domain:science.org
EverythingIsFine
Well, is eventually fine
1d ago
I'm struggling to imagine how on earth you could possibly think that this is a distinction without a difference. There's a massive difference!
Especially if you then turn around to use this conflation to imply other things. Maybe it's all well and good as its own end (as you point out if it kills us all it won't matter why), but you yourself advanced the following argument:
These aren't agents that were explicitly trained to be self-preserving. They weren't taught that shutdown was bad. They just developed shutdown resistance as an instrumental goal for completing their assigned tasks.
This suggests something like goal-directedness emerging from systems we thought were "just" predicting the next token. It suggests the line between "oracle" and "agent" might be blurrier than we thought.
To me that sounds like circular or tautological reasoning, you can't call it a distinction without a difference and then use only one of the two as evidence for something else. If shutdown-resistance is emergent, it has implications for agency and intelligence. You used this theory as an example of goal-seeking behavior. But, if shutdown-resistance is a normal training pattern outcome (as I believe), then we can't draw any conclusions from it.
I think the best way to put this in perspective is the philosophical debate over whether the intelligence of humans is just a means to an end (evolution opened up a niche for intelligent beings) that's ultimately purely mechanical (aided in survival and reproduction), or if it's something more special (self-awareness is a thing, life isn't deterministic, meaning is a new concept unique to humans or intelligent life) and emergent in the sense that it created something more than the sum of its parts. Religiously, I believe the second, but you don't have to be religious to feel that way. Note that the implications for how we treat AI differ greatly depending on which bucket you put in it. Similarly, although on a personal level maybe there's little reason to see a difference between the two opinions about human intelligence (you are you either way), it's hardly a distinction without a difference overall. It's a big deal!
aardvark2
1d ago
but then what's the use for link to reddit profile? It doesn't let me to view user post history, only says "This account has been suspended"
aardvark2
1d ago
For the other 98% of black people, just ask to see their degree.
USA has affirmative action so that blacks college degree ownership is not based on IQ gap at all but on what those in power want; so https://jbhe.com/2022/03/the-racial-gap-in-educational-attainment-in-the-united-states-5/ There were 7,921,000 African Americans over the age of 25 in the United States who had earned at least a bachelor's degree
I would agree however that it helps more to rich blacks and not poor ones.
Primaprimaprima
...something all admit only "TRUMP", and the Trump Administration, can do.
1d ago
My apologies, I didn't know how much you had already tried! The whole "therapists and soft safe closets" thing made you sound like the permissive type, but if you're not, then fair enough.
The design space of possible minds is very large. I suppose there are some people who would just die without drugs; and perhaps they did, for most of history. That's a bit sad though.
Do what you have to do to live a normal (and physically safe) life obviously. Although I do think you should listen to your intuition that "it doesn't sit right with you". At the very least, don't let anyone talk you into thinking that it should sit right with you. You can at least have that much.
fmac
Ask me about bike lanes
1d ago
Defensive is a hilarious word to use in this context given were rapidly approaching year 100 of this conflict (or 1000, depending on how far back you want to play the grievance game).
I don't think it would be very difficult to come up with an argument that the Palestinians are on the defensive due to $PREVIOUS_ATTROCITY.
Also quite funny to call it defensive when there's a 1000:1 military power ratio and one side has killed 6x the people of the other.
If the credible threat against you has been defeated in the field for almost two straight calendar years, are you really on the defensive?
OracleOutlook
Fiat justitia ruat caelum
1d ago
Thank you for your detailed and thoughtful reply!
FarmReadyElephants
1d ago
What makes discord and reddit similar is that there is a discussion of enthusiasts available on any topic you are interested in. If I want to learn Ableton or discuss the Byzantine empire, I know I can find that on reddit. It just comes with BLM and LGBT propaganda and ban-happy leftist mods.
OracleOutlook
Fiat justitia ruat caelum
1d ago
I have four kids. All my other kids so far are perfectly normal and well-behaved. We are not permissive parents. We are not an "everything goes" family.
