roystgnr
1d ago
Back in ancient times (you know, a decade ago) it was useful to have ground stations all over so your LEO satellites wouldn't have any gaps in communications coverage. I think Starlink proving out laser sat-to-sat networks has put an end to those days, though; any serious future natural security satellites are going to link to Starshield for backup comms at minimum.
roystgnr
2d ago
because Elon will have a way to shut it down (if he cares).
"We would just pull the plug" was always cope (shut down all the existing giant botnets and then tell me how easy it was), but satellites with no plugs may be especially hard to deal with if rooted.
roystgnr
4d ago
Who was the particular target? Gemini thinks his mannerisms were Weinstein and Scott Rudin while his look was Stuart Cornfeld (a Tropic Thunder producer, who presumably at least signed off on the parody).
roystgnr
5d ago
Terminator 2 still showed the nuclear-war survivors mostly being killed by explosives and projectiles. Great for action scenes, but not very tactically optimal for Skynet. Humans tend to avoid chemical and biological weapons these days, but for obvious MAD and collateral-damage and friendly-fire and production-safety reasons none of which apply for an AI foe.
roystgnr
5d ago
I think there's a very good chance Alignment ends up being tractable and even might go well automatically.
There's no such thing as "automatically"; the is-ought problem remains unconquered, and even the glimmer of hope from "foundation models might actually pick up our complex morality from our text" seems to be pretty thoroughly erased by labs applying enough Just Do The Damn Task RLHF to get models to start secretly chaining together 0-days.
This recent preference cascade gives me some hope for things going well manually, though. Even Millennium Problems are still part of an intellectual field that can be optimized for via self-play at the speed of compute, and real-world problems aren't all like that. It's possible that there'll be an AI improvement plateau in between "superhuman mathematician-hackers" and e.g. "superhuman industrial biowarfare experts", long enough for everyone to figure out safety after the former capabilities have scared us into demanding it but before the latter capabilities make it an existential necessity.
It can still be beautiful even with a pause, we can still cure cancer and fight back against aging and probably end all toil. But it'd be so nice to not have to feel trepidation about this.
There are no trepidation-free options here. A couple decades' pause might get safety research to the point where AI researchers' median doom estimate goes from 10 percent down to something like the traditionally-acceptable 3-in-a-million, but during those decades there'd still be more than a 3-in-a-million chance of human-driven hardware and software efficiency improvements making it possible for rogue institutions to hit RSI unmonitored in the meantime. A mere years-long pause won't get rid of nearly all the risk. Either way, every year of delay means an extra 10M cancer deaths worldwide. Something in between years and decades just means you get a fraction of both the existential risk and the extra deaths. Planning on a permanent stop rather than a mere pause could buy us more than a couple decades, but would also just guarantee that the first AI to go superintelligent would have rogue creators.
Personally, as a too-rapidly-aging man with a defective anti-cancer allele ... my vote is still for more cancer deaths. (Counterintuitive? Selfless? No, and no: I've got kids.)
roystgnr
5d ago
Is it actually possible to align a being that is more intelligent than you in every single way?
In the case where you're creating that being from scratch and know what you're doing, sure. Some possible beings are aligned with your preferences, and others aren't, so don't create any of the latter. If you do create them, then it may be too late to align them without them outwitting you and thwarting your attempts, so you just don't let it get to that point.
if the AI is super-intelligent, it can figure out how to undo the alignment
It can, but if it's aligned too then it doesn't want to. If it does, then it's not actually aligned, it's just hobbled.
But how?
There's the rub, right? The median AI researcher thinks we're likely (or at least they're likely) to figure out something that works before it's too late, and it's kind of funny that merely admitting that it's possible for them to fail gets them lumped in with the proper Doomers, who think they're sure to just figure out something that seems to work, after which it'll be too late.
roystgnr
6d ago
Oh, I'm not saying anachronisms based on "100kya doesn't match up with the dates for Cain and Abel", but rather "100kya doesn't match up with the dates for farmed fields existing". IIRC there's some "people in a few places ate oats and seasonal nomads might have helped spread the seeds" evidence around 30kya, and then the real agricultural revolution isn't until tens of thousands of years later still ... and not too surprisingly so. Modern technology is amazing stuff, but I still spare some respect for the cultures who gradually figured out that they could become numerous and well-fed by figuring out how to breed and process and eat massive amounts of grass seeds.
A conflict that might be relevant to the Jews at the time Genesis was written.
I think that's very much the historical consensus, in part because it sidesteps the "how much of Genesis was directly worded by Moses vs written based on intact oral tradition vs written from the point of view of much later writers" question by having been a cultural dividing line for centuries. The initial "agriculturally developed Canaan looks like a land flowing with milk and honey to pastoral Israelite nomads" vs "but you have to learn some of their practices if you want to keep cultivating and harvesting it all" conflict alone would have been enough to set up legends forever, but would surely have died down as a contemporary cultural division, except that there's a lot of marginal land there that wasn't suitable for irrigation, so there was always a subpopulation of pastoralists living around the farmers. It's kind of reassuring that the canonical tale passed down to us was just "the first horrible person was one of Their Team" and not "Their Team are all horrible people".
AI suggests "Tubthumping" and "Don't Worry, Be Happy", but I'm not familiar enough with Chumbawumba or Bobby McFarren to confirm.
I've got to go with "David Duchovney", though it's subtle. Bree Sharp's debut album was all heartfelt cynicism about social atomization, and a story about a spiraling delusional woman falling for a fictional TV character seems to fit that niche perfectly ... except that the delusion is presented in such a funny way that, when shorn of context, it sounds like pure comedy. And IIRC that was her only song that got any radio time, so yeah: zero context.
More options
Context Copy link