TechTakes

Oh no I was looking for more German flashcard programs (my favorite flashcard website, Seedlang, went down hopefully temporarily) and pretty much everything is forcing AI integrations of some sort.

For example Memrise goes so far as to be condescending and user hostile to people who ask for no AI: https://memrisebeta.zendesk.com/hc/en-us/articles/24937487873937-Can-I-disable-Conversations-the-AI-chatbot

It's not possible to disable the suggestions to do Conversations. [...] So, the reason it might seem like we are pushing conversation exercises is that we truly believe immersion is the key to successfully acquiring a language.

2 out of 19 found this helpful

Well excuse me for wanting to get immersion by talking to actual humans and not your shitty chatbot.

I might have to just use Anki like everyone says (my problem with Anki is I spend more time fiddling with database entries and JavaScript than actually studying)

While browsing some german news media outside my usual territorry (DW and tagesschau), and was fooled by this chameleon of an ad on the front page of WELT (trying for classy, but obvious conservative bias).

The heading means "Bitcoin could protect from inflation". If you want to check out some retail investor shilling in the wild, here you go!

Not strictly related to our normal fare, but it is on a website. HHS has been stepping up their search for snitches on people who provide gender-affirming care to trans kids. I don't know exactly what they're going to do with those reports, but it's feeling real bleak.

don't want to spam the stubsack with US politics but guys I am awfully scared

Today in relevant skeets:

::: spoiler transcript Skeet: If you can clock who this is meant to be instantly you are on the computer the perfect amount. You’re doing fine don’t even worry about it.

Quoted skeet: 'Why are high fertility people always so weird?' A weekend with the pronatalists

Image: Egghead Jr. and Miss Prissy from Looney Tunes Foghorn Leghorn shorts.

404 media: I Tested The AI That Calls Your Elderly Parents If You Can't Be Bothered

It's a service that makes an AI voice chatbot call your parents daily, so you don't have to, and then it even sends you a notification to your phone with an AI summary of what your parent told the AI.

I really didn't think that people can come up with new AI-based ideas anymore that would astonish me, but there, I was wrong, they did it. This is so cold and fundamentally alienating to me, it reminds me of that recently much-quoted Miyazaki phrase, "an insult to life itself".

jesus christ:

previous stubsack guest star Cursor rejoins the show, using a shitty liarsynth to automatically tell users broken behaviour is expected (cw: orange site), followed by people mass-killing their subscriptions

Earlier today Cursor, the magical AI-powered IDE started kicking users off when they logged in from multiple machines. Like,you’d be working on your desktop, switch to your laptop, and all of a sudden you're forcibly logged out. No warning, no notification, just gone.

Naturally, people thought this was a new policy.

So they asked support.

And here’s where it gets batshit: Cursor has a support email, so users emailed them to find out. The support peson told everyone this was “expected behavior” under their new login policy

One problem. There was no support team, it was an AI designed to 'mimic human responses'

haven’t gotten into the replies to look for sneers yet but I bet there will be some

X Is Ditching DMs Hours After News of Musk Messaging Women About Impregnating Them Breaks

New sneer from Tante, this time aimed at the entire field of CS:

"In hindsight, CFAR co-founder Anna Salamon told NBC, “we were creating conditions for a cult.”", if only there had been people warning about this, sadly such a club didn't exist. Anyway, im sure this will lead to them reflecting and changing things.

Source "Two Vegan Lovers, an AI ‘Cult,’ and a Trail of Dead Bodies", Archive

LessWronger puts in the work and determines that LLMs can't spacially vizualize for shit, comments are like "well you're prompting it wrong" (paraphrased) as well as "why not pay experiences machinists to videotape what they're doing os their work can be automated"

https://www.lesswrong.com/posts/r3NeiHAEWyToers4F/frontier-ai-models-still-fail-at-basic-physical-tasks-a#comments

The article itself is worth reading for some insights in the challenges of using current "AI" (LLMs) to work in the real world.

I got a spam message with a phishing link.... Via Github? Seriously? Are we really doing this?

Not a completely unusual comment.... From the URL it was very obvious that this was a phishing link though. Curiosity got the better of me. The site shows you a "cloudflare" captcha. OK, let's click the checkbox. The usual loading animation starts, then this is shown:

Yeah ok, right....

I'm actually a bit impressed with this, these captchas are so common, I didn't even really think about checking the box. But of course, that interaction means the browser will allow the site to add something to your clipboard.

But like.... Why distribute it via Github? I cannot think of a worse audience to try and con into "paste something random into your windows console". Am I just being naive here? Is this something common I somehow never experienced before?

article about the tactics that felon employs against the women bearing his children

features some notable sentences from his fixer, too. the sort of shit that just barely doesn’t qualify as him threatening to top off your kneecaps

New BlueSky post from Baldur Bjarnason: