@WhiningCoil's banner p

WhiningCoil

Ghost of Quokka's Future

6 followers   follows 0 users  
joined 2022 September 04 23:24:47 UTC

				

User ID: 269

WhiningCoil

Ghost of Quokka's Future

6 followers   follows 0 users   joined 2022 September 04 23:24:47 UTC

					

No bio...


					

User ID: 269

Probably the lingering effects of America being a frontier nation once upon a time. Where self reliance is a virtue.

It's fading. That is clear as day. The learned helplessness I see all around me is depressing. People acting like any sort of manual labor or recreational competence is beneath them, and the only worthy use of their time is byzantine credentialism scams. Why be the chump making things when you are supposed to be ordering them around?

Rarely do I see any of these people living lives as lofty as their view of their place in the world. But that's a topic for another day, and thread.

Speaking for myself, there is something mythical about woodworking. My household had items that grandfathers had made, my wife's household had items her grandfathers had made, I myself feel a deep calling to leave my daughter with some things I've made for her eventual household. It's like a totem of generational competence. Some proof that "We are people who make things". And it's a lot less ephemeral than the code I sling as a career.

Yeah, if it were me, I'd have used some sort of solvent to remove the paint or lacquer. My wife actually used this orange based paint stripping stuff to take the paint off all the countertops in the kitchen. Worked amazing, and it took zero effort with a paint scrapper to remove the resulting goo.

So, there are a few areas where sanding will go wonky on you. The first is you have shitty sandpaper that wears out too quickly. I recommend sticking with 3M sandpaper. The second is that the sandpaper clogs up. All those palm sanders have a slight suction effect, but attaching them to a dust extractor will have you humming along effortlessly for hours if need be. And with the specific type you linked, that has more of a vibration mechanism than a scraping mechanism, I find I have to move them just to get the dust out from under them. I use them as a last resort when they are the only tool that fits into the space.

And if you are removing a lot of paint, I'd recommend a scraper first, as I've found paint gums up sandpaper even faster than normal.

But I'm a relative layman.

I have to ask, what sort of sander were you using, with what grit sandpaper? Did you have any sort of dust extraction?

Cause I can totally see you sanding and sanding away with a palm sander, using 400 grit sandpaper, with nothing but the bag it comes with.

I can see this. But funnily enough, I didn't get into woodworking for it's own sake. I got into doing what carpentry repairs seemed at my skill level for the house I bought. Because finding contractors willing to do small repairs is borderline impossible because reasons, and the ones that are willing to slum it doing handyman tasks instead of flipping houses want an arm and a leg. So primarily I invested in a table saw for those tasks. Saving $60 on a random frame is a fringe benefit of having built a skill set to save hundreds, if not thousands of dollars on relatively basic home repair tasks.

Lord of the Maps ever so vaguely invokes what I thought was a staple of nearly every home that was established in the 80's that I ever visited or lived in. An artistically taken photograph of the house when it was first built. Always with the color subtly faded, either by time or by process. And I remember in my profound summer boredom staring for extended periods of time at the one in my house, noticing how much trees grew, or which pushes had been taken out, etc, etc.

Counting my time? I lost several hundred dollars.

You know it was funny. After I made it, and it was so damned cheap and easy, I did a quick google of "Are custom picture frames a rip off". Almost universally the results were "Absolutely not, framing is super difficult and technical, and takes a very trained eye." Random posts on Reddit by professional framers justifying how the average person should not, under any circumstances, attempt to build their own frames.

The internet can be a strange place.

Ok, this reply finally moved the needle for me, and I'll shift my position from "LLMs are a neat statistical trick" to "Maybe LLMs use language to perform some form of 'thinking' in ways not entirely dissimilar from how language facilitates our own thinking."

Hrrrm. Maybe half an hour one Sunday doing the initial milling, then another afternoon a different Sunday to finish the project.

Work was kicking my ass the last month. But that finally ended, so I got out to my workshop to finally frame a fun map I got.

The Algorithm thought I would be interest in Lord of the Maps, and I was. Purchased a nice map of my state around Father's Day. But I skimped on the frame. $70 for a frame? Fuck that, I can probably build one at least as good, right?

Took me a month to get around to it, but I eventually did. I didn't do anything too fancy. Milled down the red oak cause I had plenty on hand, I let it get out it's movement in my shop for about two weeks, but it actually didn't move any more. So I cut it to final dimensions. Put a slight 15 degree chamfer on the inside edge because I saw someone else do it and I liked it. Then I worked it with some aged barrel stain. I did also do some splines on the corners to reinforce the relatively weak glued miter joints.

A few things happened. My miter gauge must have been a fraction of a degree off, because the miter joints in aggregate opened ever so slightly towards the outside after I glued the whole thing up. So I guess I'll be recalibrating that. I calibrated it when I got it, but I suppose I didn't do a good enough job. The second thing that happened was the stain was slightly blotchy in a few places where I may have left it on too long. A lot of cleaned up well when I hit it with some mineral spirits, but some of it didn't. Oh well. Next time. In retrospect, I'm not sure I'm pleased with how the splines turned out either, and I might use small dowels next time.

The cost breakdown is insane compared to a store bought frame. I used probably $8 worth of red oak, the 16x20 glass pane was also $8. I used about $2 worth of hardboard for the backing, and then it used pennies worth of stain, glue and brad nails. So about $18 all told, versus $70 ish.

The goal is the pluck out your lying eyes. There can be no feedback in the system to cast any doubt what so ever on whoever the regime picks to rule next.

I don't think it's plausible that all the examples of GPT-4 doing that sort of thing are faked, not when anyone shelling out the $20 can try it themselves. And people use it for things like programming, you can't do that without reasoning, just a less familiar form of reasoning than the example I gave.

My problem is, while I'm sure that not all the examples of GPT-4 seeming to get complex reasoning tasks are fake, if they cannot be replicated, what good are they? If GPT-4's ability to "reason" is ephemeral and seemingly random, is it really reasoning, or is it just occasionally getting lucky at ordering abstract tokens for it's monkey overlords?

You know, it's funny, I went through the linked whitepaper. Skimmed mostly. It made few positive, objective claims about GPT4's ability to reason. It mostly said it could reason "better" than previous iterations, and had been trained on a dataset to encourage mathematical reasoning. Notably they say:

It can sometimes make simple reasoning errors which do not seem to comport with competence across so many domains

I saw some the prompts where they asked GPT-4 to explain it's reasoning, and was underwhelmed. They were extremely rudimentary mathematical tasks of the 5th grade word problem sort, and it's purporting "reasoning" could have easily been imitating training. When I saw that, I took a closer look at how it performed in assorted test, and saw it comprehensively failed the AP English Language and Composition and AP English Language and Literature tests. Which makes sense to me, because a lot of those tests involve more generalized and flexible reasoning than the sorts of formalized mathematical logic examples it might plausibly be trained to imitate.

You don't need to do statistical tests of its calibration to see that, because something that was truly just doing statistical prediction of text without having developed reasoning or a world-model to help with that task wouldn't be able to do even the most basic reasoning like this unless is already appeared in the text it was trained on.

I opened up Bing Chat, powered by GPT4, and I tried that example. I got "The diamond is still inside the thimble inside the coffee cup on the kitchen counter". In fact, I've yet to see a single example of an LLM's supposed ability to reason replicated outside of a screenshot.

I mean, I can agree people are stupid. Hell, I'm probably stupid in a lot of ways too! My wife reminds me of it every time we are in a social setting and I alienate her friends.

Even so, LLMs lack most of the faculties that allow humans to get closer to truth. They have zero interaction with base reality. They take Socrates allegory about the cave, and turn it into literally how they experience the world, through a training dataset. And, as I keep mentioning, their "cognition", such as it is, isn't even based on the statistical correlations of things being true, but of what words come after other words. Without even knowing what any of those words mean! It's all just abstract tokens to them.

Imagine this were all being done is some sort of unrealistically massive mechanical or analog computer! Would you still consider it thinking?

I earnest disagree. If you check the GPT-4 white paper, the original base model clearly had a sense of internal calibration, and while that was mostly beaten out of it through RLHF, it's not entirely gone.

They have a genuine understanding of truth, or at least how likely something is to be true. If it didn't, then I don't know how on Earth it could answer several of the more knotty questions I've asked it.

It is not guaranteed to make truthful responses, but in my experience it makes errors because it simply can't do better, not because it exists in a perfectly agnostic state.

I think you are flatly wrong about this. I've tried to find literally anything to back up what you are saying, and come up with zilch. Instead, I wound up with this.

https://www.scribbr.com/ai-tools/is-chatgpt-trustworthy/

A good way to think about it is that when you ask ChatGPT to tell you about confirmation bias, it doesn’t think “What do I know about confirmation bias?” but rather “What do statements about confirmation bias normally look like?” Its answers are based more on patterns than on facts, and it usually can’t cite a source for a specific piece of information.

This is because the model doesn’t really “know” things—it just produces text based on the patterns it was trained on. It never deliberately lies, but it doesn’t have a clear understanding of what’s true and what’s false. In this case, because of the strangeness of the question, it doesn’t quite grasp what it’s being asked and ends up contradicting itself.

https://www.scoutcorpsllc.com/blog/2023/6/7/on-llms-thought-and-the-concept-of-truth

Thus far, we’re really just talking about sentence construction. LLMs don’t have a concept of these as “facts” that they map into language, but for examples like these - it doesn’t necessarily matter. They’re able to get these right most of the time - after all, what exactly are “inferences” and “context clues” but statistical likelihoods of what words would come next in a sequence?

The fact that there is no internal model of these facts, though, explains why they’re so easily tripped up by just a little bit of irrelevant context.

https://fia.umd.edu/comment-llms-truth-and-consistency-they-dont-have-any-idea/

They have zero idea what's true. They only know the probabilities of words in text. That's NOT the same thing as "knowing" something--it's a bit like knowing that "lion" is the most likely word following "king of the jungle..." without having any idea about monarchies, metaphor, or what a king really is all about.

The folks at Oxford Semantic Technologies wrote an interesting blog post about LLMs and finding verifiable facts. They call the fundamental problem the "Snow White Problem." The key idea is that LLMs don't really know what's true--they just know what's likely.

I keep meaning to dick around with some LLM software to see for myself how some of the nuts and bolts work. Because my layman's understanding is that they are literally just a statistical model. An extremely sophisticated statistical model, but a statistical model none the less. They are trained through a black box process to guess pretty damned well about what words come after other words. Which is why there is so much "hallucinated information" in LLM responses. They have no concept of reason or truth. They are literally p-zombies. They are a million monkeys on a million typewriters.

In a lot of ways they are like a con man or a gold digger. They've been trained to tell people whatever they want to hear. Their true worth probably isn't in doing anything actually productive, but in performing psyops and social engineering on an unsuspecting populace. I mean right now the FBI has to invest significant manpower into entrapping some lonely autistic teenager in his mom's basement into "supporting ISIS". Imagine a world where they spin up 100,000 instances of an LLM do scour Facebook, Twitter, Discord, Reddit, etc for lonely autistic teens to talk into terrorism.

Imagine a world where we find out about it. Where a judge forces the FBI to disclose than an LLM talked their suspect into bombing the local mall. How far off do you think it is? I'm guessing within 5 years.

I've long though ChatGPT would replace offshore work. Or at least executives would give it a try. They'd probably still discover that everything it ever does is wrong, just like offshore workers. But they'll give it the old college try for a decade or two.

And the poor bastard who has to oversee the quality of work it produces will be driven just as mad by ChatGPT pretending it understands the request and returning nonsense, as he was by the offshore workers who did the same. If anything it will be worse, since ChatGPT will return results instantly, and the Indians take a week or two. He will have zero reprieve from the stupidity. Just complete idiocy, in his face, all the time, that he must fruitlessly attempt to wring productivity out of.

Are Indians actually smart? I honestly don’t know. We get higher caste Indians here and heavily filtered from a giant population. If they are then one can probably make an assumption India could replace China. I have some doubts here because India has always been the “next big thing” for perhaps even centuries and it never happens. Sort of feels like a better Africa. They were never the leader or a great power of civilization like China has often been.

I've had... bizarre experience with Indian's professionally. Every Indian coder, support staff, offshore worker, project manager, etc I've ever worked with over 20 years, had this peculiar mode of communications failure. It's hard to even describe. On my end it looked like some part of their cognition was just broken. Like they'd make all the right mouth sounds that they understood certain questions, comments and directives, and then proceed to perform tasks or give answers that were totally detached from any context we were operating under. It's how I imagine p-zombies would behave. It actually reminds me of trying to get ChatGPT to understand my point, and it just failing to over and over again. Because ChatGPT doesn't think, or really comprehend anything at all. It's just a statistical model of what words come after what other words.

Like, one time I was having a conversation with the Indian project manager we were subcontracted to. There were three requirements on the project, that when taken all together, only a subset of 2 could be satisfied. And trying to have this conversation with the PM just went round and round in circles. I explained that I implement requirement 1 and 2, then requirement 3 cannot be satisfied. They'd come back to just do 3. I'd return that if I implement requirement 3, requirement 1 cannot be satisfied. They'd come back and insist I implement requirement 1. I'd explain if I implemented requirement 1 and 3, requirement 2 could not be satisfied. We went round and round and round like this, with me giving detailed technical and logical reasons why these requirements in aggregate created a paradox, or a double bind. After two weeks of beating my head against a wall, they handed me off to a non-Indian and we had the issue sorted in 5 minutes as he immediately understood what I was pointing out.

This person was a project manager.

I've had far worse experiences with lower level Indians I've had to work with. Simple instructions like "DO NOT RESTART THE SERVER" just get flatly ignored routinely. Like, I tell the guy, to his face, looking directly into his lazy eye not to do it. He proceeds to do exactly that 5 minutes later.

So, I just donno. I hate to generalize. Generally I'm not a fan of "different forms of intelligence" discussions. But my every professional interaction with an Indian, at all levels of an organization, have been frustrated by them seeming to just not ever grok anything I tried to explain to them. Conversational knowledge just refused to penetrate their skull in any meaningful way. I simply cannot account for it. I've had people tell me it has to do with their culture, and never ever under any circumstances admitting they don't understand, or don't know something. I guess I can't rule it out.

The male autist who spends all day discussing deep lore on MawInstallation is not consooming a lot of product doing so.

They used to. Who else do you think dozens of source books for a Star Wars RPG appeals to? Or dozens, if not hundreds, of books that take the canon seriously? If Disney cared at all about giving those people material they could meaningfully engage with, they'd lap it up like the paypigs they are. They certainly used to, and are more or less the only reason Star Wars had any cultural relevance 20 years after it first came out.

Once upon a time, I would have said it requires meaningfully engaging with the core themes of the work. The Zen-like sayings of Yoda which are some of our only glimpses into Jedi philosophy. The broad conflict between the light side and the dark side, and all the things those sides represent. With the light side representing love, friendship, individualism, courage and heroism, and the dark side representing fear, anger, hate, conformity and cowardice.

Star Wars wasn't queer. Star Wars wasn't commie or capitalist. Star Wars wasn't feminist and/or against the patriarchy. I get that people point to certain lines or characters and try to spin it at "Star Wars was really X" all along. But no. They were not core themes of Star Wars. You can see anything in anything if you are willing to squint hard enough at 6 hours of footage.

I do wonder sometimes what made Star Wars nerdy. In the sense that when I was a kid in the 90's, being a Star Wars fan was weird. I wasn't alive for it, but it was a massive nationwide phenomena when it came out. It's not like only nerds saw it. But I guess it was the nerds that stuck around for 20 years, watching the original Trilogy over and over again. Playing the West End Games RPG because they just didn't want to let go of the world. The EU didn't really start proper until the 90's. Return of the Jedi came out in 83! None of the Video Games were that good until Lucas Arts was finally allowed to make some, and that was in the 90's as well. Being a Star Wars fan in the late 80's and early 90's must have been profoundly lonely and boring.

I mean, Christ, I was hoping to find some documentation, but look at this downright quaint blog talking about Star Wars Celebration in 1999. Or this threadbare article about a 10 year anniversary convention for Star Wars in 1987.

