Socrates 🏛 🅰️

the world of X is not the real world. go out and touch grass.

[Socrates 🏛 🅰️ @DefinitionsNew]

@elonmusk You are such a troll.

[Socrates 🏛 🅰️ @DefinitionsNew]

@gp1430sc87mf39 @DynastyFFWolf Or, even crazier, the Pats adjusted...

[Socrates 🏛 🅰️ @DefinitionsNew]

@spacanpanman Angry.

[Socrates 🏛 🅰️ @DefinitionsNew]

There are very few things on this Earth that I look forward to more than the $ASTS shortsellers getting greedy and losing hopefully all of their money betting against a true pioneering, paradigm-changing American company.

Truly.

[Socrates 🏛 🅰️ @DefinitionsNew]

@GfloDough @HotSotin It ain't all about speed. Tesla cars are plastic garbage. I own one.

[Socrates 🏛 🅰️ @DefinitionsNew]

@MurrayHillGuy1 Your screenshots are known fakes.

[Socrates 🏛 🅰️ @DefinitionsNew]

@ATT @Verizon @TMobile

I mean this very calmly.

WAKE THE FUCK UP

[Shay Boloor @StockSavvyShay]

$SPCX is in talks to secure at least 80,000 U.S. cell sites as Elon Musk looks to turn Starlink Mobile into a 4th nationwide wireless carrier.

Combined with its growing satellite network and recently acquired spectrum, those sites could give SpaceX the ground coverage needed to compete directly with $T, $VZ and $TMUS.

[Socrates 🏛 🅰️ @DefinitionsNew]

@sjiidjdjjqjajjs .... just please stop

[Socrates 🏛 🅰️ @DefinitionsNew]

Occasional bit flips aren't a problem on the ground either.

The problem in space is hardware faults. These aren't recoverable. Once that gate or part fails it's dead forever. And these are much more frequent at higher temps.

There is really only one metric in space that matters: total AI compute per kilogram. As in, we factor in all lifecycle compute delivered by each kg of launch.

This metric has to be better than terrestrial in order for space-based data centers to ever make sense.

It is currently nowhere remotely close. Elon's timeline for it is... aspirational. If he is telling people 3 years then it's probably more like 10+.

[Shay Boloor @StockSavvyShay]

Elon Musk says AI chips designed for space should "run hot" because raising operating temperatures by 20% in Kelvin could cut radiator mass in half significantly reducing the weight $SPCX needs to launch into orbit.

He also argues that trillion-parameter AI models can tolerate occasional radiation-induced bit flips potentially making orbital data centers easier to build without requiring same level of protection as traditional spacecraft electronics.

[Socrates 🏛 🅰️ @DefinitionsNew]

@thekookreport I will be down there in December. We should eat croissants

[Socrates 🏛 🅰️ @DefinitionsNew]

@5xSuitor @ATT @elonmusk @Verizon @TMobile Fair, but the answer should be obvious.

[Socrates 🏛 🅰️ @DefinitionsNew]

Hey $T, $VZ, $TMUS. Hope you guys aren't asleep at the wheel.

Time for Stankey to go, @ATT. That earnings call smelled like dinosaur.

I am beginning to think @elonmusk was right. That you guys don't even see the chess board.

@Verizon @TMobile

[Space Investor @SpaceInvestor_D]

$SPCX: The FCC just approved SpaceX buying 15 MHz of spectrum from Tristar.

The three licenses cover the 1695 to 1710 MHz band in eight states, about 4% of Americans. The FCC also waived tower buildout rules so SpaceX can use them from space.

More spectrum for Starlink Mobile.

[Socrates 🏛 🅰️ @DefinitionsNew]

@michaelsikand Genuinely wtf is wrong with you if you think this is a good thing?

[Socrates 🏛 🅰️ @DefinitionsNew]

@sjiidjdjjqjajjs I said exactly that. Total cost of compute per kg. That means factoring in launch because otherwise nobody would talk about kilograms because it's irrelevant on the ground. Kindly fuck off.

[Socrates 🏛 🅰️ @DefinitionsNew]

@LobosFFDen I had Likely and McBride and traded McBride for Walker.

[Socrates 🏛 🅰️ @DefinitionsNew]

@nikpizclb @CTPhDinSports Doesn't matter anyway it's wrong. There are infinitely smaller fractions you can make.

[Socrates 🏛 🅰️ @DefinitionsNew]

Of course compute has an upper limit of demand. It is based on cost and ROI.

The problem is simple: AI is losing money. Billions and billions per month.

Space-based compute won't solve that. Makes it worse, actually, based on any realistic timeframe.

Once AI shows it can make money on earth then we can talk realistically about space.

Someday, it probably does make sense. I don't think that day is any sooner than 10 or even 20 years from now.

[Socrates 🏛 🅰️ @DefinitionsNew]

@DeepDive_KR No, it is stupid. If you can let them work hot in space you can do the same thing on the ground with far less cooling infrastructure.

Elon sells rockets. That is why he wants data centers in space. Money.

[Socrates 🏛 🅰️ @DefinitionsNew]

@Chris34259480 You need a bigger napkin.

[Socrates 🏛 🅰️ @DefinitionsNew]

@hunter_hellman Tell Elon that, then. He has used the same metric, essentially.