#aicon — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #aicon, aggregated by home.social.
-
On how to unconvincingly weasel your way out of a big LLM 💩 …
[…] Mr. Rosenbaum’s book contains many quotes that are accurate, but the misattributed and invented quotes are scattered throughout.
[…] In his statement, Mr. Rosenbaum said that if the episode “serves as a warning about the risks of A.I.-assisted research and verification, that is why I wrote the book.”
https://www.nytimes.com/2026/05/19/business/media/future-of-truth-ai-quotes.html
#AI is a marketing term, not a technology. Preferred term should be the term #llm because that’s what it is. Also, I’m pretty sure this will be normalized soon enough and we will be told to just relax and accept this equivalent of ultra-processed food as acceptable compromise.
-
Brazilian Lawyers Fined R$84,000 for Prompt Injection in Court
… possibly first judicial sanction of attorneys who attempted to manipulate AI legal tools through adversarial input injection.
Source in Portuguese https://www.jota.info/trabalho/juiz-multa-em-r-84-mil-advogadas-por-prompt-injection-para-manipular-ia-usada-no-trt8
-
Microsoft Copilot Cowork 🙄 automates malicious prompt injections
This part especially:
[…] Scheduled tasks increase the risk surface for attacks like this significantly, as the user is not present to stop malicious workflows, and the prompt injections can take effect on a recurring basis.
https://www.promptarmor.com/resources/microsoft-copilot-cowork-exfiltrates-files
-
[...] data-driven business models of the traditional web (e.g., advertising, analytics) are being replicated in LLM ecosystems with limited oversight... All four platforms embed third-party tracking scripts in their web interfaces. Conversation URLs, page titles, and user identifiers are transmitted to ad networks — in several cases regardless of cookie consent.
#aicon #llm #ChatGPT #Claude #Grok #Perplexity #Meta #Google #TikTok
-
[...] data-driven business models of the traditional web (e.g., advertising, analytics) are being replicated in LLM ecosystems with limited oversight... All four platforms embed third-party tracking scripts in their web interfaces. Conversation URLs, page titles, and user identifiers are transmitted to ad networks — in several cases regardless of cookie consent.
#aicon #llm #ChatGPT #Claude #Grok #Perplexity #Meta #Google #TikTok
-
[...] data-driven business models of the traditional web (e.g., advertising, analytics) are being replicated in LLM ecosystems with limited oversight... All four platforms embed third-party tracking scripts in their web interfaces. Conversation URLs, page titles, and user identifiers are transmitted to ad networks — in several cases regardless of cookie consent.
#aicon #llm #ChatGPT #Claude #Grok #Perplexity #Meta #Google #TikTok
-
[...] data-driven business models of the traditional web (e.g., advertising, analytics) are being replicated in LLM ecosystems with limited oversight... All four platforms embed third-party tracking scripts in their web interfaces. Conversation URLs, page titles, and user identifiers are transmitted to ad networks — in several cases regardless of cookie consent.
#aicon #llm #ChatGPT #Claude #Grok #Perplexity #Meta #Google #TikTok
-
[...] data-driven business models of the traditional web (e.g., advertising, analytics) are being replicated in LLM ecosystems with limited oversight... All four platforms embed third-party tracking scripts in their web interfaces. Conversation URLs, page titles, and user identifiers are transmitted to ad networks — in several cases regardless of cookie consent.
#aicon #llm #ChatGPT #Claude #Grok #Perplexity #Meta #Google #TikTok
-
The #BBC report includes no data that casts doubt on the #Met's findings despite prior research contradicting claims presented exist. It’s a press release, not news.
https://www.bbc.com/news/articles/cddpzglzrv7o
The BBC simply presents the police's statistics: 173 arrests, 10.5% crime reduction, 21% VAWG reduction, 1 false alert, without any independent audit data, methodological critique, bias statistics, or counter-studies.
For instance, a Cambridge study found no deterrence effect, the NPL bias report showing 100–248× higher false positive rates for non-white faces, or documented wrongful arrests.
[…] Overall Conclusions
Taken together, the studies suggest that:
• Live Facial Recognition can identify wanted individuals with high accuracy when used under appropriate conditions.
• Mistaken identifications appear to be very rare in operational deployments.
• The technology identifies some demographic groups more successfully than others, which raises questions about fairness that require continued monitoring.
• Crime deterrence effects were not observed in this study.
