@jkf's banner p

jkf


				

				

				
0 followers   follows 0 users  
joined 2022 September 04 19:07:26 UTC

				

User ID: 82

jkf


				
				
				

				
0 followers   follows 0 users   joined 2022 September 04 19:07:26 UTC

					

No bio...


					

User ID: 82

Farmers typically stockpile quite a bit of diesel just by nature of living far from a gas station and getting regular deliveries -- not enough to farm, like, forever, but surely enough to get in a bit of a harvest and bootstrap some veggie oil or whatever. (which is kind of exactly "making your own diesel" -- not enough to sustain operations at current levels, but surely enough to keep things running on a survival basis)

Similarly, "far from parts/services" means that farmers are absolutely exceptional at jerry-rigging things to get the harvest in when necessary -- modern equipment is definitely more complex than it used to be, but the basic operating principles have hardly changed at all since the 70s and "rip out that damn computer and make the thing work" is well within the capabilities of a clever farmer in a long-term collapse situation. (The ones I know never actually sell anything either, and will probably have several combines of various vintages rusting in a hedgerow someplace that could be reactivated if the need arises)

A farmer with a welder is dangerous and effective at the same time...

The output of a wheat farm is... pretty much wheat berries? ("Combine" is short for "Combine harvester", so called because the threshing is combined with the harvesting)

You could make it into porridge in extremis, but making flour is not real hard either -- a grain farmer ain't gonna starve, and if you happened to put in canola the year the balloon went up you could probably trade veggie-diesel for anything you can find until the cows come home.

More importantly I'm pretty sure Starlink relies on ground stations for actual connectivity for the time being -- I've no doubt that Elon's goal might be to phase those out with sat-sat connectivity and maybe some of these space servers, but if there's no internet to connect to I think the subnet goes down as well in this case?

4 miles per hour is pretty sustainable if you are fit and on smooth terrain -- see, uh, Jack London IIRC? 15 hours of constant fast-walk would be tiring, but I guess if there's a horde of urban zombies at your back it would be good to pick up the pace to... somewhat above average!?

When I say that it's a more polite way of saying "you are overlooking this obvious thing; don't be retarded"

eg:

  • "Trump is obviously corrupt; look at how much money he has!"

  • "I mean, Trump had a lot of money before!?"

Oh I know -- that's the point, broken promises all around!

Anyways I'm not say "FSD never", I'm saying that the skeptics were correct to be skeptical of the promises made around FSD in the early 2010s...

(heavily influenced by the left tail that is responsible for 80% of the fatalities). ... if everyone used it 95% of the time like I do I believe we'd have a lot fewer accidents.

These points interact though -- that left tail will be the very last to use it; probably not unless regular cars are banned altogether.

All that said, the Cybercab is real, is it not??

Not in my town -- like I said, I was riding around with a guy trying out his 'FSD' not long ago, and as a dancing bear it was fantastic! Then we got to a street with some (partial) construction barricades, which it veered around (well done) into the left turn lane for oncoming traffic (umm, ok) pissing off the lady who was trying to turn left and get out of the clusterfuck quite a bit. (not great)

The car carried on (because we wanted to see what it would do!) until it got to the complete barricade and big trench in the road -- which it did not try to drive through (nice!) but did turn 180 degrees to the road and shuttled back and forth for a while until we got tired of it and turned around manually.

I was not scared at any point, and would feel like a total cad if I complained that the bear was a bad dancer -- but if I were paying for a taxi ride (and/or a trembling chickenshit millenial) I think I'd be a bit pissed -- I assume that the cybercabs are (similarly to Alphabet) relying on some heavy mapping and maybe remote operators on-call to avoid such situations? IDK but I can't imagine we wouldn't have heard about something in the 5% by now if they didn't have some kind of backup plan. (speaking of which, where's Uber gone in this space? I was under impression that their whole entire play relied on beating everyone to FSD -- if Elon has beaten them to something like it with his crazy ideas, I'm not sure what that says?)

Elon said that he would do it -- I don't think he has...

Weird because I'm at 95% self-driving

See Pareto...

Elon's timeline on FSD was (IIRC and the last time he said much about it; it shifted a few times and he's kind of shut up lately) SF-NY without touching the wheel by 2020; he still hasn't done it AFAIK?

I also have experience with a self-driving Tesla -- it did pretty great until getting boxed in by unmapped roadwork.

Has any AI skeptic ever said "I was fundamentally wrong"?