I physically pick up my daughter and carry her to the car. We leave when she starts freaking out and she misses out on a lot of stuff. We left the fourth of July party early and she missed fireworks. She misses a lot. We don't go places usually because I just have to pick her up and leave. Leave the library before we check out books. Leave the grocery store with a shopping cart half full. Leave the park. Leave leave leave. That has been my life. Babysitters have quit. I can't go anywhere with her and I can't go anywhere without her and I can't go anywhere.
And she's getting heavier and heavier. And when I pick her up she fights with everything she has. She is STRONG, crazy strong, frightened animal strong, and it's getting more and more difficult. If I don't figure this out soon, I will NO LONGER be able to carry her safely to her closet to calm down. And then if she's attacking her siblings, I have to attack her? If I can't carry her safely, it's just getting rougher and rougher to her.
And yes, I take away her toys. Yes, she is consigned to her closet often. We are stricter than most people we know. Our kids know they need to say please and thank you or they aren't getting fed. You really have the wrong idea if you think we just don't try to change her behavior at all through normal parenting means.
We even tried spanking for a few months when she was four. She kept doing this one behavior where she would get water out of the bathroom sink, fill containers, and then leave them places. The water would spill and make a mess and we were worried about rot. So we had a rule - when we saw her do this, she would be spanked immediately. The consequence would be immediate, and it was only for this one specific behavior. Well... nothing changed. Nothing at all. Except we felt like jerks, because it really seemed like, if she could stop herself, she would have. She didn't like getting spanked.
The word "Pathological" means that it prevents you from living normal life. I think our experience with her qualifies.
That's why it's so weird that I can actually go places with her taking L-Theanine.
She ran straight into a moving car. This isn't something a normal seven year old does. She is going to die if I just treat her like my other three kids.
Are you sure that when someone demands too much of you, you have an adrenaline rush? Start attacking people? Run away like a lion's after you? Freeze like a gun is pointed at your heart? Several times a day? That is what PDA is supposedly. And my kid acts like it.
And the fact that I can give her a supplement that completely changes her behavior, so she becomes perfectly behaved, when if I give the same supplement to another kid it doesn't change anything at all... doesn't that hint at something?
TheDemonRazgriz
1d ago
I would argue that “unalived” already looks quite stupid.
EverythingIsFine
Well, is eventually fine
1d ago
I personally don't think it's a good measure for intelligence, but I will actually try and defend the "it's intelligence if it makes enough money in the proper contexts" argument. It's not saying on the immediate level that monetary effects show intelligence, that's obviously silly, it's got one or two more links in the logical chain. It probably goes something like this:
-
people pay money for things that matter to them, that is to say, money is a good proxy for value
-
people pay money for [knowledge] work, that requires intelligence
-
if enough people pay money for [knowledge] work, to AI...
-
therefore, we can infer that people, in aggregate, and based on reliable links, have judged that AI has enough intelligence to count as intelligent
I disagree with it because I think thinking about intelligence is inherently a philosophical-only type of question, but I certainly respect the opinion above anyways because it does have a degree of sense to it. While jobs and tasks also notably are affected by supply and demand, I do personally think that money is an excellent proxy for value in the vast majority of cases. People are whining about museums closing? Evidently the museums aren't valuable enough. STEM degrees pay more than humanities degrees? Their degrees are more valuable to society. Conclusions like that.
Bullet points 2 and 3 are I think the points of confusion here? Advocates of this definition emphasize that if there is a 'big enough' amount of money involved, the judgement has effectively already been made by the wisdom of the masses + the laws of capitalism. In terms of what number counts, this is nebulous as anyone will admit, but serious people have put numbers to it, in fact as I referenced in my original comment, the Microsoft-OpenAI contract itself uses this definition for "AGI" and puts a 100 billion number as the cutoff. So I don't think we can dismiss the argument out of hand!
I should also note that this argument strongly implies (bullet 3) that replacement of humans (humans specifically because nothing else is 'intelligent' enough to do the task) for certain types of work is required. Bullet 2 is another sticking point: is help-desk support, for example, something that actually requires intelligence? Humans are replaced by machines for purely physical tasks already and that doesn't result in claims of intelligence. Still, you can see the appeal of the argument if something previously thought impossible for a machine due to its perceived complexity and adaptability is suddenly possible on a large scale. However, it's important to distinguish between picking apart the underpinning details of this logical chain, from the overall claim, they are different. A dispute about bullets 2 or 3 is somewhat a factual dispute, or a definitional one, and doesn't invalidate the overall claim necessarily.
