domain:anarchonomicon.substack.com
Amadan
Enjoying my short-lived victory
1d ago
Your contributions on AI are always interesting and worth reading (not that I agree with them, but I enjoy reading them). But as much as moderation here has been accused of running on the principle "Anything is okay as long as you use enough words," it did not escape me that you used a lot of words to basically say Jane you ignorant slut!. No, burying the insults (repeated) under a lot of words does not make it okay to be this belligerent. And on a topic that should not require this much emotional investment. Your lack of chill is a you problem, but your lack of civility is a Motte problem. You do not win the argument by plastering as much condescension and disdain as you can between links.
BreakerofHorsesandMen
Sweet Sejenus
1d ago
ENTJ
I just did the Meyers-Briggs a few months ago at the suggestion of some people I work with. Unknown to me at the time, the office consensus had settled on ENT already and was only undecided on J or P.
Once I got my results, J seemed pretty obvious to me, but I guess I hide it well or come off differently to people in real life.
WhiningCoil
Ghost of Quokka's Future
1d ago
It might have been the friend I was doing it with, and how the trails were rated in our area. It was 15+ years ago, so I'll probably get all the details wrong. But there was some sort of rating system that didn't seem dissimilar from rock climbing ratings, and he was really into getting to the next difficulty, and mastering X, Y and Z skills necessary for doing so, and upgrading his bike with fancy brakes and tires and shocks. Where as I just had some dinky street/trail hybrid bike with none of those things and found myself completely incapable of keeping up. I just enjoyed doing the same trail or two when I could.
urquan
Hold! What you are doing to us is wrong! Why do you do this thing?
1d ago
do you like to categorize people because it's another interesting data point about how they work
This is basically the thing that got me interested in psychometrics. My problem is I have a bad tendency to categorize people in my head as lab rats with identifiable characteristics and try to predict what they’re going to do. It can make it a little hard to actually connect, because I’ve already formed an impression of what box someone fits it, and my box is oddly specific.
That said, what you said earlier about the most interesting part of Meyers-Briggs being the type functions is also my view. I think the categorizations are bogus, but I’m definitely the sort of person whose most conscious experience is thinking through things like I’m making a logical argument, pulling in information from the environment to try and enhance that logic, and then dragging my feelings along.
Except when anxiety hits and the processor gets interrupted by the amygdala. That’s when things go off the rails.
Gaashk
1d ago
Sure, it's good advice. It's just good advice in almost all contexts, hardly anyone gets enough exercise nowadays. It's worse for the purposes of differentiating various personalities.
ToaKraka
Dislikes you
1d ago
Despite having already signed a contract to have a custom house built, I remain obsessed with designing houses. On Monday, I managed to get my coworkers to participate in a house-design contest by offering prizes for the winners (a one-ounce silver coin for first place, half-ounce for second, and quarter-ounce for third). Is anybody interested in having a contest on this website as well?
self_made_human
Grippy socks, grippy box
1d ago
Call me crazy, but I think bring advised to get more exercise is quasi-universally applicable, and beats the advice given by the average horoscope. Just because something is universally applicable doesn't mean it's worse!
rae
1d ago
It comes across as a bitter nasty commentariat incredulous that someone would dare to have a different opinion from you.
I don't think the issue is OP's opinion. The issue I had was listing off credentials before making completely incorrect technical explanations, doubling down instead of refusing to admit they made a mistake, and judging researchers based on the fact that they don't hold any US or EU patents.
Ioper
1d ago
you know full well you'll never enter the top 10% in most activities
Yes I will.
Top 10% is nothing though. Even top 1% is nothing.
Practically nothing I do recreationally will mean anything to anyone outside of my immediate family, regardless of how good I am.
Thinking that people would care if you were a bit better at some skill is autistic delusion. No-one gives a shit.
anon_
1d ago
Of course. But it goes to who is at fault -- the siege ends when Hamas surrenders. That Hamas has constructed itself to make that impossible to surrender doesn't change the fact that the lack of surrender is the but-for cause that perpetuates the siege.
