TechTakes

1430 readers
103 users here now

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

founded 1 year ago
MODERATORS
401
 
 

Follow up to https://awful.systems/post/1109610 (which I need to go read now because I completely overlooked this)

Now OpenAI has responded to Elon Musk's lawsuit with an email dump containing a bunch of weird nerd startup funding drama: https://openai.com/blog/openai-elon-musk

Choice quote from OpenAI:

As we get closer to building AI, it will make sense to start being less open. The Open in openAI means that everyone should benefit from the fruits of AI after its built, but it's totally OK to not share the science (even though sharing everything is definitely the right strategy in the short and possibly medium term for recruitment purposes).

OpenAI have learned how to redact text properly now though, a pity really.

402
 
 

Sometimes what is not said is as sneerworthy as what is said.

It is quite telling to me that HN's regulars and throwaway accounts have absolutely nothing to say about the analysis of cultural patterns.

403
 
 

starting out[0] with "I was surprised by the following results" and it just goes further down almost-but-not-quite Getting It Avenue

close, but certainly no cigar

choice quotes:

Why is it impressive that a model trained on internet text full of random facts happens to have a lot of random facts memorized? … why does that in any way indicate intelligence or creativity?

That’s a good point.

you don't fucking say

I have a website (TrackingAI.org) that already administers a political survey to AIs every day. So I could easily give the AIs a real intelligence test, and track that over time, too.

really, how?

As I started manually giving AIs IQ tests

oh.

Then it proceeds to mis-identify every single one of the 6 answer options, leading it to pick the wrong answer. There seems to be little rhyme or reason to its misidentifications

if this fuckwit had even the slightest fucking understanding of how these things work, it would be glaringly obvious

there's plenty more, so remember to practice stretching before you start your eyerolls

404
 
 

Yes, I know it's a Verge link, but I found the explanation of the legal failings quite funny, and I think it's "important" we keep track of which obscenely rich people are mad at each other so we can choose which of their kingdoms to be serfs in.

405
 
 

Musk sues Altman over OpenAI going for-profit, lol

406
 
 

just think: an AI trained on depressed social justice queers

wonder what Hive is making of Bluesky

"you took a perfectly good pocket calculator and gave it anxiety"

407
 
 

inspired by

https://awful.systems/comment/2382662

from the link, we obtain HODL gang and versace bedouin (which might be the worst song I've ever heard period).

I would also like to add predatory capitalist feminism anthem WAGMI.

on an ideological level I would nominate the technofascist propaganda piece we appreciate power. banger tho

408
9
submitted 9 months ago* (last edited 9 months ago) by [email protected] to c/[email protected]
 
 

This is not so much about a particular post but rather to document Jakob Nielsen's relentless generative AI boosting.

His weekly updates are so saturated with AI subject matter and every image is AI generated they are unreadable and I can only assume the text is AI generated as well. It really doesn't matter if it isn't, in fact, because he's demonstrating in real-time how damaging the AI aesthetic is to a brand.

He also seems to be mentioning his 40 years of expertise a lot more, which might be a reaction to some negative feedback. I want to dig deeper, but I don't like the feeling that I'll have to read generated stuff carefully.

His latest newsletter triggered this post because he links to a terrible AI generated song he made (with the line "Jakob Nielsen with UX fame, forty-one years, still in the game") and spends most of the newsletter talking about the process.

https://www.youtube.com/watch?v=bYt12jr5yUY

409
 
 

410
411
 
 

(here’s a Verge article about the Waymo car getting burned during a Chinese New Year celebration)

a self-driving car got destroyed (to a round of applause from the crowd) in San Francisco! will the robot car fans on the orange site take this opportunity to explore why the tech seems to be extremely unpopular among the populations of the cities where it’s deployed?

of course the fuck not, time to spin the wheel of racist dog whistles and see which one we land on! a note to the roving orange site fans (hi, fuck off), these replies are either heavily upvoted or have broad agreement in the thread (or I’m posting them here cause I want to laugh at some stupid shit, you don’t dictate the terms of my enjoyment)

This isn't a revolt against AI. SF attracts anarchist mobs and they'll vandalize buses, trains, police cars, bikes, whatever is around.

we’re off to a strong start with some bullshit straight from musk’s twitter (which he stole from the fever dreams of the conservatives on his platform)

Alternatively: this is San Francisco where on a good day the locals don’t need much excuse to set fire to a car (although I usually associate it with the Giants winning a World Series) and this poor dumb stupid driverless Waymo drove into a celebratory and by the looks of it somewhat drunken crowd on the Streets of Chinatown during the Chinese New Year where in following its prime directive to do no harm, it got itself stuck up the creek without a paddle so to speak. Waymo probably should have accounted for that ahead of time and told their cars not to go near Chinatown this evening.

remember that no matter what, the robot car is the victim here. there’s no chance Waymo was doing anything dangerous or assholeish in the area; much like robocop, the car is an innocent victim of its fucking prime directives??? and you wouldn’t set fire to robocop, would you?

This is a hilarious take. A few youths went bonkers and defaced private property. Has nothing to do with philosophical beliefs or a Big Tech agenda. You should debate the finer points of the Big Tech agenda with them while they run up to you in a maddened rage.

yeah! I can’t wait until these angry mobs set fire to your robot car body! then you’ll see!

