Discussion
Loading...

#Tag

Log in
  • Sign up
  • About
  • Code of conduct
  • Privacy
  • Users
  • Instances
  • About Bonfire
Gerry McGovern boosted
Ciara
Ciara
@CiaraNi@mastodon.green  ·  activity timestamp 1 hour ago

“This flood of low-quality and fake content from AI is particularly affecting minor languages. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.”

Reading this with Danish-speaking eyes, I realised that this problem had never occurred to me. Hvor dejligt. A whole new consequence of contemporary AI to dislike.

From '99th Day: A Warning About Technology' by Gerry McGovern

#AI #Books
#Language

Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
  • Copy link
  • Flag this post
  • Block
Ciara
Ciara
@CiaraNi@mastodon.green  ·  activity timestamp 1 hour ago

“This flood of low-quality and fake content from AI is particularly affecting minor languages. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.”

Reading this with Danish-speaking eyes, I realised that this problem had never occurred to me. Hvor dejligt. A whole new consequence of contemporary AI to dislike.

From '99th Day: A Warning About Technology' by Gerry McGovern

#AI #Books
#Language

Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
Passage 99th Day: “AI is efficiently churning out fake books on how to get rich quick or to how to code quick. Once you’ve found an AI author on Amazon, with fake face and fake bio, then Amazon will recommend you a range of other fake AI authors. AI can churn out books in the blink of an eye. If you write a popular book, AI will efficiently write fake workbooks to accompany your book, riddled with lies and errors. When a celebrity or even semi-celebrity dies, the next morning AI will have a book published on them, full of lies, misquotes and misinformation. This flood of low-quality and fake content from AI is particularly affecting minor languages, which don’t have a lot of original content published on the Web to begin with. A lot of content initially gets written in major languages such as English, and then, if it is translated at all into minor languages, it is usually done by machine translation. AI acts as a type of hyper machine translation. These minor languages then become overwhelmed by AI translations and other fake AI content.’ “We find that indiscriminate use of model-generated content in training causes irreversible defects in the resulting models in which tails of the original content distribution disappear." AI begins to feed on itself, to train and learn from its own crap. They call this “model collapse” because the data will get worse & worse and all sorts of weird things will begin to happen. If AI were a digestive system, it would have no capacity to poop"
  • Copy link
  • Flag this post
  • Block
Autonomie und Solidarität boosted
Autonomie und Solidarität
Autonomie und Solidarität
@autonomysolidarity@todon.eu  ·  activity timestamp 13 hours ago

Algorithmen auf Streife: Bremens Straßenbahnen werden zur KI-Überwachungszone

Mit #AI-Watch führt die #BSAG eine Echtzeit-Analyse von Fahrgästen ein. Was als Sicherheitsgewinn verkauft wird, markiert eine neue Stufe der #Videoüberwachung….

https://www.heise.de/news/Algorithmen-auf-Streife-Bremens-Strassenbahnen-werden-zur-KI-Ueberwachungszone-11185075.html

#Bremen #Überwachung #KI #Technology #Digitalisierung

https://todon.eu/tags/%C3%9Cberwachung
heise online

Algorithmen auf Streife: Bremens Straßenbahnen werden zur KI-Überwachungszone

Mit AI-Watch führt die BSAG eine Echtzeit-Analyse von Fahrgästen ein. Was als Sicherheitsgewinn verkauft wird, markiert eine neue Stufe der Videoüberwachung.
https://todon.eu/tags/Video%C3%BCberwachung
  • Copy link
  • Flag this post
  • Block
bobocubed
bobocubed
@boelder@shakedown.social  ·  activity timestamp 12 hours ago

Please expect more and more of this from our shitty media, as AI and climate change makes potable water impossible to afford.

(I do not know if this is real or satire, and am having an increasingly hard time determining the difference between the two)

#fuckai #ai #endstagecapitalism

A screen cap of a New York Post article on social media (have not checked as to the legitimacy of the headline) with two people showering, lathered up with soap and water. Headline is 'Daily showers are purely 'performative' and have no real health benefit, experts insist.' 

A reply from 'Ed' says, 'The health benefit is so you don't smell like shit. Hope this helps.'
A screen cap of a New York Post article on social media (have not checked as to the legitimacy of the headline) with two people showering, lathered up with soap and water. Headline is 'Daily showers are purely 'performative' and have no real health benefit, experts insist.' A reply from 'Ed' says, 'The health benefit is so you don't smell like shit. Hope this helps.'
A screen cap of a New York Post article on social media (have not checked as to the legitimacy of the headline) with two people showering, lathered up with soap and water. Headline is 'Daily showers are purely 'performative' and have no real health benefit, experts insist.' A reply from 'Ed' says, 'The health benefit is so you don't smell like shit. Hope this helps.'
  • Copy link
  • Flag this post
  • Block
Arnold Knijn boosted
Wladimir Mufty
Wladimir Mufty
@wlaatje@social.edu.nl  ·  activity timestamp 14 hours ago

Raising a child uses energy too…. says #SamAltman.