International law can't make anyone do anything -- but it does assign normative responsibility based on the practices of nations. Doubly so when there the construction that prevents the resolution of the conflict based on that practice is itself against that practice.
DaseindustriesLtd
late version of a small language model
1d ago
I really haven't entered a pissing contest (typo).
I find OP's text exceptionally bad precisely because it is designed as a high-quality contribution but lacks the content of one; what is true is not germane to the argument and what little is germane is not true, its substance is mere sneer, ideas about reactivity and perceptivity are not thought through (would we we consider humans modulo long term memory formation unintelligent?), the section on hallucinations is borderline incoherent. This is LLM-like in the worst sense possible. I've said many times that superficial adherence to the letter of rules of polite discussion while ignoring its spirit is unacceptable for me. Thus I deem it proper to name the substantial violations. If mods feel otherwise they should finally give me a time out or a block. I am not a very active participant and don't intend to rely on any residual clout.
Multiple people in this post were able to disagree with OP without resorting to prosaic insults in their first sentence.
Multiple people should be more motivated to call out time-wasting obfuscated bullshit before wasting their time. I am grateful to @rae for doing the tedious work of object-level refutation, but the problem is that the whole dismantled section on word2vec math is not relevant to OP's argument about lack of reactivity (which isn't supported by, well, anything), so OP doesn't feel like it is anything more than a nitpick, a pedantic challenge to his domain-specific technical competence. Why should anyone bother with doing more of that? Let's just get to the meat of the issue. The meat is: are LLMs intelligent? I've shown that rigorous, good faith objections to that have a poor track record.
At the risk of getting into it with you again. What did you think of this when it made its rounds 2 months ago: https://ml-site.cdn-apple.com/papers/the-illusion-of-thinking.pdf
I think I've already responded to that but maybe not. The meta issue with Apple papers is that their DL team is coping about repeated failures to build a competitive system (it may be that such philosophical handicaps get in the way). The object level issue with their tests is covered in this series of posts on X. One relevant piece:
- If you actually look at the output of the models you will see that they don't even reason about the problem if it gets too large:
"Due to the large number of moves, I'll explain the solution approach rather than listing all 32,767 moves individually"
- At least for Sonnet it doesn't try to reason through the problem once it's above ~7 disks. It will state what the problem and the algorithm to solve it and then output its solution without even thinking about individual steps.
Does this mean “0% accuracy”? I guess for people who believe “LLMs create billions of value by doing stuff like autonomously optimizing CUDA kernels, agriculture creates value by growing wheat, ergo wheat is as intelligent as an SWE? heh” is a clever dunk, it does.
There is a massive gulf in efficiency of understanding between people who approach LLMs with some rigid preconceived notions and people who can fucking look at the outputs and think about them. The gulf is so large that the former group can go through the motions of "empirical research" and publish papers proving how LLMs inherently can't do X or Y and not notice that they can, in their own setup, moreover that the setup is nonsensical. It's no longer a matter of polite disagreement, it's pure refusal to think, hiding your head in the sand. It's on par with paranormal research and homeopathy and should be treated as such: pushed out of the field and into self-funded fringe journals to die in obscurity.
3472094371
1d ago
There has been a discussion of Death Note lately, and I feel obliged to shill my favorite fanfiction of it: Silent Partner, Unfinished Business, which is the best thriller novel I've ever read, including original literary works. And a few days back one forum participant was disgusted by the canon's treatment of Naomi's death — here she survives Light's sadistic execution, and... well, that would be spoilers. But to that forum participant (I honestly don't remember the name): it could be to your liking, and wouldn't even require knowledge of the original series past the episode with Naomi.
anon_
1d ago
Well, we can dunk on them and they richly deserve to be dunked on.
But also it's equivocation again. Are we talking about a utopia -- on how we should organize society on ideal terms. Or are we talking about how one should live within a real society in its real terms.
anon_
1d ago
Even that's not getting it. Warriors dream of slaying the enemy. 95% of being a warrior is marching for days on end in shitty (figuratively) boots, living in a shitty (literally) camp and dying of cholera.
