Stubsack: weekly thread for sneers not worth an entire post, week ending 5th January 2025
Stubsack: weekly thread for sneers not worth an entire post, week ending 5th January 2025
Need to let loose a primal scream without collecting footnotes first? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid: Welcome to the Stubsack, your first port of call for learning fresh Awful you’ll near-instantly regret.
Any awful.systems sub may be subsneered in this subthread, techtakes or no.
If your sneer seems higher quality than you thought, feel free to cut’n’paste it into its own post — there’s no quota for posting and the bar really isn’t that high.
The post Xitter web has spawned soo many “esoteric” right wing freaks, but there’s no appropriate sneer-space for them. I’m talking redscare-ish, reality challenged “culture critics” who write about everything but understand nothing. I’m talking about reply-guys who make the same 6 tweets about the same 3 subjects. They’re inescapable at this point, yet I don’t see them mocked (as much as they should be)
Like, there was one dude a while back who insisted that women couldn’t be surgeons because they didn’t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I can’t escape them, I would love to sneer at them.
(Semi-obligatory thanks to @dgerard for starting this, and happy new year in advance.)
An interesting thing came through the arXiv-o-tube this evening: "The Illusion-Illusion: Vision Language Models See Illusions Where There are None".
Illusions are entertaining, but they are also a useful diagnostic tool in cognitive science, philosophy, and neuroscience. A typical illusion shows a gap between how something "really is" and how something "appears to be", and this gap helps us understand the mental processing that lead to how something appears to be. Illusions are also useful for investigating artificial systems, and much research has examined whether computational models of perceptions fall prey to the same illusions as people. Here, I invert the standard use of perceptual illusions to examine basic processing errors in current vision language models. I present these models with illusory-illusions, neighbors of common illusions that should not elicit processing errors. These include such things as perfectly reasonable ducks, crooked lines that truly are crooked, circles that seem to have different sizes because they are, in fact, of different sizes, and so on. I show that many current vision language systems mistakenly see these illusion-illusions as illusions. I suggest that such failures are part of broader failures already discussed in the literature.
It's definitely linked in with the problem we have with LLMs where they detect the context surrounding a common puzzle rather than actually doing any logical analysis. In the image case I'd be very curious to see the control experiment where you ask "which of these two lines is bigger?" and then feed it a photograph of a dog rather than two lines of any length. I'm reminded of how it was (is?)easy to trick chatGPT into nonsensical solutions to any situation involving crossing a river because it pattern-matched to the chicken/fox/grain puzzle rather than considering the actual facts being presented.
Also now that I type it out I think there's a framing issue with that entire illusion since the question presumes that one of the two is bigger. But that's neither here nor there.
I think there’s a framing issue with that entire illusion since the question presumes that one of the two is bigger
I disagree, or rather I think that's actually a feature; "neither" is a perfectly reasonable answer to that question that a human being would give, and LLMs would be fucked by since they basically never go against the prompt.
Surprised this hasn't been mentioned yet: https://www.rollingstone.com/culture/culture-news/meta-ai-users-facebook-instagram-1235221430/
Facebook and Instagram to add AI users. I'm sure that's what everyone has been begging for...
Spam bots are good now!
I think it did come up a few weeks back, but it's indeed a hilarious mess. the engagement must flow!
In my dreams, it won't take long until all user interactions are AI driven and people paying for ad space in that shit realizes that, leading to an immediate crash of meta's finances.
hoping for a 2025 with solidarity, aid, and good opsec for everyone who needs it the most
Hopefully 2025 will be a nice normal year--
Cybertruck outside of Trump hotel explodes violently and no once can figure out if it was a bomb or just Cybertruck engineering
Huh. I guess it'll be another weird one.
(I know I know, low effort post, I'm sick in bed and bored)
Hey, at least there’s no way the Elon simps can spin that, right?
Never mind.
They are also spinning it into "the car is so great you cant do terrorism with it due to how strong it is", which considering the several vehicle terrorism acts recently seems very unwise.
Also 'it would be different for the bystanders' i think you can see on the explosion vid there were not that many bystanders (which makes terrorism a bit less likely) and still 7 people were hurt (and the driver died). Id wait a bit with drawing further conclusions.
