Agreed, with caveats. I don't think that's always the case - I think there are times when he's selling a story that even he himself isn't convinced of, if he feels it's for "the greater good" (which, in his head, is always "whatever his goals are at the time"). But usually, I agree that he smokes his own supply. His general formula for success has been:

* Find some sci-fi thing that nerds really want to happen, that's technically possible but difficult and has seen little progress
* Sell a big story about how you're going to pull it off
* End up getting tons and tons of applications from said nerds with the engineering background needed. Get your pick of the best, who you can work hard in long hours and they won't complain because they want to be there.
* Also end up getting lots of investment from nerdy investors who also want to make said story happen. Get them to approve whatever you want/need to by convincing them that you're the only one who can make the story happen.
* Let your engineers do totally new stuff that can fail but will be very advantageous if it succeeds, so you can iterate faster. Burn money to buy speed.
* Either succeed or not, but you actually have a shot because of all of the talent and money you attract
* On the ones that you fail at, either push it off further into the future, or roll the failure up into some bigger venture and let it slowly fade out in the background. Your ability to keep doing this requires maintaining the image that whatever story you pronounce, you'll succeed at it eventually.

It's the sci-fi equivalent of "fake it until you make it". But it legitimately can work.

It's not really comparable, though. Data centres are (comparatively) point sources of heat; they can't be "spread out" or laid out flat in 2d. Starlink satellites have a, flat surface with its heat spread across it. Also,Sstarlink antennas are their own radiators. They get very hot and correspondingly radiate quickly (waste heat from the thrusters is also quite high temp). Your chips by contrast are running at like 75-80C, and your cooling liquid thus has to be significantly below that to draw heat away quickly enough (on Earth, like 35-45C). You can use heat pumps to cool the coolant below the chip temperature while having a higher radiation temperature, but that increases the needs for everything (power, total heat dissipation, mass, propulsion, etc), esp. if you want a significant thermal "lift", where heat pumps become inefficient.

It's also not a question about whether things are "solved". Nothing butts up against basic physics here. This is an economics question, datacentres on Earth vs. datacentres in space. And you can't talk about the power advantages without also talking about the disadvantages like thermal management.

He's a Martian robot spy send to sabotage the Earth prior to the invasion.

He probably bought shares in the company that makes the bars and considered that good enough.

Duggan: You can't make an omelette without breaking eggs.
Romana: If YOU wanted an omelette, I'd expect to find a pile of broken crockery, a cooker in flames and an unconscious chef.

I dunno, I.... just connect this scene to Musk's diabolical plans.

Misread that as "lunches" and wondered when SpaceX started hiring hobbits from the Shire.

Hmmmm. Microsoft did just fine with lying (even in court), and Enron would have gotten away with it too if it wasn't for those pesky kids and their mangy brownouts.

Psychologists argue that a primary trait of a good CEO is psychopathy, since it requires a personality that has no remorse or compassion and a willingness to do whatever it takes.

Think of how easy it would be for a nation like Russia, the China or the US to blackmail foreign entities that own space datacentres. They can be attacked with far greater plausible deniability, and they're not located in the rival nation's territorial jurisdiction. Massive amount of value all in one place on an eminently predictable orbit that's easy to toss a piece of quote-unquote "space debris" at.

The unspoken issue here is that getting rid of heat (on Earth, all those people complaining about water use of datacentres) is far harder in space. The panels at least radiate *their* heat away from their large rear surface area, but the datacentre itself has to have large amounts of fluid cycling out and back to roughly comparably large radiators.

It doesn't make space datacentres "undoable", by any stretch, but omitting mentioning it and talking only about the power advantages is really dishonest.

I'm assuming they would need Starship working first, and that's going to drop the cost per kg to orbit by a factor of about 10. But I agree there's lots of other challenges.

There was some recent research showing that we are actually much closer to Kessler Syndrome than was originally thought. Basically, if we were to lose control of the existing constellations (and therefore the ability to dodge debris) then the next significant collision would happen in a matter of days, and then the cascading collisions would produce more and more debris. Some of the orbits would stay unusable for years. I don't think you want to launch satellites with these large solar panel arrays into the midst of that.

But to be clear to the GP, that doesn't mean "it's a 10% better model". For most queries that one does for any two models, most of the generations / fixes will be "good", and so it's just basically a coin flip as to which model to choose ("I like this one's documentation more", "This one's fix was more concise", "This model was more polite", etc). 10% is actually a pretty big difference and reflect the cases where one model was unambiguously better than the other.

I'm not seeing a paywall on the site. uBlock Origin must be doing something.

The UHF app on our Apple TVs & iOS devices and the UHF Server in Docker to act as a PVR gives us everything for a few $ a month paid in crypto.
We haven't had cable since ~1999-2000. Downloading and the *arrs have kept us happy, but the better half wanted to check out some live sports. So IPTV it was.

I've been frustrated by AI constantly making shit up and stating things as fact when it's clearly wrong. But then I realized if you read around the internet that it's full of posts that are just factually untrue. Like I'm not talking about opinion stuff... I mean people posting absolutely untrue "facts" even on technical forums where a very cursory google search with turn up multiple independent sources that show they're completely wrong.

So, it turns out that if you can make a program that's only correct 90% of the time, you've already created a program that's correct more often than the average employee, so it's probably useful. This is completely backwards to how I think about computers, but I can't deny it. Humans are just a lot less useful than I thought, particularly if you need them to make correct decisions.