Most of the army isn't even fighting at all! The tooth to tail ratio is, at best, 1:3. So odds are that you're not even a soldier, you're just hauling supplies or dealing with logistics or guarding the rear/base. All while still in your trenchfoot boots and without a sewage system. They don't even get the glory of saying "I risked my life for the poets".
Maybe one angle of what I'm getting at is that any functioning system (whether it's an army or a commune or a technological liberal/capitalist society) requires a large amount of tedious, unglamorous and unrewarding work. When people imagine a system other than their own, they either forget that or imagine that someone else will do it.
It's like because they are transposing reality by imagining another system, they also transpose reality by ignoring or eliding this fact about all systems.
Rov_Scam
1d ago
As an avid mountain biker, I'm curious as to what you think was gamified about the whole experience. Most people who get into the sport start riding relatively easy trails and progress to harder ones as they get better, but the whole concept of difficulty is vague and not necessarily related to how fun a trail is to ride. What most people don't do is start off by taking lessons and sticking with it to "unlock" various achievements by passing certain thresholds. Easy trails can still be a blast for experienced riders, and a beginner can always walk anything he's uncomfortable with (most difficult trails are only truly difficult for relatively brief stretches). Most people, though, will be good enough in a year that they'll be able to ride whatever they want to, within reason, and the only thing that differentiates riders is speed, which isn't important if you aren't racing and which no one cares about on casual rides. Skills improvement usually just means getting faster by being able to navigate tricky sections better, like having the technique to navigate tight turns without slowing down too much or being able to find lines in rock gardens. The end result of developing these skills is that you end up finding certain kinds of trails more enjoyable, but it's a completely personal gain.
oats_son
1d ago
If kids were tattoos, sure! Instead, parents are generally viewed well by everyone except antinatalists, and responsibly having kids in a healthy marriage has got to be a good indicator of lack of criminality or impulsiveness. They also do a lot more than sit on you making a statement. People become parents by accident or they become parents for reasons other than signaling something. This provides pretty good cover for the people who are signaling something.
Getting more than one costly, expensive, loud, and permanent signal directly on your body, on the other hand, seems to correlate a lot with low impulse control and many other problems. Most of my arguments would not apply (as much) to a small, invisible tattoo, or a singular large one. I still would dislike them, though.
The_Nybbler
In the game of roller derby, women aren't just the opposing team; they're the ball.
1d ago
International law can require command authority all it wants; it can't make it actually exist. Hamas's leadership cannot surrender and retain command authority.
P-Necromancer
1d ago
Yeah, a real organization with rigid, non-democratic decision making processes can avoid this dynamic, at least so long as those processes hold. Japan's surrender in WWII is instructive, here: There was a cabal of officers who tried to prevent the surrender, but discipline held and they were rebuffed. The difference with amorphous groups is that there's just no one who can do the rebuffing; 'leaders' last only so long as the rest of the movement deigns to listen to them.
DradisPing
1d ago
Don't know your familiarity with coding, but Java isn't a good place to start.
Ok so coding is the part of this project I know, I'm just not super familiar with all of the Java tooling. Unfortunately Java is a necessity for Minecraft mods. The MCP interface stuff is in Python.
Getting a jar file with dependencies into a working Docker container was more of an adventure than I expected.
Do you mean Claude Code ?
Nope, Claude desktop is a thing with MCP support: https://claude.ai/download
Using an IDE with an open project to do things unrelated to that project seemed peculiar so I didn't go down that path. I tried out using the MCP in VSCode and it worked well, although the agent did try to re-write my MCP code when it couldn't initially connect. The downside to Claude desktop was that it didn't let me try other llms.
I wasn't aware of the VSCode support, so that's neat.
anon_
1d ago
First, on a normative level, this is why international law requires command authority. An armed force must be “under a command responsible for the conduct of its subordinates”. I'm sure I"m preaching to the choir here, but "We have constituted our armed forces to be incapable of X, therefore it isn't our fault if we don't do X" is a game theoretic self-own. It's invites the very conduct we seek to proscribe.