Arguments about driverless cars aside, the youth in this country are seriously lost. It only takes one generation of poor parenting and poor civic policies to ruin a culture.

this one is downvoted, but this reply isn’t:

Sounds like they were right. The youth at that point was lost, and are now raising people who will literally burn down a waymo for fun, or because of some horrifically ignorant idea about fairness.

oh you poor woke kids don’t like when shitty dangerous robot cars are on the streets? are you gonna start crying about how it’s “unfair” they’re covering up pedestrian injuries and traffic accidents now? your grandpa would never stand for this

412
 
 

(via mastodon)

413
 
 

So, there I was, trying to remember the title of a book I had read bits of, and I thought to check a Wikipedia article that might have referred to it. And there, in "External links", was ... "Wikiversity hosts a discussion with the Bard chatbot on Quantum mechanics".

How much carbon did you have to burn, and how many Kenyan workers did you have to call the N-word, in order to get a garbled and confused "history" of science? (There's a lot wrong and even self-contradictory with what the stochastic parrot says, which isn't worth unweaving in detail; perhaps the worst part is that its statement of the uncertainty principle is a blurry JPEG of the average over all verbal statements of the uncertainty principle, most of which are wrong.) So, a mediocre but mostly unremarkable page gets supplemented with a "resource" that is actively harmful. Hooray.

Meanwhile, over in this discussion thread, we've been taking a look at the Wikipedia article Super-recursive algorithm. It's rambling and unclear, throwing together all sorts of things that somebody somewhere called an exotic kind of computation, while seemingly not grasping the basics of the ordinary theory the new thing is supposedly moving beyond.

So: What's the worst/weirdest Wikipedia article in your field of specialization?

414
 
 

Possibly the worst defense yet of Garry Tan's tweeting of death threats towards San Francisco's elected legislature. In yet more evidence for my "HN is a Nazi bar" thesis, this take is from an otherwise-respected cryptographer and security researcher. Choice quote:

sorry, but 2Pac is now dad music, I don't make the rules

Best sneer so far is this comment, which links to this Key & Peele sketch about violent rap lyrics in the context of gang violence.

415
 
 

Also a bunch of somewhat less heinous cringe shit.

416
 
 
417
 
 

Apologies for the link to The Register...

Dean Phillips is your classic ratfucking candidate, attempting to siphon off support from the incumbent to help their opponent. After a brief flare of hype before the (unofficial) NH primary, he seems to have flamed out by revealing his master plan too early.

Anyway, apparently some outfit called "Delphi" tried to create an AI version of him via a SuperPAC and got their OpenAI API access banned for their pains.

Quoth ElReg:

Not even the presence of Matt Krisiloff, a founding member of OpenAI, at the head of the PAC made a difference.

The pair have reportedly raised millions for We Deserve Better, driven in part by a $1 million donation from hedge fund billionaire Bill Ackman, who described his funding of the super PAC as "the largest investment I have ever made in someone running for office."

So the same asshole who is combating "woke" and DEI is bankrolling Phillips, supposed to be the new Bernie. Got it.

418
 
 

Actual framing:

Cruise wasn’t hiding the pedestrian-dragging video from regulators — it just had bad internet / An independent review of an incident in which a driverless Cruise vehicle dragged a pedestrian over 20 feet concludes the company has connectivity issues.

My expectations for journalists are low but goddamn

archive link: https://web.archive.org/web/20240126195607/https://www.theverge.com/2024/1/25/24050791/cruise-pedestrian-dragging-video-driverless-report

419
420
 
 

It's just like communism in that it entirely ignores the needs and capacity of an individual and makes them slaves to the idealism of the state.

They're no better than a homeless encampment because you can't give people a house and then expect them to act like a homeowner overnight. It was the wrong problem to solve, and I think people rush to solve it out of self conscious guilt and a desire to quickly make the apparent parts of the problem disappear from common view.

It's a cruel and ridiculous strategy.

HN has had enough of those damn moochers, too:

I have found that many people who receive housing benefits are very poor stewards of their personal resources.

The first thing someone does when they get housed is to invite all their homeless friends over, to shower, to eat, to crash, to do drugs, to play games, whatever.

So your typical Section 8 housing recipient is not just a single person/family benefiting from housing, they're dragging in their entire circle of loser friends who don't have those benefits, and so now you've got a cluster of mooches who aren't invested nor responsible for disruption or damage in that community.

It's really an unfortunate thing, and I just saw it over and over again. So many people lose their benefits very quickly because they can't resist helping other folks out, but that's not what you do with welfare and entitlements.

421
 
 

remember, regardless of how outspoken you are in life, nothing will stop the capitalists from reanimating your defiled corpse into a shitheaded centrist zombie if there’s a buck in it:

“I'd just like to say that as much as I think billionaires are destroying the fabric of society with unchecked greed and blatant self-interest at the expense of basic human rights for everyone else, it is a little strange to me that people get mad at them. People are the ones who gave them the money in the first place," the AI Carlin said.

(editor’s note: the above is supposed to be a joke from the comedy special these fucking assholes hijacked Carlin’s corpse to promote. I can’t find the punchline, but it’s supposed to be a joke)

422
 
 

Choice quote:

Actually I feel violated.

It's a KYC interview, not a police interrogation. I've always enjoyed KYC interviews; I get to talk about my business plans, or what I'm going to do with my loan, or how I ended up buying/selling stocks. It's hard to empathize with somebody who feels "violated" by small talk.

423
 
 

The day just isn't complete without a tiresome retread of freeze peach rhetorical tropes. Oh, it's "important to engage with and understand" white supremacy. That's why we need to boost the voices of white supremacists! And give them money!

424
 
 

It was only a matter of time that we saw a TechTake from this guy. I'm sorry to inflict Peterson on y'all, but this was too funny not to post.

425
 
 

Global outage on fetching posts. Funny enough, some features are still working as evidenced by the fact #TwitterDown is trending.

Two HN threads about this now, looking forward to some excellent takes

https://news.ycombinator.com/item?id=38717367 https://news.ycombinator.com/item?id=38717326

view more: ‹ prev next ›