When you start comparing feeding babies to data centers, you’re not making an “intelligent” analogy, you’re revealing a worldview as tech leader where humanity is measured in megawatts and inefficiencies…

#chatGPT #Coplot #ai #humanity #values

Your browser does not support the video tag.
GIF
GIF
Open
GIF
  • Copy link
  • Flag this post
  • Block
Autonomie und Solidarität
Autonomie und Solidarität
@autonomysolidarity@todon.eu  ·  activity timestamp 2 days ago

‚Neun bis elf digitale Kameraaugen pro Wagen wachen über die Geschehnisse in den Bremer Stadtbahnen. Bis Ende 2026 soll in diese Kameras eine künstliche Intelligenz (KI) eingebaut werden, die potenzielle Gefahrensituationen wie aggressives Verhalten und Gewalttaten in Echtzeit erkennen können soll….‘

KI-Überwachung im NahverkehrBig Brother in der Tram

https://taz.de/KI-Ueberwachung-im-Nahverkehr/!6155933/

#Ki #Überwachung #Videoüberwachung #Bremen #Digitalisierung #Tech

taz.de

KI-Überwachung im Nahverkehr: Big Brother in der Tram

Die Bremer Straßenbahn stellt ein KI-Tool vor, das aggressives Verhalten erkennen soll. Es soll helfen, schneller einzugreifen und Gewalt vorzubeugen.
1 more link(s)
Autonomie und Solidarität
Autonomie und Solidarität
@autonomysolidarity@todon.eu replied  ·  activity timestamp 13 hours ago

Algorithmen auf Streife: Bremens Straßenbahnen werden zur KI-Überwachungszone

Mit #AI-Watch führt die #BSAG eine Echtzeit-Analyse von Fahrgästen ein. Was als Sicherheitsgewinn verkauft wird, markiert eine neue Stufe der #Videoüberwachung….

https://www.heise.de/news/Algorithmen-auf-Streife-Bremens-Strassenbahnen-werden-zur-KI-Ueberwachungszone-11185075.html

#Bremen #Überwachung #KI #Technology #Digitalisierung

https://todon.eu/tags/%C3%9Cberwachung
heise online

Algorithmen auf Streife: Bremens Straßenbahnen werden zur KI-Überwachungszone

Mit AI-Watch führt die BSAG eine Echtzeit-Analyse von Fahrgästen ein. Was als Sicherheitsgewinn verkauft wird, markiert eine neue Stufe der Videoüberwachung.
https://todon.eu/tags/Video%C3%BCberwachung
  • Copy link
  • Flag this comment
  • Block
Wladimir Mufty
Wladimir Mufty
@wlaatje@social.edu.nl  ·  activity timestamp 14 hours ago

Raising a child uses energy too…. says #SamAltman.

When you start comparing feeding babies to data centers, you’re not making an “intelligent” analogy, you’re revealing a worldview as tech leader where humanity is measured in megawatts and inefficiencies…

#chatGPT #Coplot #ai #humanity #values

Your browser does not support the video tag.
GIF
GIF
Open
GIF
  • Copy link
  • Flag this post
  • Block
guerrilla stickers 通信 boosted
SirAldo Canaveni
SirAldo Canaveni
@psicologo@mastodon.uno  ·  activity timestamp 20 hours ago

Siamo circondatə! 💀
@psicologia

#psicologia #IA #AI #chatbot


Assistente AI
per Psicologi: 

- Come ti posso aiutare oggi?
- Aiutami con una diagnosi di
disturbo d'ansia generale.
- Chiedi all'Assistente A.l.
Assistente AI per Psicologi: - Come ti posso aiutare oggi? - Aiutami con una diagnosi di disturbo d'ansia generale. - Chiedi all'Assistente A.l.
Assistente AI per Psicologi: - Come ti posso aiutare oggi? - Aiutami con una diagnosi di disturbo d'ansia generale. - Chiedi all'Assistente A.l.
  • Copy link
  • Flag this post
  • Block
SirAldo Canaveni
SirAldo Canaveni
@psicologo@mastodon.uno  ·  activity timestamp 20 hours ago

Siamo circondatə! 💀
@psicologia

#psicologia #IA #AI #chatbot


Assistente AI
per Psicologi: 

