*See Bible for terms and conditions
(certain restrictions may apply)
Benn Jordan was banned from a conf he paid for and his room reservation was cancelled... So they stayed outside and used their “unblocked electromagnetic frequencies” 🤓
Flock values their privacy. Oh well.
YouTube
RE: https://mas.to/@zzt/117134157732461816
This highlights one of the big problems with LLMs for code generation.
If a thing has been done often enough that there are are loads of examples of it for an LLM to train on, you probably shouldn't be doing it again. You should be using some abstraction over the thing and not copying and pasting it.
If a thing is rare enough that it has been done a handful of times in the training set, there's a good chance that the LLM output will meet the legal definition of a derived work, you may be accidentally violating a license / infringing copyright.
If a thing is so rare that there are no examples in the training set, the LLM will probably produce something that looks right. There is even a chance that it is right. Good luck.
@david_chisnall "If a thing is so rare that there are no examples in the training set, the LLM will probably produce something that looks right."
I tried using one for "rewrite my code to this backwards-incompatible version of this library's API which was released two days ago and which I am required to use whether I like it or not to fix a CVE".
It went round in loops with a sequence of random guesses none of which compiled.
"Ethical use of generative AI" discusses the wrong subjects and it's the AI industry using the "CO2 footprint" trick all over again.
There's an interesting insight in the podcast discussion linked below.
Right now, there's a lot of discourse about the "ethical use of (generative) AI". You'll also find job offers and government grants about this topic and a slew of research institutes and NGOs putting their focus on it.
But: This "ethics" discussion is usually around the end user's use cases. It's the end user who supposedly decides whether using (generative) AI is ethical or not. It's the end user who chooses right or wrong and will be judged for it.
We're making fun of those weak people forming friendships or falling in love with "their" chatbot, we're making fun of the lazy students failing to learn. It's their individual moral failure of having chosen the wrong way to use generative AI.
However, the questionable ethical choices happened prior to this, the power lies with the company offering the generative AI product - which is intentionally forcing to use AI onto the end user, with addictive dark UI patterns.
The "ethical use of (generative AI)" discourse is therefore a repeat of the "CO2 footprint" trick that the fossil industry used to avoid regulation and shift responsibility for the industry's negative societal consequences onto the end consumer.
https://omny.fm/shows/it-could-happen-here/how-ai-companies-are-monetizing-intimacy-ft-bridget-todd
This one is gold!
It's not only YouTube, but also Facebook, Instagram, and any other network.
Disclaimer: Not on Mastodon 🙏
#Mastodon #fediverse #Socialmedia #Advertising #Jokes #Reality #Check
Fun Fact: When I interviewed Forrest J Ackerman many years ago, he admitted to me that he indeed had been Edward D. Wood, Jr.'s "il-literary" agent. Sidenote: Maila Nurmi ("Vampira") recorded a promo for me.
Are you going to Worldcon (LACon V) in Anaheim next week?
I'm going; it's been a long time -- my last Worldcon (or any SF con at all) was LACon III in 1996. I'm looking forward to seeing a lot of old friends (now mostly septuagenarians, yow!).
Tante on AI: ‘When this thing crashes, I want to see apologies’
‘Many people from very different groups can come together hating AI’
https://www.youtube.com/watch?v=beUZ2scYWUA&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20260821-tante-on-ai-when-this-thing-crashes-i-want-to-see-apologies - podcast
Cleaned-up transcript for patrons: coming shortly
time: 49 min 52 sec
Are we plugging our phone physically with a wire into our computers and moving files between them?
@danirabbit I like to maintain a local music library and copy everything from computer to the phone. Plugging my android device in is quicker and less flaky than anything wireless regardless of what OS the computer is running.
This is about as close as you can get to explicitly saying, "don't worry, she doesn't *need* to know anything or think for herself."
RE: https://bsky.app/profile/did:plc:4llrhdclvdlmmynkwsmg5tdc/post/3mtm2ishtni2j
It's still a little buggy and unoptimized, but I am finishing up a GPUI port of Runebender today. Here it is on Omarchy Quattro Linux.
100% Rust
"Shape your brand’s future with .AI. Lock in the domain extension that defines the next decade at the best price in the market."
say what you will about the myriad valid concerns with current LLMs but I personally will never get over the fact that Aaron Swartz was prosecuted to his literal death for downloading JSTOR and then a decade later, they downloaded & ingested the entire world and our IP laws just went ¯\_(ツ)_/¯
@partim Ha! Yeah, locals told me about some roads where there were frequent accidents involving cars and ‘roos.
I don’t know about crocodiles here. I am now in Brisbane right in the core of the city section, so wild animals are not around.
@danyork I bet you have similar stories about moose in your part of the world.
I had one on the road just behind a bend in New Brunswick once. Luckily it was a narrow and very slow road.
@ed @emanuelecariati@vernissage.photos @emanuelecariati@hub.vocalcat.com @Donutsaurus @thornsinnercircle
FediProfile funziona come LinkStack. La tua pagina è un attore di activitypub: https://hub.vocalcat.com/emanuelecariati + https://hub.vocalcat.com/
Vernissage è una sorta di Flickr federato. Più incentrato sulla fotografia: https://joinvernissage.org/
Spero possano aiutarti questi link 😅
@emanuelecariati@varese.social @ed @emanuelecariati@vernissage.photos @emanuelecariati@hub.vocalcat.com @thornsinnercircle grazie, do un'occhiata a Vernissage che sembra interessante! In generale comprendo però il fediverso quanto mia nonna potrebbe comprendere la trama di Kingdom Hearts, quindi sono molto confuso da tutto ciò che ho letto 😂
delete a toot every morning and be free!
- delete toots _over_ a certain interaction threshold instead of under
- delete toots that break containment
Bored cat develops telekinesis after eating dubious brand cat food
Habe endlich mal ein Mini-Projekt von mir “fertig” gemacht: https://ardsounds.tisba.de
Mich störte es schon super lange, dass die ARD in ihrer Audiothek (“ARD Sounds”) zwar alles “Podcasts" nennt, aber so gut wie nichts einen Podcast Feed hat 🤬
Nun kann man einfach die URL von einer Sendung eingeben und bekommt eine Podcast Feed URL für den Podcatcher der Wahl. Die Daten dazu kommen von der offiziellen GraphQL API der ARD. Downloads kommen direkt von der ARD.
Feedback Welcome 🙇♂️💕
I dislike it so when people say “the planet is dying” as a result of anthropogenic global heating. It’s so navel-gazing and self-involved.
*Humanity* may not survive the current epoch, but “the planet” will. The planet’ll be just fine.
@adamgreenfield As an English major I'm forced sometimes to give up and regard usage as the final social arbiter, which suggests "the planet" as shorthand is likely to win the day, along with "the calvary is coming," "do you want some sherbert" and "my prostrate is acting up." That said, whenever I repost from elsewhere anything significant with "the planet" in the headline, I know someone will savage me for failing to make the distinction on behalf of the source, so I self-censor for the pedants. I'll do my best. It's tiring, though.
EFF’s Lena Cohen joined Tunnel Vision’s Christophe Haubursin to explain how your personal data is being shared, recorded, and used in the milliseconds before you see an online ad. https://youtu.be/U0-8SrbG0xA?si=GbBdpiyU13I1ud3B
@eff I keep mine in another room when I'm at home. And I never see ads on it.
👾 Super ZSNES v0.300 - RetroRGB