Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
as previously discussed, the rabbit r1 turns out to be (gasp) just an android app.
in a twist no one saw coming, the servers running "rabbit os" report to just be running Ubuntu, and the "large action model" that was supposed to be able to watch humans use interfaces and learn how to use them, turns out to just be a series of hardcoded places to click in Playwright.
Nick Bostrom's advertising his new book about what if AIs let us sleep until noon and wouldn't that be grand.
Also call me paranoid, but "A winter wonderland glittering with possibilities for discovery and play" sounds exactly like the sort of thing an LLM might generate.
Also there was a whole thing about how CGTP tends to use 'delve' a lot because it's been RLHF'd by speakers of nigerian english, so yeah.
This sneer sponsored by me getting surgery last Friday and being extremely sore
https://www.lesswrong.com/posts/bkr9BozFuh7ytiwbK/my-hour-of-memoryless-lucidity
Surgery is the perfect opportunity to test one's "Am I in the matrix?" mental sigils and other fun rationalist mind games!
Several years ago, I was thinking about worthwhile precautions to take against strange scenarios and wanted a way to defend against erasure of my short-term memory, e.g. by the CIA or alien abductions.
Of course someone made a follow up post encouraging readers to experiment with dangerously large drug dosages for giggles.
This is presumably the sort of thing which is tough to get past an institutional review board these days, but easy to do yourself over the weekend with a friend or two.
Little mention that this might be a terrible idea, but at least someone pointed it out in the comments
Important notice: benzodiazepines are serious business: benzo withdrawals are amongst the worst experiences a human can go through, and combinations of benzos with alcohol, barbiturates, opioids or tricyclic antidepressants are very dangerous: benzos played a role in 31% of the estimated 22,767 deaths from prescription drug overdose in the United States.
Lesswrong and drug abuse. A match made in heaven.
Jesus, that second post.
[Benzos are] not only a great amnestic, it’s also apparently one of the most heavily prescribed drug classes historically, and used recreationally - which puts very strong lower bounds on the drug’s safety in practice, and means it’s probably readily available.
Yeah, because no widely used drug has ever turned out to be harmful, right?! Right?! Now if you'll excuse me, I'm off to huff some glue. After all, glue huffing is widely practiced recreationally - by kids, even - so it's probably safe, right?
Benzos
Apparently JBP might still be on benzos, after he tried the russian coma method (which seems to have blasted his mind further, he was already weird before, but now he just seems even dumber). So GG there LW.
it sounds like what they're trying to do is Wada test, for no particular reason, with zero knowledge what they're trying to do or what tools do they need for that
about everything what could they learn from that is already described in psychiatry and psychology textbooks, but it's too hard for homeschooled first-principlers like them
Nothing illustrates the difference between rational and Rationalist than ignoring the advice of trained professionals and instead roping in a SO to help perform ad-hoc personality tests. Dude's gonna need to work to keep that girlfriend coz it's unlikely he's going to find someone else willing to put up with his shit.
this is peak 18-century science
Inorite! “Here, let me replace my blood with blood from a horse, it should make me stronger”
we have mkultra at home
on the other hand, that's just speedrunning being jordan peterson
on the third hand, our very good friend apparently never had been drunk
on the fourth hand, our very good friend also didn't clock that maybe his ability to assess his own mental state was disturbed (it was) (severely)
on the fifth hand, this is an effect of many drugs, iv line suggests midazolam but maybe ketamine was an option. neither are exactly easily available for regular civilian (other benzos are)
first principling that benzos are completely safe, yeah way to go, 10/10. also not their first rodeo, but lwers on drugs would be megaminds on adderal, at least that's what could be noticed previously
Several years ago, I was thinking about worthwhile precautions to take against strange scenarios and wanted a way to defend against erasure of my short-term memory, e.g. by the CIA or alien abductions.
What...
Makes you wonder if LW just attracts people with weird paranoid thoughts, or that you get more weird paranoid thoughts due to being a lot on lesswrong. Right
This is another example of how LW is a cult, if some other commenter had replied like this you’d expect the reply to be ignored but the respondent is of the same tribe, so author has no problem exposing them (and himself, by extension) as cranks to an appreciative audience (us).
Google has supplemented their AI customer service backend with a just as stupid AI front-end. They're pushing you to make refund claims with "Google AI". It's a chat bot, and you're pushed to use it to see the status of your refunds. Clicking a button and viewing a table of refund requests is apparently too Web 2.0, so instead you can ask a chatbot to pull up that status of your refunds one...by...one. It's the second least useful use of a chatbot I've ever seen personally, right behind the chatbot they put into my office's seating software.
someone re-invented email-threading. It's billed as a way to have "long and complex discussions." Their prototypical example? Yud's utterly pointless thoughts on imaginary scenarios conjured up by his immense brain. Orange site is skeptical and yearns for 4chan instead.
The feature I am looking forward to the most in comminication apps is having a machine learning model listen to those "quick calls", generate summary and action items and post them right back in the thread. You get the benefits of both worlds that way.
Oh, you do, do you?
Fractal sealions
Their entire manifesto could have been replaced with:
"Oh hey check out our threads in threads demo so you can comment on comments!"
Like it's not that deep. It doesn't need a manifesto.
This manifesto could have been an e-mail!
Hi, Im threadGPT, I am an advanced AI which will make concise bullet points out of longer email or comments, here is a bulletpointed list of the above comment.
The slatestar subreddit is doing its regular so what's up with all the racists constantly crawling out of the woodwork around here surprised pikachu thread, in response to Scotty doing Hanania Week in the substack.
You can like a thinker without endorsing all of their beliefs, even if their beliefs are evil. Why do people like Schmitt and Heidegger even though they were fascists? Or Foucault given his views on the age of consent? I agree that Hanania's views are relevant context, but I think it's fine to write a book review that doesn't try to analyse the author's motivations or the book's place in a wider political context.
Hanania is clearly analogous to Foucault and Heidegger, and also is it even wrong to completely divorce a work from all context.
I think Scott was simply more interested in writing an article on arguments aginst civil rights law than an article on whether Hanania is engaged in an insidious project to smuggle rascist ideas into the mainstream via his legal arguments, and frankly I find that kind of review more interesting too. Perphaps this is irresponsible, but at the end of the day Scott is a modestly influential blogger that just likes to write about things he finds interesting.
uwu smolbean blogger with absolutely no agenda besides the pursuit of truth and civility strikes again.
Scott is of the opinion that being able to maintain peaceable discourse with people who you deeply disagree with on political issues is an important feature of society which we shouldn't readily make exceptions to.
"Scott being nice to racists and reviewing their books positively actually means he's less racist" is a good rhetorical trick.
HBD is a legit line of scientific inquiry you guys, it's not just eugenics obsessed weirdoes and fascists trying to bring back birthright as the primary path to privilege.
Amazing how much they memoryholed the Scott emails.
Ohio State commencement speaker was grifter Chris Pan, who tossed a suggestion to buy bitcoin into the middle of his speech, got audibly booed. Full speech (around 1:33:45)
Molly has some fun details on Musk's hell site.
More fun detail's from Ohio's The Rooster (definitely read this one!)
From the notes, it starts with normal inspirational speech garbage-- life pro-tip: do not repeat stupid parables saying that blind people have limited perspectives to large audiences-- then bitcoin comes up in the middle:
I know this might feel polarizing but I encourage you to keep an open mind. Right now, I see Bitcoin as a very misunderstood asset class. It is decentralized and finite which means no government can print more at will. In the early days, the exchanges for Bitcoin were prone to hacks and fraud. But this issue has been solved with the recent launch of bitcoin ETFs backed by the world’s 2 largest asset managers, BlackRock and Fidelity. And you can hold these ETFs in your retirement accounts just like you hold the S&P 500.
I’d love to do a demo for you: So here are 4 quarters. Inflation after 4 years has turned this into 3 quarters of purchasing power. Now if we apply some innovation and open-mindedness… Investing in your financial literacy will unlock so much freedom and possibilities for you.