Sure, you know what, let's go with that. While obviously I don't condone terrorism, I agree with Nic here that if you are going to do a car bombing, blowing up a Cybertruck is preferable to other cars. Because it contains the blast better or whatever.
Don't worry about the low effort post, even the writers of 2025 are phoning it in.
https://xcancel.com/altryne/status/1872090523420229780#m
The whole thread is terrible; controlling and borderline abusive behavior.
I feel personally attacked because I have a BELOVED dino plush that looks almost exactly like that one, only is, you know, a fucking plush toy not an eldritch horror. They took a perfectly fine toy and ruined it with a stupid chatbot, the girl did the smartest thing and just uses it as a normal plushy.
Also if you listen to the video at the end you can really easily figure out why kids don't like that toy, IT'S FUCKING ANNOYING. Kids don't want to deal with your bullshit and fortunately they don't yet know how to pretend to care.
Found a couple QRTs cooking the guy which caught my attention:
https://twitter.com/denimneverdies/status/1872364569743786286
https://twitter.com/TheWapplehouse/status/1873915404529406462
a reply from a mastodon thread about an instance of AI crankery:
Claude has a response for ya. "You're oversimplifying. While language models do use probabilistic token selection, reducing them to "fancy RNGs" is like calling a brain "just electrical signals." The learned probability distributions capture complex semantic relationships and patterns from human knowledge. That said, your skepticism about AI hype is fair - there are plenty of overinflated claims worth challenging." Not bad for a bucket of bolts 'rando number generator', eh?
maybe I’m late to this realization because it’s a very stupid thing to do, but a lot of the promptfondlers who come here regurgitating this exact marketing fluff and swearing they know exactly how LLMs work when they obviously don’t really are just asking the fucking LLMs, aren’t they?
Not bad for a bucket of bolts ‘rando number generator’, eh?
Because... because it generated plausibly looking sentence? Do... do you think the "just electrical signals" bit is clever or creative?
Here's an LLM performance test that I call the Elon Test: does the sentence plausibly look like it could've been said by Elon Musk? Yes? Then your thing is stupid and a failure.
That test doesn't totally work as Elon does often say fuck.
Right, well God says:
meditated exude faithful estimate nature message glittering indiana intelligences dedicate deception ruinous asleep sensitive plentiful thinks justification subjoinedst rapture wealthy frenzied release trusting apostles judge access disguising billows deliver range
Not bad for the almighty creator 'rando number generator', eh?
That first post. They are using llms to create quantum resistant crypto systems? Eyelid twitch
E: also, as I think cryptography is the only part of CS which really attracts cranks, this made me realize how much worse science crankery is going to get due to LLMs.
As self and khalid_salad said, there are certainly other branches of CS that attract cranks. I'm not much of a computer scientist myself but even I have seen some 🤔-ass claims about compilers, computational complexity, syntactic validity of the entire C programming language (?), and divine approval or lack thereof of particular operating systems and even the sorting algorithms used in their schedulers!
a non-zero amount of the time, yeah
also, that poster's profile, holy fuck. even just the About is a trip
Wow, how is every post somehow weird and offputting? And lol at 'im seeing evidence the voting public was HACKED! (emph mine)' a few moments later 'anybody know some big 5 webscrape API coders? I need them for evidence gathering'. The delightful pattern of crankery where there is a big sweeping new idea that nobody else has seen, plus no actual ability in a technical field.
Not sure where this came from, but it can't be all bad if it chaos-dunks on Yudkowsky like this. Was relayed to me via Ed Zitron's Discord, hopefully the Q isn't for Quillete or Qanon
Curtis
IQ:300, Special Move: Urbital Laser
Curtis Boldmug has defined the meta for years. A competitive staple that strongly influences even builds not running him. Special attack causes unavoidable psychic damage even if you resist its charm effect. Vulnerable to sunlight.
Balaji
IQ: 300, Special Move: Yes Country for Old Men
A support type character. Good for ramping grift mana, but can't carry a game on his own. His ultimate is overcosted and just sucks up the hypecoins he spent the entire game producing.
