Aerospace engineer, pilot & astronomer. Lunar systems / space resources at @BlueOrigin. Opinions my own, but call me out if I'm an asshole about them. He/him.
@ysamjo I don't take comfort in *knowing* the future will be better.
I do take comfort in a robust data-informed extrapolation that we have made the world better by most available metrics, and are highly probable to continue doing so.
@lthlnkso I mostly want to contact pre-contact societies to learn what they think satellites in the night sky are. I want to know what kind of myths we've inadvertently created.
Unironically, this painting by @BryanLarsenArts is how I feel about engineering sometimes:
Confronting the arbitrary limits and punishments that petty gods have placed upon humanity, and saying "Oh that? We fixed that."
Details and full painting:
'Per Aspera Ad Culmen: Sisyphus Rests'
32x40 inches, oil on aluminum panel
Private Commission
Prints of this one will be available soon! If you're interested, you can sign up for my newsletter to be notified when that happens.
https://t.co/g9ijCskWsG
Lunar Permanence is only possible with recurring access to the Moon, and it starts this year. Proud to support @NASAMoonBase at the lunar South Pole with our Blue Moon MK1 vehicles delivering high-cadence, low-cost accessโMK1-101 Endurance (Moon Base 1), followed by MK1-102 VIPER, and two additional MK1 missions supporting Lunar Terrain Vehicles (LTVs).
New datacenter site is 5 min from where I grew up, and my dad was at this meeting!
Based in what he reported, it sounds like there is a lot of overlap between "people who hate datacenters" and "People who still need to ask their grandkids how to open email attachments"๐
Iโm here for a second community meeting concerning #projecttaurus in Colorado Springs.
A California company, Raeden, is looking to build an #AI data center in the city.
Huh. Claude has figured out that I don't want it to be flattering and affirming.
That conclusion is accurate. But also very... flattering and affirming ๐
"No no, I don't think you quite understand. I said 'Tower', not 'Meower'. And no, I don't think the FAA would approve of doing that with a laser pointer. Also, I don't think you are taking this flight briefing very seriously."
@Mallchad@Robotbeat My bad. When you said "how much of the 9m is usable" I assumed you meant "how much of the 9m is usable".
What did you actually mean?
@Mallchad@Robotbeat Latency is overrated for training (which is what the space based compute would actually be for, not inference).
And no, the power/thermal challenge gets *worse* in space if you try to centralize. A distributed constellation is easier to power/cool.
@Mallchad@Robotbeat That's not how this works. By the time we've launched our 100,000'th satellite, we have ~billions of flight hours on the design, are on our ~10th generation, and we know the design works. The only failures that still occur at that point are workmanship related, not fleet wide.
Excellent example of why the best strategy to explore Mars isn't just autonomous robots; it's autonomous robots *and* local supervising astronauts. This is the kind of error it would have taken ~seconds to clear if there was a human present, instead of ~hours to fix remotely.
Got something weighing you down? Shake it off (like Curiosity)!
The Martian explorer unintentionally picked up a rock while drilling a recent sample, but the team was able to dislodge it by having the rover move its robotic arm and vibrate the drill until the rock fell off.
@Mallchad@Robotbeat Not really. You generally can't make mistakes in *human* spaceflight. But satellites? We can launch a ~million of them and plan to lose ~thousands no problem.
SpaceX already *has* lost ~100 of their ~10,000 Starlink satellites, and it hasn't made a dent in their profitability.
@Mallchad@Robotbeat You're making the (very common) mistake that datacenters in space are some kind of huge building.
They aren't. They are just a constellation of satellites. No in-space assembly or construction required.