EA forum: Stock prices in a VC-forced bubble show that society will make trillions from AI!!!
first comment: "uh, wtf, that's not how anything works."
reply to this comment: "oh yeah? well DEBATE ME"
For such ostensibly smart people, the basics of finance and economics are surprisingly elusive!
"Familial Transmission of Personality Is Higher Than Shown in Typical Studies" - posted to HN
Another set of ~~social darwinisms~~ dark bio troofs feat. 2 year old accounts made by promptfondeling libertarians
Oh look, another company’s “we won’t data scrape against our users’ wishes to feed spicy autocomplete” mask has just crumbled. I am so surprised.
Let’s see how many of my comments I can delete or overwrite with garbage before they ban me.
i couldn't delete the one question i had on stackoverflow, so i used a text generator to overwrite the body and title of the question. fight garbage with garbage
Is the AI lying? wonders the Guardian and turns to this guy:
“As the deceptive capabilities of AI systems become more advanced, the dangers they pose to society will become increasingly serious,” said Dr Peter Park, an AI existential safety researcher at MIT and author of the research.
Who of course validates the sentient AI frame. They should have asked him if this means that we are closer to Terminator or Matrix.
Orange Site denizen plays Dr. LLM: https://news.ycombinator.com/item?id=40331850
Show NH [sic]: "data-to-paper" - autonomous stepwise LLM-driven research
data-to-paper is a framework for systematically navigating the power of AI to perform complete end-to-end scientific research, starting from raw data and concluding with comprehensive, transparent, and human-verifiable scientific papers
The example "research paper" was some useless fluff about diabetes, based off an existing data set (read: actual work produced by actual humans), and mad-libs.
The study identifies an inverse correlation between physical activity and fruit and vegetable intake with diabetes occurrence, while higher BMI is positively correlated
I'm too sleepy and statistics-impaired to check how nonsensical the regression "analysis" or findings are, so instead let's check out the references (read: the actual humans who were plagarized to make this fluff)!
Reference #5
[5] T. Schnurr, Hermina Jakupovi, Germn D. Carrasquilla, L. ngquist, N. Grarup, T. Srensen, A. Tjnneland, K. Overvad, O. Pedersen, T. Hansen, and T. Kilpelinen. Obesity, unfavourable lifestyle and genetic risk of type 2 diabetes: a case-cohort study. Diabetologia, 63:1324–1332, 2020.
This incredibly managed to mangle all non-English alphabet names:
Hermina Jakupović, Germán D. Carrasquilla, Lars Ängquist, Thorkild I. A. Sørensen, Anne Tjønneland, Tuomas O. Kilpeläinen
I guess AI has an easier time advancing science than producing a PDF with non-ascii text in it
The fact that actual engineers have been trying to educate newcomers on Unicode for at least 20 years and not only is it still pervasively ignored but the hottest, newest, cutting edge AI that Will Change Everything™ with billions of dollars and so many manhours behind it gets absolutely dumbfounded when it sees é is the exact combination of funny and sad that will eventually result in me turning into a Butlerian Jihad Joker.
Every respectable programming language has functionality in its standard library that recognises letter characters
As a C++ programmer I've never been so offended by something I so entirely agree with.
Introducing AI-Powered Audiobooks of Rational Fiction Classics
You know, sometimes I wish my words had touched more lives, but at least I can console myself with the discovery that nothing I've written has fucked people up that badly.
https://less.online/ - they couldn't be bothered to write the 5 lines of js to do their karma discount automatically
"A Festival of Writers Who are Wrong on the Internet"
How is it possible to be so thoroughly immune to self-reflection.
Bumble founder Whitney Wolfe Herd says the app could embrace AI
"Your dating concierge could go and date for you!" this woman actually said out of her mouth
previously both a Black Mirror and an Oglaf (highly NSFW)
I haven't read it yet (and I might not even bother) but Google's probably lying about their ML models again
"I can predict the structure and interactions of all of life’s molecules"
