon cost ~50$. That’s a low cost for shipping computer chips.
In addition iron/alluminion is very plenty on the moon so you can easily create things that are to be used on the moon but that isn’t complex like casing out of them.
The LHC seemed to cost around $4.75 billion to build and weights around 14,000-tonne. That suggest you need something like 140 starship trips to bring it up. At $5 million per trip that’s $1 billion.
If a LHC that that would operate at −240 C would cost half of the amount to build you might save a billion by shipping it up to the moon.