- Come ti posso aiutare oggi?
- Aiutami con una diagnosi di
disturbo d'ansia generale.
- Chiedi all'Assistente A.l.
Assistente AI per Psicologi: - Come ti posso aiutare oggi? - Aiutami con una diagnosi di disturbo d'ansia generale. - Chiedi all'Assistente A.l.
Assistente AI per Psicologi: - Come ti posso aiutare oggi? - Aiutami con una diagnosi di disturbo d'ansia generale. - Chiedi all'Assistente A.l.
  • Copy link
  • Flag this post
  • Block
Lord Caramac the Clueless, KSC boosted
tagesschau
tagesschau
@tagesschau@ard.social  ·  activity timestamp yesterday

Studie: Wenn KI nicht weniger, sondern mehr Arbeit macht

Laut einer US-Studie entlastet Künstliche Intelligenz Arbeitnehmer nicht, sondern kann zu Mehrarbeit und nachlassender Effizienz führen. Ein deutscher Forscher macht ähnliche Beobachtungen. Von Frank Bäumer.

➡️ https://www.tagesschau.de/wissen/forschung/ki-arbeitsintensitaet-folgen-100.html?at_medium=mastodon&at_campaign=tagesschau.de

#KI #AI #KünstlicheIntelligenz #Arbeit #Job

https://ard.social/tags/K%C3%BCnstlicheIntelligenz
tagesschau.de

Studie: Wenn KI nicht weniger, sondern mehr Arbeit macht

Laut einer US-Studie entlastet Künstliche Intelligenz Arbeitnehmer nicht, sondern kann zu Mehrarbeit und nachlassender Effizienz führen. Ein deutscher Forscher macht ähnliche Beobachtungen. Von Frank Bäumer.
  • Copy link
  • Flag this post
  • Block
tagesschau
tagesschau
@tagesschau@ard.social  ·  activity timestamp yesterday

Studie: Wenn KI nicht weniger, sondern mehr Arbeit macht

Laut einer US-Studie entlastet Künstliche Intelligenz Arbeitnehmer nicht, sondern kann zu Mehrarbeit und nachlassender Effizienz führen. Ein deutscher Forscher macht ähnliche Beobachtungen. Von Frank Bäumer.

➡️ https://www.tagesschau.de/wissen/forschung/ki-arbeitsintensitaet-folgen-100.html?at_medium=mastodon&at_campaign=tagesschau.de

#KI #AI #KünstlicheIntelligenz #Arbeit #Job

https://ard.social/tags/K%C3%BCnstlicheIntelligenz
tagesschau.de

Studie: Wenn KI nicht weniger, sondern mehr Arbeit macht

Laut einer US-Studie entlastet Künstliche Intelligenz Arbeitnehmer nicht, sondern kann zu Mehrarbeit und nachlassender Effizienz führen. Ein deutscher Forscher macht ähnliche Beobachtungen. Von Frank Bäumer.
  • Copy link
  • Flag this post
  • Block
Lord Caramac the Clueless, KSC boosted
Dr. Fortyseven 🥃 █▓▒░
Dr. Fortyseven 🥃 █▓▒░
@fortyseven@defcon.social  ·  activity timestamp last month

One of these photos is AI generated using the Z-Image Turbo model (running on local hardware), and the other is a real photo taken in Minnesota.

I've placed the answer in the alt text.

If you'd asked me which one was the generated image, I'd have probably chosen the real one.

Maybe, MAYBE with meticulous, slow examination I could have figured it out. Or at least attributed a chance percentage of it being generated.

But man... we've really fucked ourselves, I think. This is just the state of it all TODAY. And that's just here at home on consumer hardware. And this isn't even the best, but where the hell else is there to go in terms of image fidelity for uses like this?

And it comes at the absolute worst time in US history, when video and photograph evidence is the most crucial...

#genai #zimageTurbo #ai

2 media
REAL.
REAL.
REAL.
ZIMAGE TURBO.
ZIMAGE TURBO.
ZIMAGE TURBO.
  • Copy link
  • Flag this post
  • Block
Lord Caramac the Clueless, KSC and 1 other boosted
Ben Ramsey
Ben Ramsey
@ramsey@phpc.social  ·  activity timestamp 2 days ago

Quote du jour: “‘Top-down mandates to use large language models are crazy,’ one employee told Wired. ‘If the tool were good, we’d all just use it.’”