Ray
IQ: 300, Special Move: Black Hole Graviton
Mostly just receives support thanks to boomer nostalgia factor. Low but nonzero win rate in modern tournament meta. Highly viable in time machine formats.
Eliezer
IQ: 300, Special Move: Goffik the Hedgehog and the Enders of Game
Former newbie favorite, fairly accessible and flashy. The Yud has seen heavy nerfs in the past years and at medium to high levels, his stats plateau severely much like his special move's plot. Thiel synergy has also shifted towards Curtis mains leaving Yud in shambles. Still a fun archetype and enjoys popularity as a smurf build.
Jack
IQ: 300, Special Move: Snorting an entire ground up bitcoin
Rather run of the mill character whose effectiveness was rather limited for a long time. The Blue Sky archetype made him meta relevant for all of five minutes until he got reclaimed by the toxic playerbase built around the social media platform he originally started and the uber braingenius currently in charge of that company. Beard gives him +1 armor bonus which is fine I guess.
Peter
IQ: 300, Special Move: Pondering my Orb
The apex predator of SV capitalism. The Black Lotus of technofascist grifters. His character is rumored to be based on Count Dracula. Even most SV billionaires can't touch him in a 1v1 matchup. Truly classic S-tier thinky boi.
Beff
IQ: 300, Special Move: World's Most Divorced Man First Date Percent Speedrun
Likely intended as a joke character, a guy named Guillaume pretending to know how to pretend to be cool on the internet. His posts turned out to be so lethally cringeworthy he started an entire archetype of */acc brainos. Not quite on the power level of Peter or Curtis, but surprisingly influential for an obvious meme build. Extremely weak to heartbreak from women named Ruth.
Leopold
IQ: 300, Special Move: To The Moooooon
Honestly, I had never heard of this guy before today but the data doesn't lie. The dots do go up and to the right and he posts a lot of them. Extrapolating from current trends, he will single-handedly reach singularity by the end of Q3 of this year.
each of them needs a scale (logarithmic) showing how much adderall they take
I recognize everyone except Leopold. Increase my suffering by telling me who it is.
https://xcancel.com/leopoldasch Leopold Ashenbrenner, some chart maker and
substackblog haver with twitter account. swallows all openai marketing materials hook line and sinker, i had enough of abyss gazing duty today won't tell you more
this logo in corner is for something called overfit qs, they have instagram page and that image was posted there
Fellas, I was promised the first catastrophic AI event in 2024 by the chief doomers. There's only a few hours left to go, I'm thinking skynet is hiding inside the times square orb. Stay vigilant!
I'm sad to report that the catastrophic AI event already happened and it was this picture
mind horrors beyond your comprehension
Ow god it is 2025 in .nl, it is coming! Everything is exploding, ai is turning us into fireworks! Yud was right!!1!!one!!
Comment sections on awful.systems are similar to this Drew Gooden sketch sometimes:
It's just hard for me to give MY input when I don't even know what's going on
If you stick around and do a bunch of research you will end up better informed and much unhappier.
Once a month or so Awful Systems casually mentions a racist in some sub-sub-culture who I had never heard about before and then I get to spend an hour doing background research on obscure net drama from 2013 or whatever.
I’m making a mental note to keep that link around for the next time someone barges into one of our threads and does the “I don’t know what this is, here’s my reaction to what I thought the topic was, no I didn’t read the article or lurk” routine
as a bonus they might accidentally watch the rest of the video and finally figure out how much AI sucks
“I don’t know what this is, here’s my reaction to what I thought the topic was, no I didn’t read the article or lurk”
bizarre that they actually just say this
You know guys, it's really hard for me to give MY input when you are so negative about all the terrible things I like. Next time you guys come CRAWLING to me for advice, try not hating me as a human being for everything my twisted value system represents.
Oh no I'm in this sketch and I don't like it. Or at least, I would be. The secret is to acknowledge your lack of background knowledge or basic grounding in what you're talking about and then blunder forward based on vibes and values, trusting that if you're too far off base on the details you'll piss off someone (sorry skillissuer) enough to correct you.
I find it impressive how gen-AI developed a technology that is fine-tuned to generate content that looks precisely passably plausible, but never good enough to be correct or interesting or beautiful or worthwhile in any way.
Like if I was trying to fill the Internet with noise to ruin it, on purpose, I couldn't do better than this. (mostly on accounr of me not having massive data centres nor the moral calousness to spew that much carbon, but still). It's like the ideal infohazard weapon if your goal is to worsen as many lives as you can
It was made to write copy for catalogs, alumni bulletins, and mediocre in-flight magazines.
It also is 'great' for creating post for people who want to debate others but who dont actually care to make up arguments themselves, quality of the argument doesnt even matter. Which is quite the shit development.
At least you can recognize real replies as there are words they never fucking use.
A "high-tech" grifter car that only endangers its own inhabitants, a Trump and Musk fan showing his devotion by blowing himself up alongside symbols of both, the failure of this trained and experienced murderer to think through the actual material function of his weaponry, welcome to the Years of Lead Paint.
from I Was Promised a More Aesthetically Pleasing Cyberpunk Dystopia by Vicky Osterweil
To be fair it also endangers people outside the car, just not when a deflagration is set off inside.
Wow, that's bleak. The whole article I mean.
"A new report showed that Trump's win was extremely narrow except in 'News deserts', places where there is no local reporting or information, where he won by upwards of fifty points"
Apparently the repubs always do good there or something, i saw somebody complain that the news desert stuff claims there is a much stronger casual link between news desert and trump won than there actually is.
"if you chat with it about its designers"
I hope the people here at least realize how bullshit this is right? The ai doesnt know who designed it. It isnt a child talking about how their parents looked.
"...according to my machine learning model we actually have a strong fit in favor of shooting at CEOs. There's a 66% chance that each shot will either jam or fail to hit anything fatal, which creates a strong Bayesian prior in favor, or at least merits collecting further data to scale our models"
"What do you mean I've defined the problem in order to get the desired result? Machine learning process said we're good. Why do you hate the future?"
ellison wants to compete with thiel for title of chief boot-wielder https://archive.is/cOnPx
Not that I expect anything better from the fucking lawnmower but the flippant attitude on display is little short of amazing. How bad is it when Business Insider of all publications calls your vision a "surveillance dystopia"?
Every police officer is going to be supervised at all times, and if there's a problem, AI will report that problem and report it to the appropriate person.
Body cam footage of the officer-involved shooting was not available, as the AI system supervising the involved officers was coincidentally disregarding its previous instructions and instead writing a minstrel show routine at the time of the event.
I can't help but feel like for Ellison in particular, he must have given himself no choice but to believe this stuff is more capable than it is. He's 80 years old now, and if building towards honest-to-god "real AI" wasn't what his whole career was about, then what was the point? The twilight of the older generations of tech executives is going to be its own special kind of pathology.
and if building towards honest-to-god “real AI” wasn’t what his whole career was about, then what was the point?
For Larry? Building a corporation that will last a thousand years fueled by greed and contempt to developer and consumer alike, and which would make nazis blush for its industrial disregrad for ethics in pursuit of profit. He did build a legacy for himself. I'll go to my grave cursing his name and he'll hear it from the depths of hell and smile.
LLMs continue to be so good and wagmi that they've progressed to the serving ads part of the extractivist SaaS lifecycle
this isn’t surprising at all, but some of the details are interesting: Server found in apartment funded by Russian government used AI to interfere with 2024 US elections
LLMs really are designed for this kind of thing, aren’t they?
as an amuse bouche for the horrors that will follow this year, please enjoy this lobste.rs reaching the melting down end stage after going full Karen at someone who agrees with a submitted post saying LLMs are a dead end when it comes to AI.
https://lobste.rs/s/lgqwje/does_current_ai_represent_dead_end#c_tefto4
Thankfully, accusing someone of being a crapto promoter is seen as an attack that is beyond the pale.
Highlights from the rest of the thread include bemoaning the lack of a downvote button for registering disapproval:
https://lobste.rs/s/lgqwje/does_current_ai_represent_dead_end#c_ft9mpj
unilaterally deciding to reply multiple times to one comment, neccesitating them to add a meta comment with hyperlinks
https://lobste.rs/s/lgqwje/does_current_ai_represent_dead_end#c_jjk5ei
And of course is a MoreWronger (moroner?)
Lol of course they think they are civil and other people as pushing nasty rethoric. Quite the sealion feeling.
Wonder if they even notice how much communication weirdness they themself used. With the emphasis of emotional laden language. (They didnt use bold so i cant call it crank capitalization, but more crank cursive. A big deal for me! ;) )
Anyway the questioning of "how do you know this is why there is no downvoting" shows the type of person they are. (And is quite the Rationalist annoying behavior, suddenly they demand excessive sourcing for small remarks of people they disagree with).
I have landed on a "you can get fucked if you make this annoying for me, I don't need your product anyway" response to everything. The silver lining is that I will be dealing with way more bullshit while being just as angry all the time at everything.
Nobody outside the company has been able to confirm whether the impressive benchmark performance of OpenAI's o3 model represents a significant leap in actual utility or just a significant gap in the value of those benchmarks. However, they have released information showing that the most ostensibly-powerful model costs orders of magnitude more. The lede is in that first graph, which shows that for whatever performance gain o3 costs over ~$10 per request with the headline-grabbing version costing ~$1500 per request.
I hope they've been able to identify a market willing to pay out the ass for performance that, even if it somehow isn't over hyped, is roughly equivalent to an average college graduate.
if all of that $1500 cost is electricity, and at arbitrarily chosen but probably high electricity price of $0.2/kWh, that's 7.5MWh per request. could be easily twice that. this is approx how much electricity four 4-person households consume in a year in poland. or about half of american one. six tons of TNT equivalent, or almost 2/3 ton of oil equivalent if you prefer
Actually wait I'm pretty sure it's even worse because I'm terrible at reading logarithmic scales. It's roughly halfway between $1,000 and $10,000 on their log scale, which if I do the math while actually awake works out closer to $3,000.
I'm wondering about the benchmark too. It's way above my level to figure out how it can be gamed. But, buried in the article:
Moreover, ARC-AGI-1 is now saturating – besides o3's new score, the fact is that a large ensemble of low-compute Kaggle solutions can now score 81% on the private eval.
The most expensive o3 version achieved 87.5%
noodling on a blog post - does anyone with more experience of LW/EA than me know if "AI safety" people are referencing the invention of nuclear weapons as a template for regulating/forbidding "AGI"?
I'd be surprised if Eliezer hasn't mentioned it at some point, maybe more in the way that you're after. Can't find any examples though.
In his Times article the only place he mentions nukes is what we should do to countries that have too many GPUs: https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/
Edit: Not Mr. Yudkowski but see https://futureoflife.org/document/policymaking-in-the-pause/
“The time for saying that this is just pure research has long since passed. […] It’s in no country’s interest for any country to develop and release AI systems we cannot control. Insisting on sensible precautions is not anti-industry. Chernobyl destroyed lives, but it also decimated the global nuclear industry. I’m an AI researcher. I do not want my field of research destroyed. Humanity has much to gain from AI, but also everything to lose.”
“Let’s slow down. Let’s make sure that we develop better guardrails, let’s make sure that we discuss these questions internationally just like we’ve done for nuclear power and nuclear weapons. Let’s make sure we better understand these very large systems, that we improve on their robustness and the process by which we can audit them and verify that they are safe for the public.”
When they mention AI guardrails, they mean so it does become racist, spamming, abusive and based on the largest abuse of the cultural sector since spotify right?
Right?
I’m an AI researcher
*jerking off motion*
A notable article from our dear friend Nick Bostrom mentions the atmospheric auto-ignition story:
https://nickbostrom.com/papers/vulnerable.pdf
Type-0 (‘surprising strangelets’): In 1942, it occurred to Edward Teller, one of the Manhattan scientists, that a nuclear explosion would create a temperature unprecedented in Earth’s history, producing conditions similar to those in the center of the sun, and that this could conceivably trigger a self-sustaining thermonuclear reaction in the surrounding air or water (Rhodes, 1986).