Does FSD count as AI? 'Cos I've been pretty dam skeptical about that since about 2012, both IRL and online, and I'm at the point where I'm willing to say -- I was fundamentally right? I didn't say it was impossible, I said it was oversold and probably going to be pretty niche for a long time -- which is pretty well how it's panning out. Maybe AI will be similar? I don't believe that we've hit Pareto stuff with it yet, not even close.

I don’t want to sound like a leftist but this basically never happens.

I don't see how it could not -- I did an (adult) coach's clinic awhile back where we did some scrimmaging -- most of the other dudes were beer-dads like me, but there was also a pretty serious woman -- quite a bit younger, high level rec player and probably the most talent there. When she tried to go around me to the net, I had to dial it down a lot -- if I played even rec-level full bore defense, she would have been injured. This was not the case with the other (old, out-of-shape) dudes -- we did not hold back on each other; it's not fun that way. I wouldn't count on all teenage boys to have this kind of restraint.

So that's a tidy few billion there for recovering anything regardless of what it's made of.

Still, it doesn't quite sound, ah, profitable?

I could see us going to A.C. at some point in the next thousand years too -- moderately low probability but it could happen. But I have a hard time seeing us doing it more than once-ish, and setting up a colony seems right out barring extreme (read: unlikely) circumstances on Earth. If your oppressed minority can build a starship powered by atomic bombs, I'd think you might be more inclined to set yourselves up somewhere on Earth and tell your oppressors to fuck off?

And if there's no colonization, the Fermi Paradox still isn't really a paradox? There could be an awful lot of advanced civilizations in the galaxy without them ever bumping into each other if all they can do is go to the next star system and poke around a bit.

Nah, I'm not that degenerate -- I'm not saying that the market thinks Huang is irrelevant (although he probably is) -- but the anti-datacentre sentiment last week is also significant. As to why AMD and Broadcom investors seem fine with it, IDK -- Broadcom has a pretty different business model I think? AMD... I guess they released some good Q3 numbers? Dunno, that's why I'm not an active trader lol.

It was obviously his vision that ultimately brought the company back from the dead, but it's no sure thing that the company would have survived if Jobs had stayed.

Probably worth noting too that Jobs' interim hardware venture didn't exactly set the world on fire, either -- and apple continued to struggle in the desktop space when he got back... pretty much to this day? iPod/Phone kind of saved the company, and one could argue that this was the same kind of wave-surfing that nVidia did with machine learning frameworks for GPUs. "Right place, right time, and didn't fuck it up" is great, but it doesn't make you indispensable.

Check out the 5 day performance of NVDA (-9%) vs Broadcom (up 25%) and AMD (up 44%)

Jensen resigned today -- refresh my memory, did anthing else at all newsworthy happen last week?

I think the space-junk shielding aspect of Orion is a pretty big question mark though? But even if it worked, what's in Alpha Centauri that will make me rich in 20 years when I get back with it? The motivations that drove the age of sail just don't exist.

Endoliths and sponges are not sentient -- sponges are barely even animals! I find it somewhat plausible that we might find a way to live as long as the sharks, but wouldn't be at all surprised if that were a hard limit. (and even less surprised if there were some even shorter limit for big-brained animals in general)

hard for me to believe that for every species, it's impossible to use technology to substantially extend lifespans. And even assuming that, it's extremely hard for me to believe that it's infeasible for all species to reach a nearby star system simply using generation ships.

It's not hard to believe at all if you reason from observation instead of from science fiction -- we don't know of any sentient species that lives more than a few centuries; maybe there's a reason? As I said, generation ships may be a possibility, but you need to also think about the probability of somebody wanting to do this -- if it's low enough, it doesn't really change things much unless intelligent life is really quite common in the galaxy.

Well there is an existence proof for self-replicating machines with human level intelligence -- humans ourselves.

Of course -- but if one were building a machine capable of self replication, deliberately building it to do so in an uncontrolled manner seems extremely unwise -- as you note, these machines tend to be pretty dangerous to those around them!

Not at all -- the difference is negligible because they are making use of resources which are just sitting around in the systems they visit. If they can self-replicate and travel well enough to explore the galaxy, it would be trivially easy to leave a few in each visited system.

It depends how easy it is to build these machines, doesn't it? If it's trivial, I think we are wandering back into the realm of science fiction; after all, so far as we know it's not actually possible to build a self replicating machine that can do so over the span of millions of years and survive travelling through interstellar space at some reasonable speed. It sounds kind of hard, if you think about it?