https://futurism.com/artificial-intelligence/jack-dorsey-block-falling-apart-ai

#AI

Futurism

Jack Dorsey's New Company Falling Apart as It Forces Employees to Use AI

Twitter founder Jack Dorsey is running into some serious issues while overhauling his financial services company, Block.
  • Copy link
  • Flag this post
  • Block
Glowing Cat of the Nuclear Wastelands liked this activity
Ben Royce 🇺🇦 🇸🇩
Ben Royce 🇺🇦 🇸🇩
@benroyce@mastodon.social  ·  activity timestamp 2 days ago

if you like #publictransportation and dislike autonomous #robots and #ai

oh boy do i have a video for you

Your browser does not support the video tag.
This video cannot be previewed
Open original
a delivery robot on train tracks gets obliterated by a passenger train
  • Copy link
  • Flag this post
  • Block
Glowing Cat of the Nuclear Wastelands boosted
Ben Royce 🇺🇦 🇸🇩
Ben Royce 🇺🇦 🇸🇩
@benroyce@mastodon.social  ·  activity timestamp 2 days ago

if you like #publictransportation and dislike autonomous #robots and #ai

oh boy do i have a video for you

Your browser does not support the video tag.
This video cannot be previewed
Open original
a delivery robot on train tracks gets obliterated by a passenger train
  • Copy link
  • Flag this post
  • Block
Deborah Pickett
Deborah Pickett
@futzle@old.mermaid.town  ·  activity timestamp yesterday

I’m starting to notice that there’s an AI slop typeface. Or more accurately, lettering. A particular style of letters that prompted image generators gravitate to. Sans serif (serifs are hard). No stroke variation (easy to get wrong if you forget how typography evolved from writing). A little bit condensed. Quite naive and reminiscent of abandoned grotesque fonts from early the 1900s. Very rarely used in the real world. #ai #llm #typography

Detail from an AI-generated postcard advertising a Melbourne that doesn’t exist. There’s a yellow sign with the inexplicable sequence of letters MADERN, in ~that font~.
Detail from an AI-generated postcard advertising a Melbourne that doesn’t exist. There’s a yellow sign with the inexplicable sequence of letters MADERN, in ~that font~.
Detail from an AI-generated postcard advertising a Melbourne that doesn’t exist. There’s a yellow sign with the inexplicable sequence of letters MADERN, in ~that font~.
  • Copy link
  • Flag this post
  • Block
ophiocephalic 🐍 boosted
Autonomie und Solidarität
Autonomie und Solidarität
@autonomysolidarity@todon.eu  ·  activity timestamp last week

Disgusting. Frontex is building an ‘AI chatbot app’ to encourage repatriations

‚The Border Agency of the European Union, Frontex, is building a mobile application for individuals who are voluntarily or forcibly deported from Europe, as part of its EU Reintegration Programme (EURP).

The application will come equipped with an ‘AI chatbot’ feature, which allows deportees to learn about the available services within the program, such as the opening hours of the nearest return counseling centers, or the financial incentives offered to individuals who accept to travel back to their country of citizenship (given the relative lack of choice people face, some scholars prefer the term ‘self-deportation’)...‘

#AbolishFrontex

https://algorithmwatch.org/en/frontex-is-building-an-ai-chatbot-app-to-encourage-repatriations/

#Migration #Refugees #Chatbot #AI #Surveilance #Technology #Europe #FortressEurope #Frontex

  • Copy link
  • Flag this post
  • Block
Ben Ramsey
Ben Ramsey
@ramsey@phpc.social  ·  activity timestamp 2 days ago

Quote du jour: “‘Top-down mandates to use large language models are crazy,’ one employee told Wired. ‘If the tool were good, we’d all just use it.’”

https://futurism.com/artificial-intelligence/jack-dorsey-block-falling-apart-ai

#AI

Futurism

Jack Dorsey's New Company Falling Apart as It Forces Employees to Use AI

Twitter founder Jack Dorsey is running into some serious issues while overhauling his financial services company, Block.
  • Copy link
  • Flag this post
  • Block
F-Droid
F-Droid
@fdroidorg@floss.social  ·  activity timestamp 2 days ago

This week in #FDroid (TWIF) is live:

* Client banners, did you forget something?
* Basic 2.0-alpha3, test test test!
* Google Push w/o proprietary libs
* #DolphinEmulator huge updates and stories
* #ImageToolbox gets #AI
* #Luanti #Minetest texture fix
* #Nextcloud Hub 26 Winter upgrade
* #ProtonVPN w/o #OpenVPN
+ one new app
& 291 updates
- five archived

Where were you during the 2026 Android freedom wars? Get involved: https://f-droid.org/2026/02/20/twif.html

devSJR :python: :rstats:
devSJR :python: :rstats:
@devSJR@fosstodon.org replied  ·  activity timestamp 2 days ago

@fdroidorg #ImageToolbox gets #AI is correct. It is local and works if the phone is not providing compute, this new feature is not fun to use. However, #DeJPEG, #upscale, #colorization, #enhancement is now available. Background removing offers also more models for the #FLOSS version.

  • Copy link
  • Flag this comment
  • Block

Bonfire social

This is a bonfire demo instance for testing purposes

Bonfire social: About · Code of conduct · Privacy · Users · Instances
Bonfire social · 1.0.1 no JS en
Automatic federation enabled
Log in Create account
  • Explore
  • About
  • Members
  • Code of Conduct