Orbital AI datacentres come with latency issues.
-
Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.
The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.
Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.
@david_chisnall Good idea. An additional advantage is that boosting an AI bro into LEO requires considerably less energy than the popular alternative proposal of shooting him into the Sun.
-
Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.
The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.
Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.
@david_chisnall I see a fast path to having more than $1 trillion USD worth of assets in LEO then!
-
@david_chisnall @catsalad I’m giggling just imagining this … https://youtu.be/IBYS6VdRzyA
-
@tuban_muzuru No shade intended, just curious: how much have you personally interacted with adult Americans lately?
I'm a US expat--moved to Canada 36 years ago, naturalized (so dual citizen) long since, but still in touch with friends and colleagues down there, still visit a few times a year (except last couple of years, for personal rather than political reasons), and when I visit I travel by train so plenty of informal, non-stressful interactions with strangers. Since my travel is often centred on attending a technical conference, usually about computing systems research or system administration, I see plenty of tech people and often visit the Bay Area or other tech centres.
It's rare that I encounter a MAGA follower or other creep or jerk. Most people I meet are still pretty decent. Even those with political views I oppose are generally decent, civil, and non-extremist. Even (indeed especially) the non-tech people in the service industry--staff in hotels and shops and on trains, for example.
-
@tuban_muzuru No shade intended, just curious: how much have you personally interacted with adult Americans lately?
I'm a US expat--moved to Canada 36 years ago, naturalized (so dual citizen) long since, but still in touch with friends and colleagues down there, still visit a few times a year (except last couple of years, for personal rather than political reasons), and when I visit I travel by train so plenty of informal, non-stressful interactions with strangers. Since my travel is often centred on attending a technical conference, usually about computing systems research or system administration, I see plenty of tech people and often visit the Bay Area or other tech centres.
It's rare that I encounter a MAGA follower or other creep or jerk. Most people I meet are still pretty decent. Even those with political views I oppose are generally decent, civil, and non-extremist. Even (indeed especially) the non-tech people in the service industry--staff in hotels and shops and on trains, for example.
@tuban_muzuru So although billionaire tech CEOs and their toadies are indeed loathesome, the vast majority of people are decent. Even some of the 30% or so who blindly support Trump are more deluded than fundamentally nasty, and are probably salvageable.
Or so it feels to me.
-
Yes...
The humor in the final clause of the final sentence gets too dark for me, though (in the context of actual space flight).
I merely express a reaction; this is not a veiled call for censorship.@vnikolov @david_chisnall I counter that it would actually get pretty bright and isn't dark at all.
-
Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.
The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.
Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.
@david_chisnall I strongly disagree. Their exceedingly divine nature leaves the very centre of the solar system as the only acceptable place for AI and bros.
-
@david_chisnall I strongly disagree. Their exceedingly divine nature leaves the very centre of the solar system as the only acceptable place for AI and bros.
@hllizi @david_chisnall That takes far too much delta-v, unfortunately. I think we'll have to settle for LEO.
-
Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.
The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.
Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.
@david_chisnall that does not even account for heat dissipation and power needs.
-
Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.
The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.
Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.
@david_chisnall Remember that Tesla Elon sent into solar orbit some years ago with the dummy driver? In retrospect he missed a business opportunity there.
He should've started a regular "Ride a Tesla to Space" service, sending his billionaire mates, on grand tours of the solar system in Teslas. When they'd all been launched and the process had become routine, we could all chip in a few bucks each to give Elon his own ride around the Sun.
-
P pelle@veganism.social shared this topic