SkoomaDentist
1d ago
You have to be able to find the drone and guide the loiterer close enough fast enough.
FeepingCreature
1d ago
I think the synthesis here is that we should have enough knowledge that if we were to build an ASI, and turned it on, it would in fact do what we tell it to, interpreted in the way that we mean it, and that this is table stakes for getting any sort of good outcome. - That is, our problem at the moment is not so much that we don't know what the good is as that we can't reliably make the AI do anything even if we want it very much and it is in fact good.
Lazuli
1d ago
Dangerous professional voice is a superpower.
Can you elaborate? Is this simply bombarding bureaucratic drones with requests for specific documentation so you can create a paper trail for yourself on their inability to process something up the chain?
I actually was required to take Gallup's Clifton Strengths test for my business communications college class (at a discounted cost) which I actually found pretty good and a step above the other tests I've taken.
For example, one of my top traits was "Context". Basically, that I enjoy thinking about the past, like to think about cause and effect, etc. This is helpful for I think some obvious reasons, but also a weakness, because change can be tough, and it can sometimes slow me from looking for current opportunities. I think that's actually pretty spot-on for me, and at the same time it's not true of all people (a lot of people find the past boring) so it suffers a little less than some of the other tests from the generic-advice trap common to astrology and horoscopes. Most of the traits highlighted has some kind of pros and cons list, with the idea being to better understand yourself and to double down on what you're good at (and be aware of the blind spots for what you aren't good at).
Or, "Harmony" was another top trait. It can be helpful for sensing conflict beforehand, finding common ground, staying practical, etc. but also means that I might sometimes avoid conflict, seek too-easy band-aid fixes, or get stressed when people don't agree.
Now, I will say that it's oriented towards corporate-like utility, rather than some kind of 'accuracy', and any system of personality with cleanly separated domains with suspiciously similar numbers of sub-categories is a little suspect, but I also kind of like that aspect of it. Also it's identifying the top "strengths", but really it's just saying these traits are your strongest traits, somewhat divorced from if they are good/bad or on some kind of sliding scale. In that sense it's a bit more honest because it's not so much about "you are X category" but more "this blend of traits represents you best".
Stellula
1d ago
I don’t fantasize about being a warrior or a poet. I like democracy, and I think I live in the most peaceful, best time in all of human history. I think both of these fantasies are stupid.
Primaprimaprima
...something all admit only "TRUMP", and the Trump Administration, can do.
1d ago
idk, when I was super little and I would start acting up in public my dad would physically pick me up, carry me to the car, and say "we are never taking you anywhere ever again if you're gonna act like a brat". And I would usually shut up pretty fast after that. For in-house infractions they'd hide my toys or something until I calmed down. Seemed to work well enough. It's possible I was a more "mild" case though, because by the time I was in first grade I had already become a relatively docile teacher's pet.
Basically I'd throw out all the psychiatry shit and say "sink or swim kid, up to you". That's how people did it for, you know, all of human history up until the last ~50 years or so. You think they had L-Theanine 1,000 years ago? No they said pick up a fuckin' shovel kid or we're all gonna starve this winter.
I think we could all be diagnosed with a little PDA, yeah? I got PDA for days. I'm still a lazy piece of shit as an adult who doesn't like to do anything. The only thing that makes me actually acquiesce to the "demand" is a hard deadline (with consequences) and a swift kick in the ass. It never goes away, you just gotta learn to deal with it. People like me appreciate the kicks in the ass, trust me!
Primaprimaprima
...something all admit only "TRUMP", and the Trump Administration, can do.
1d ago
Yeah. You just gotta accept it and move on and make the best of what you got.
Honestly I think you probably could get it to work okay right now with current models. However, for something like this, you really need to have some above-average skills in prompting. You'd find it helpful to read something like Anthropic's prompting guide, although that one's specialized a bit more for Claude than OpenAI's stuff. Some of the advice is non-intuitive, and you might need some tweaking. For example, for Claude (has some unique preferences like wrapping sections in XML tags), they recommend something kind of like the following in terms of general structure, and yes, before you ask, order can matter. If you don't want to read through it, here's my abbreviated notes for a good prompt structure for something like this:
You are __. The Task is __ (simple one-sentence summary).