Second, I'm not sure that's right. Hamas is more than capable of butchering domestic opposition. They did it to the PLO, they can impose what they want at the tip of a bayonet and be obeyed. Perhaps though.
the_Culture_is_great
1d ago
they facilitate coexistence between different groups with divergent aesthetic values
Beyond just different groups, even just between individuals. Most people have a few preferences that are weird or non-conforming, even if they're otherwise very similar. It's just nicer to not sweat the little things in general, and I'd argue the vast majority of tattoos are in that category.
Nah, I agree with the others below: If you need to gamify something to enjoy it, then you don't actually enjoy it. It's like people who get gym memberships on January 2 with the goal of trying to lose that stubborn 20 pounds and finally "get into shape". But the goal is more important to them than the exercise, which they find sucks, and they have to force themself to get to the gym and quit by March. the fit people who go to the gym aren't there because they have exceptional self-discipline; they're there because they like going to the gym. It's not something they have to force themselves to do; it's something they look forward to doing. I'm an avid cyclist, and I regularly go on long rides on the weekend. But I'm not putting in 60 miles because I need to tick some box that says I have to do 60 miles today and maybe I get some kind of reward for doing it. I ride the 60 miles because that's the length that corresponds to the amount of time I want to spend riding. And if I get sick of it and turn back early I don't care, because I'm not trying to force myself to do anything, or unlocking any achievement.
I feel that this is a problem of box tickers and speed-runners in general, and especially in the outdoor scene. About a decade ago I was hiking on the Appalachian Trail in Massachusetts when I came across a through-hiker eating lunch at the saddle between two mountains. I told him I was surprised that he was so far north about a month before most hikers would get that far. He excitedly told me that there were people who had finished already. I continued up the mountain and was enjoying the panoramic view at the top when he passed me. He plowed forward without even looking at the scenery. What's the point of doing a hike like that if you aren't even going to stop at the summits? It was clear that he was eating at the saddle so he could carb load before the climb and make better time.
Years later I was hiking Mt. Harvard in Colorado when I came across a guy from Kansas City who was trying to hit all of the fourteeners in the state. We hiked together for a while until he decided that I wasn't moving fast enough for him, but he did talk about how his wife was very supportive of his mission. I never would consider a hobby something that required suport from my family unless it was some kind of obsession that kept me away, which it appeared to be for him. When we got to the top we ran across two guys who were hiking together. From the summit the trail continues across a ridge to another fourteener, Mt. Princeton. It was a clear, warm day, and while the trail looped back around to the trail we hike in on, it looked like a long, hot, sunburned, high-altitude slog. The guy from KC and one of the guys decided to do it, while me and the other guy hiked down to the parking lot together. The thing about it, though, was that the guy from KC was staying with a friend in Denver who was getting him into a show at Red Rocks. If he had hiked straight out to his car from Harvard it would have been about the average time you'd leave to get back to Denver and change before heading to the concert. The guy acted like he had to get back to the car by five if he wanted to make it and thought it was possible, but he was effectively skipping the show. And since there was no cell service there, he was leaving his friend high and dry. Skipping an activity to do something else is one thing, but the guy seemed so concerned about bagging an extra peak that he was willing to risk pissing of a friend who gave him free passes to a band he really liked.
Gaashk
1d ago
I may have underrepresented how much I tried getting into it, though it's been most of a decade. I bought and read a book (not sure which), had coffee with a neighbor who was a certified counselor and used it in her work, who also lent me a book, and put probably about 20 hours into it, with no results, just confusion. Meanwhile, MBTI people say things like "use your second function more," which is much more actionable.
GBRK
1d ago
So I just ate an automated 3-day reddit ban for saying we should bomb the tigrayan militants responsible for their genocidal strategy of raping and genitally mutilating women. I can't really complain about that: I was knowingly in violation of reddit's "no advocating violence" policy. I have been before, and I will be again, probably until I get permabanned, because sometimes violence is the solution. Thomas Aquinas will back me up there.