Sure -- those could be other people doing the same thing, or people already in the market trading up and taking on a manageable mortgage.

Boomers might be changing this, but typically I think a great many house purchases come from people just moving to roughly equivalent places for work, better schools, location preference, etc. Locational arbitrage has been a big thing since even before covid -- if your place in the city is worth $1M+, buying an equivalent (or better) place elsewhere for half the money and pocketing the difference gets pretty attractive, and if anything drives up prices in the other places (low demand), while being lost in the noise of the urban real estate markets. (ie. not many people do this relative to the total housing stock in a big city, but it's still enough to distort the much smaller rural markets)

Can you be more specific as to why?

Space is really big and the lifespan of every known creature is extremely short as compared to the timescales inherent to interstellar travel -- science fiction is fiction. Same goes for extracting resources -- Columbus thought he was going to China, which was known to have lots of resources that could feasibly be moved back home and sold at a profit. If he thought it would take 1000 years to get to China, he would not have gone.

Strong disagree. Leaving an unmistakable message would cost essentially nothing. For example, you could just leave an arbitrary number of probes in the system, have them watch for signs of intelligent life, and if they see them, physically visit any such beings and deliver a message face to face. This would cost essentially nothing.

I depends on your probes, doesn't it? Complex mechanical devices tend to have a pretty limited lifespan too -- if we hypothesize for the moment that self-replicating machines are not also science fiction, it still costs a lot more to have your probes hanging around replicating in every system they visit for hundreds of millions of years than not.

For one thing I guess it means that there's a pool of buyers (current homeowners) who are not sensitive to interest rate or other affordability concerns when looking at houses -- so demand does not (necessarily) exceed supply?

(3) is probably the main source of disagreement -- I think interstellar colonization attempts are highly improbable due to the physics of the thing. (as mentioned, with possible edge cases where a highly advanced civilization has literally no alternative but to flee their home system -- seems improbable too though?)

Given ¬3, (1) no longer makes sense either -- why would you "lay claim" to something that you can never visit?

Messages... yeah maybe. I guess you need to include the odds of us finding the messages? Messages that are literally impossible to miss (on a millions of years timescale) would also be resource intensive to produce, and some black obelisks on Mars might very well exist! (or else your probes are designed to shoot some proto-DNA onto habitable planets and we are the message, of course)

the argument is "sending probes for conquest/destruction is normal, somebody might do it", and your counter-argument is "nobody would ever do it"

No, my counter-argument is "nobody CAN do it", in the case of conquest -- space is too big for conquest to be meaningfully possible. You can posit some scenario (generation ship fleeing a dying sun, say) in which it might make some sense -- but then one can counter-posit that this too sounds pretty rare, and just add a few more (small) terms to the Drake equation.

Similarly with destruction -- I can see how it might be possible for some civilization to send probes around that blow up and/or harvest all the matter from every system that they encounter, but my argument is that there's no obvious motivation for this behaviour, so it would probably be pretty rare too.

One can reject the Dark Forest hypothesis on the basis that interstellar conquest is largely impossible, therefore nobody has any reason to hide or become omnicidal -- and one can further note that "all the matter/energy in a star system" is way more than any self-replicating Von Neumann exploration setup would have reason to harvest.

Which leaves you with adding terms for "probability that any given intelligent lifeform wants to destroy the galaxy for no reason" and maybe "probability that any given lifeform is incredibly sloppy and wasteful to the point of accidentally destroying the galaxy" to your modified Drake equation -- I really do think these terms can be small to the point of negligibility, although of course they interact with your guesses at the other ones. If you think life is extremely common in the universe you might make it work -- but IMO it would need to be common enough that we'd know of it happening more than once for the chance to be worth considering.

Wake me up when Elon finds some Martian dinosaur skeletons or something.

I think maybe you misunderstood my point. I'm saying that exploring, conquering, and extracting resources is pretty normal. That's what most species do if they can.

I understand your point perfectly -- my point is that if only the first of those is something that a species can do vis a vis space exploration, which certainly seems to be the case, then they will not do the other things.

If your Von Neumann probes are star-hopping around the galaxy, they really only need enough resources from the solar system to hop on to the next few stars in your path -- where they will do it again. That doesn't seem like a lot of resources, and burning more resources than you need is wasteful, which is also something that most species are evolved not to be.

Yeah, transhumanism is dumb -- I think aliens who are roughly subject to the same constraints and desires as humans are very much more likely than otherwise; you can put a separate 1E-20 in your Drake equation if you want trans, uh, -alien aliens.