Anyways, the point I was winding to, is that Star Wars has no core themes anymore. There is no longer anything there for a true fan to intellectually engage with. So the teen girl diddling herself to 80's Luke and modern Kylo Ren dicking down may as well be just as much a fan as anyone else geeking out over laser swords, cool looking space ships, or pausing the VHS at their favorite frame of the Twilek in the slave dress dancing and rubbing one out.

The rather obvious problem for the LGBT community and the rest of us is that we cannot even point out the bootleggers without being labeled. No matter how nicely you point out the connection between letting small children make sexual decisions (or that the adults are pushing, often covertly for sexual discussions and books without parents consent) the answer is you are a horrible bigot for even thinking like that.

Not only can we not point it out, but they can't point it out either. I've seen numerous L's and G's trying to point out the bootleggers, only to find themselves kicked out, because it turns out the bootleggers are the core of their coalition. Turns out after you get all the rights you always wanted, the people who stick around and don't go back to living their normal lives like all the slogans claimed, are bootleggers.

I wouldn't count "Anakin thinks it's a great idea to spend the day at an amusement park, also having something else up his sleeve. Confessing his love to Obi-Wan!" as being genuinely a fan of the series.

They are now. Which is why I'm not.

Back in my day, Star Wars fans played the West End Games version of the Star Wars RPG, the miniatures game, and fantasized over the source books. The latest and greatest Star Wars video games were X-Wing and Tie Fighter, which were super nerdy and demanding flight sims. Dark Forces had just come out too, and was super heavily teen boy coded. It was a 90's FPS after all.

That was the fandom.

Funnily enough, all the source material written for West End Games Star Wars RPG was sent to Zahn, and he used it to bootstrap his Thrawn trilogy.

So yeah, there was no "slash fiction". A lot of the creativity of the young fans I interacted with as a young fan myself was spent coming up with RPG ideas, sharing Star Wars total conversion WADs for Doom, using the source books to try to game out if a Wampa could defeat an AT-ST walker, etc. Or, since nobody had read all the EU fiction because there were no online book stores, or ebooks, or even a wikipedia of what all the EU books even were, we'd hear tales of this or that awesome droid or alien from the random jumble of EU books our parents had haphazardly found on a shelf for one of us.

I’ll summarize the link but you really should read it for yourself, it’s astounding how bad Disney gets the Star Wars franchise: for instance, the books need to be (in addition to all the old touchstones of diversity etc) ANTI WAR. And then there’s the characters, who are somehow all androgynous.

You know, the idiocy of Disney's "The Force is Female" push doesn't take a genius to figure out. I was talking to my wife about it, and I just asked:

"When we were kids, how many little boys did you know who liked Star Wars?"

"Tons."

"Did you know a single girl who liked Star Wars?"

"No."

It's the absolute height of fart huffing retardation to literally jettison your entire built in audience for an IP, and try to replace it with an audience that never cared about it one iota. Like I get the typical neoliberal line of "They're just trying to expand their market". Except now the experiment has been run, and they've wound up appealing to less people than before.

It was literally inconceivable to me, if you'd asked me 10 years ago, that I could stop loving Star Wars. I dragged my poor not-yet-wife to a midnight screen of The Force Awakens, well past her bed time. And she went with me because she knew how much it meant to me, god bless her.

And yet here I am. I fucking hate Star Wars now.

The ways they've seemingly purposely spit in the eye of every fan of the original Trilogy and EU is insane. Between turning all the legacy characters into beaten down pieces of shit, killing them off ignominiously and denying fans the reunions they desperately wanted. Putting aside all the glaring technical issues with the craftsmanship of the films, like poor writing, poor characterization, completely nonsensical plot contortions, etc. I forget where I heard it, but someone joked that all new Lucas Films are made for an audience of one, Kathleen Kennedy. Every film has to be about a stand in for how Kathleen Kennedy views herself, replacing a character that is a stand in for how Kathleen Kennedy views George Lucas. Once I heard that, suddenly it all made sense.

God I hope Disney sells it all back to Lucas at half the price.