But what's interesting to me is the "automated" part. Now, I've faced my fair share of human disciplinary action before. Sometimes it's fair, sometimes its not. But either way, the humans involved are trying to advance some particular ideological goal. Maybe they blew up because Ii contradicted their policies. Maybe they developed a nearly autoimmune response to any kind of NSFW post becauseof prior calamities. (Looking at you, spacebattles mods.) Maybe they genuinely wanted to keep the local standard of discussion high. But reddit's automated system is clearly not designed for any of that. Rather, its most fundamental goal seems to be the impartial and universal enforcement of reddit's site-wide rules to the best of its capability.
I agree with yudkowsky on the point that an "aligned" AI should do what we tell it to do, not what is in some arbitrary sense "right." So I'm also not going to complain about how "cold and unfeeling AI can't understand justice." That would be missing the the forest for the trees. It's not that AI aren't capable of justice, it's that the reddit admins didn't want a just AI. They wanted, and made, a rule-following AI. And since humans created the rules, by their impartial enforcement we can understand what their underlying motivations actually are. That being, ensuring that reddit discussions are as anodyne and helpful as possible.
Well, really it's "make as much money as possible." But while AI are increasingly good at tactics-- at short tasks-- they're still very lacking at strategy. So reddit admins had to come up with the strategy of making anodyne discussions, which AI's could then implement tactically.
The obvious question is: "why?" To which the obvious response is, "advertisers." And that would be a pretty good guess, historically. Much of reddit's (and tumblr's, and facebook's, and pre-musk twitter's) policy changes have been as a result of advertisers. But for once, I think it's wrong. Reddit drama is at a low enough ebb that avoiding controversy doesn't seem like it should be much of a factor, and this simultaneously comes as a time where sites like X, bluesky, and TikTok are trying to energize audiences by tacitly encouraging more controversy and fighting.
Which brings me to my hypothesis: that reddit is trying to enhance its appeal for training AI.
Everyone knows that google (and bing, and duckduckgo, and yahoo) have turned to shit. But reddit has retained a reputation for being a place to find a wide variety of useful, helpful, text-based content. That makes it a very appealing corpus on which to train AI-- and they realized that ages ago, which lead to them doing stuff like creating a paid API. This automated moderation style isn't necessarily the best for user retention, or getting money through advertisement, but it serves to pre-clean the data companies can feed to AI. It's sort of an inverse RLHF. RLHF is humans trying to change what response strategies LLMs take by making tactical choices to encourage specific behaviors. Reddit moderation, meanwhile is encouraging humans to come up with strategic adaptations to the tactical enforcement of inoffensive, helpful content. And remember what I said about humans still being better at strategy? This should pay out massive dividends in how useful reddit is as training data.
Coda:
As my bans escalate, I'm probably going to be pushed off reddit. That's probably for the best; my addiction to that site has wasted way too much of my life. But given the stuff I enjoy about reddit specifically-- the discussions on wide-ranging topics-- I don't see replacing reddit with X, or TikTok, or even (exclusively) the motte. Instead, I'm probably going to have to give in and start joining discord voicechats. And that makes me a little sad. Discord is fine, but I can't help but notice that I'm going dow the same path that so many repressed 3rd worlders do and resorting to discussion on unsearcheable, ungovernable silos. For all the sins of social media, it really does-- or at least did serve as a modern public square. And I'll miss the idea (if not necessarily the reality) that the debates I participated in could be found, and heard, by a truly public audience.
I'm a bit surprised by how visceral the anti-tattoo sentiment is in the comments. I wonder if it's a product of the average age being higher than I thought (maybe early to mid 40s?) and less interaction with the working class, specifically people younger than them.
I agree that in the past tattoos could be a valid barometer for anti-social traits, but nowadays they're so ubiquitous among young people that I don't think there's a strong correlation at all. The nature of tattoos is different as well. My impression is that they tend to be smaller than they were in the past and thus people, especially women, are more likely to have more than one. Interestingly, I've noticed that a lot of younger police officers have tattoos, usually in the forearm area. I tend to view police officers fairly positively, and so I don't have nearly as negative of an association.
More options
Context Copy link