< context to consider first, including why the task is important or needs to be done this way. Yes, telling the AI "why" actually does improve model outputs in many cases >
< input (or input set) to take action on; at least for really long inputs, it should be near the beginning, short outputs this can go later >
< details on how to do it, guiding the thought process. This is where you'd put some version of your bullet points. Your layout seems reasonable but it's possible scaffolding or flowcharting a bit more explicitly, including perhaps what to consider, could help >
< explain how the output should be formatted, and the expected output (possibly repeat yourself here about the original goal) >
< optional: 3-5 diverse examples that help with interpretation of goals and reinforce style and formatting. Also optional is you could provide the thought process to reach those answers in each case, mirroring the logic already outlined >
< any final reminders or bookkeeping stuff >
Did you know that Anthropic actually have a whole tool for that process? If you follow the link, you can get a prompt generator (literally, use AI to help you tweak the prompt to find a better one), auto-generate test cases, etc. It's pretty neat. You can also somewhat mitigate confabulation here by adding a bullet point instruction to allow it to return "I don't know" or "too hard" for the more difficult cases. Also, it's possible that, depending on the level of tool use and thinking needed per bullet, that applying it to a giant music library would require some real money.
I will note that OpenAI's guide has some slightly different advice, but still pretty similar. The main difference is a lack of XML tags, and also, OpenAI recommends this structure:
< identity, style, high-level goals >
< detailed instructions >
< examples of possible inputs with desired outputs >
< context that might be helpful >
As you can tell, it's actually pretty similar overall. Yes, you have more control (as well as more complicated stuff to manage) when doing it programmatically via the API, but I think you could probably try via the normal chat interface with decent results. I should also note that if the AI doesn't need to use very much "judgement", you might actually do better with a well-prompted 'normal' model instead of a simulated-reasoning model.
coffee_enjoyer
☕️
1d ago
They’re using fiber optics drones now which is insane
coffee_enjoyer
☕️
1d ago
But how far away do you need to be to shoot out buckshot or something? Or netting? Drone kill zone is maybe 20 meters.
coffee_enjoyer
☕️
1d ago
It does not appear that either side has figured out a way to hide from enemy drones. If you are in the settlement that they are attacking, you will eventually just die.
SerialStateLineXer
1d ago
Given that they didn't remove any of those comments, that seems unlikely. You probably just got autojannied.
EverythingIsFine
Well, is eventually fine
1d ago
Vinland Saga: Suggest finishing, partly because the anime does something I've rarely ever seen: after the first season is a (fun) orgy of violence and revenge and action, the second season is the opposite: character development, slow plot, and a message about how violence is bad.
My own suggestions/mini review list, loosely sorted in order of appeal to non-anime types, or people who have only watched one or two. Of course, it's always subjective. I don't quite agree that "anime is just a medium" because its highly-controlled production pipeline and limited set of studios creates some definite commonalities, but it's true there's a wide variety of genres.
Violet Evergarden, 9/10
THIS is really an excellent first or early anime. A woman used essentially as a special-ops child soldier is now a little older, and while the war continues, she decides to take up an unusual vocation: a typist in an era where few people know how to write (I guess), she also assists in helping people organize their thoughts to write letters. Often, these letters are emotionally charged, or offer some major catharsis; thus the show's episodes are organized roughly with a major letter per episode. Parallel to this, we should mention that the main character, the eponymous Violet Evergarden, has lost both of her arms, replaced with mechanical ones, which mirrors her emotional state, still dull and robotic from her war experiences. So we slowly get to see her open up over the course of the series. Sad and emotional at times, hopeful in others, this one is highly memorable and at times honestly you often forget it's an anime at all. Finished, a season and a movie or two.