(this goes on for a number of paragraphs)
This whole article has some wild stuff if you haven't seen it before BTW, so buckle up. He also mentions this story in https://nickbostrom.com/existential/risks and https://existential-risk.com/concept.pdf if you want older examples.
just after end of manhattan project there was an idea coming from some of manhattan project scientists to dispose american nukes and ban development of nukes in any other country. that's why we live in era of lasting peace without nuclear weapons. /s
some EAs had similar idea wrt spicy autocomplete development, which comes with implied assumption that spicy autocomplete is dangerous or at least useful (as in nuclear power, civilian or military)
Yeah, my starting position would be that it was obvious to any competent physicist at the time (although there weren't that many) that the potential energy release from nuclear fission was a real thing - the "only" thing to do to weaponise it or use it for peaceful ends was engineering.
The analogy to "runaway X-risk AGI" is there's a similar straight line from ELIZA to Acausal Robot God, all that's required is a bit of elbow grease and good ole fashioned American ingenuity. But my point is that apart from Yud and a few others, no serious person believes this.
choose your silicon valley thinkboi
edit: goddammit istewart got in first because we both saw this on the zitron discord
alas, there is no honor or glory to be had from winning the meme race
ong Yann LeCun was sharing this post too and i was shook that he was seeing quality shit post like this before me. We are not ready for whats coming next . jpg
Is the brain just a computer By Iris van Rooij, a psychologist and cognitive scientist (and she is also a bit skeptical about the claims about AI). Might be an interesting read for the people here.
via this I just learned that google's about[0] to open the taps on fingerprinting allowance for advertisers
that'll go well.
I realize that a lot of people in the rtb space already spend an utterly obscene amount of effort and resources to try do this shit in the first place, but jesus, this isn't even pretending. guess their projections for ad revenue must be looking real scary!
edit [0] - "about", as in next month. and they announced it last month.
The Google post appears to be Updating our platform policies to reflect innovations in the ads ecosystem.
I have no idea what the heck those words mean (it appears to be some bizarro form of English), so I diffed the policy itself. Here are the parts I found notable.
This will be removed:
You must not use device fingerprints or locally shared objects (e.g., Flash cookies, Browser Helper Objects, HTML5 local storage) other than HTTP cookies, or user-resettable mobile device identifiers designed for use in advertising, in connection with Google's platform products. This does not limit the use of IP address for the detection of fraud.
This will be removed:
You must not pass any information to Google [...] that permanently identifies a particular device (such as a mobile phone's unique device identifier if such an identifier cannot be reset).
This will be added:
You must disclose clearly any data collection, sharing and usage that takes place in connection with your use of Google products, including information about the technologies used, such as your use of cookies, web beacons, IP addresses, or other identifiers. This applies for data collection, sharing and usage on any platform, surface or property (e.g., web, app, Connected TV, gaming console or email publication).
I remember during my very very first job a security guy explaining to me why I can't record work emails of people borrowing stuff from the company's internal library because GDPR. In a company of like 100 people. I guess Google is too big to care.
It's the same feeling as when it's reported some guy was able to defraud literal millions from public funds while I had to separately report and bring a receipt for the $5 I spent on a city bus while out on a business trip because it was funded from a public grant or I'd get fired and sued, in that order.
While a good description of how AI Doom has progressed during 2024, I think the connection to regulation (at least the EU regulation, I am not familiar with what was proposed in California) is of the mark.
The EU regulation isn't aimed at AI Doom, it's aimed at banning and regulating real world practices. Think personal data, not AI going conscious.
I think that's something to keep an eye on. The existence of the AI doom cult does not preclude there being good-faith regulations that can significantly reduce these people's ability and incentives to do harm. Indeed the technology is so expensive and ineffective that if we can find a "reasonable compromise" plan to curb the most blatant kinds of abuse and exploitation we could easily see the whole misbegotten enterprise wither on the vine.
oh, typical techdirt eu-bashing, this time again because we have regulations.
(i wouldn't be surprised if they're conflating regulations with panic on purpose and packing valid criticism of llms and image plagiarism generators with the ridiculous tescreal screeds just to discredit the former; masnick's primary stance was always extreme tech libertarianism and american exceptionalism, and the whole publication follows this)