Apothecary Diaries, 9/10
This show is great. A nice mix of mystery, cool setting, and like the previous, much fewer anime tropes than your usual fare, this one stands out. A fairly level-headed girl but with a strange obsession with poisons, raised as an apothecary by her adopted father (read: herbal-medicine doctor for the poor, in this case often a brothel) is kidnapped into a loosely-Chinese imperial palace as a servant there. And not the cool, plot kidnapping version either, she's literally just nabbed off the street and sold and has to come to terms with her new life. Which she does, and she's pretty smart and a good investigator even though it really isn't her interest, and she gets pulled into harem politics to some extent as first a food taster, and then other adventures especially for a powerful eunuch within the palace. Ongoing story with two seasons, but with some good closure.
Frieren: At Journey's End, 10/10
Now, I'm not sure whether this score, which reflects my anime of the decade designation, translates to the general public, but it's very enjoyable. A fantasy series that explores the idea of what a long-lived elf's life is actually like! Lord of the Rings plays a bit with this idea in a way, but doesn't fully commit and it's spun differently. There, the elves are kind of tired of life, but here, we ask the question: what might Legolas be feeling, going on an adventure with some others, when he knows that they are going to die and leave him behind again? LotR dodges this a bit by both killing much of the cast, and Gimli is also of a similar long life, but Frieren tackles this a bit more explicitly. She once went on a save-the-world trip, but as the mage of the party. Living for at least a thousand years however, she doesn't fully appreciate the impact this trip had on her, and experiences regret for not emotionally engaging more after her friends pass away. She uses this as impetus to start another journey back north again to the demon lands, retracing the save-the-world steps with a new group of people who grow on her. The world-building is great, the storytelling is on point, the vibes are excellent, it's just a great watch. Ongoing, one season completed.
Kaguya-sama: Love Is War, 9/10
This one earns a rare distinction for carrying with it a strong piece of advice: try the first episode or two subbed and dubbed. A humor-first series, this one takes place in a super-elite high school, where the two highest-performing students (one old money, and one a scholarship student), on the student council together, are trying to get each other to admit a crush on the other. They play all sorts of mental gymnastics to make this work. The humor largely comes from the commentary/narrator, but the sub and dub both approach it differently (and the dub actually localizes many of the jokes, so they are funny but in a different way). The sub leans a bit more dry-humor, irony-focused, while the dub plays up the conflicts as being outrageous. At any rate, this one is just good fun and although the series starts out as a bit more like a series of connected skits, it eventually transitions a little more into a proper show with character arcs and plot and all that good stuff. Finished, three seasons, epilogue movie to come.
I don't really know how the below actually stack up but I felt like tossing them in too.
Angel Beats!, 8.5/10
Admittedly it has been a while since I watched this one, but it's good. Some nice emotional catharsis, but I don't know how much I can say without spoiling things too much. A guy wakes up all of the sudden in a sort of alternate-reality school, with a confusing 'war' between a group of kids within the school and the school student body president, who is a bit of a robot, with the drone-like other students as bystanders. Despite the presence of guns, this is a low-violence affair where the war is mostly a series of, well, pranks more or less? Despite the sort of confusing set-up, you get some good character moments, and this one is a tear-jerker at times.
Dandadan, 8.5/10
Visual flair. Panache. Dialing stuff up to 11. This anime is now in its second season and is ridiculous but fun. You probably only need to see a few minutes to get an idea about what this one's about, but for text purposes the classic hook is that a high schooler who believes in ghosts teams up with one who believes in aliens, and they're both right! He has his balls stolen by a spirit, and aliens try to kidnap her, and then they have some adventures trying to resolve that.
Code Geass, 8.5/10
This is kind of like the Ender's Game of anime in a way? The main character lives as a privileged elite in a dystopian Japan ruled by a world monarchy-autocracy, but decides to join a local Japanese rebellion. He's very much a 5D chess type of guy, takes on an alter-ego, and did I mention there's mechs for some reason? Most of the show is him outsmarting people, because in a similar kind of "hook" to the oft-recommended Death Note (which I personally don't like), he has the ability to brainwash-command anyone to do anything... but only once, ever, in their life. Which he obviously wants to keep a secret, but has to also be smart about using due to its one-time-use nature. Two seasons.
But those degrees are disproportionately in psychology or communications.
More options
Context Copy link