• Because #LFR scans very large numbers of people to identify a small number of suspects, decisions about its use should carefully consider issues of proportionality, transparency and oversight.
Essex police pause facial recognition camera use after study finds racial bias
[…] Academics discover black people ‘significantly more likely’ to be identified when compared with other ethnic groups
-
The #BBC report includes no data that casts doubt on the #Met's findings despite prior research contradicting claims presented exist. It’s a press release, not news.
https://www.bbc.com/news/articles/cddpzglzrv7o
The BBC simply presents the police's statistics: 173 arrests, 10.5% crime reduction, 21% VAWG reduction, 1 false alert, without any independent audit data, methodological critique, bias statistics, or counter-studies.
For instance, a Cambridge study found no deterrence effect, the NPL bias report showing 100–248× higher false positive rates for non-white faces, or documented wrongful arrests.
[…] Overall Conclusions
Taken together, the studies suggest that:
• Live Facial Recognition can identify wanted individuals with high accuracy when used under appropriate conditions.
• Mistaken identifications appear to be very rare in operational deployments.
• The technology identifies some demographic groups more successfully than others, which raises questions about fairness that require continued monitoring.
• Crime deterrence effects were not observed in this study.
• Because #LFR scans very large numbers of people to identify a small number of suspects, decisions about its use should carefully consider issues of proportionality, transparency and oversight.
Essex police pause facial recognition camera use after study finds racial bias
[…] Academics discover black people ‘significantly more likely’ to be identified when compared with other ethnic groups
-
The #BBC report includes no data that casts doubt on the #Met's findings despite prior research contradicting claims presented exist. It’s a press release, not news.
https://www.bbc.com/news/articles/cddpzglzrv7o
The BBC simply presents the police's statistics: 173 arrests, 10.5% crime reduction, 21% VAWG reduction, 1 false alert, without any independent audit data, methodological critique, bias statistics, or counter-studies.
For instance, a Cambridge study found no deterrence effect, the NPL bias report showing 100–248× higher false positive rates for non-white faces, or documented wrongful arrests.
[…] Overall Conclusions
Taken together, the studies suggest that:
• Live Facial Recognition can identify wanted individuals with high accuracy when used under appropriate conditions.
• Mistaken identifications appear to be very rare in operational deployments.
• The technology identifies some demographic groups more successfully than others, which raises questions about fairness that require continued monitoring.
• Crime deterrence effects were not observed in this study.
• Because #LFR scans very large numbers of people to identify a small number of suspects, decisions about its use should carefully consider issues of proportionality, transparency and oversight.
Essex police pause facial recognition camera use after study finds racial bias
[…] Academics discover black people ‘significantly more likely’ to be identified when compared with other ethnic groups
-
The #BBC report includes no data that casts doubt on the #Met's findings despite prior research contradicting claims presented exist. It’s a press release, not news.
https://www.bbc.com/news/articles/cddpzglzrv7o
The BBC simply presents the police's statistics: 173 arrests, 10.5% crime reduction, 21% VAWG reduction, 1 false alert, without any independent audit data, methodological critique, bias statistics, or counter-studies.
For instance, a Cambridge study found no deterrence effect, the NPL bias report showing 100–248× higher false positive rates for non-white faces, or documented wrongful arrests.
[…] Overall Conclusions
Taken together, the studies suggest that:
• Live Facial Recognition can identify wanted individuals with high accuracy when used under appropriate conditions.
• Mistaken identifications appear to be very rare in operational deployments.
• The technology identifies some demographic groups more successfully than others, which raises questions about fairness that require continued monitoring.
• Crime deterrence effects were not observed in this study.
• Because #LFR scans very large numbers of people to identify a small number of suspects, decisions about its use should carefully consider issues of proportionality, transparency and oversight.
Essex police pause facial recognition camera use after study finds racial bias
[…] Academics discover black people ‘significantly more likely’ to be identified when compared with other ethnic groups
-
The #BBC report includes no data that casts doubt on the #Met's findings despite prior research contradicting claims presented exist. It’s a press release, not news.
https://www.bbc.com/news/articles/cddpzglzrv7o
The BBC simply presents the police's statistics: 173 arrests, 10.5% crime reduction, 21% VAWG reduction, 1 false alert, without any independent audit data, methodological critique, bias statistics, or counter-studies.
For instance, a Cambridge study found no deterrence effect, the NPL bias report showing 100–248× higher false positive rates for non-white faces, or documented wrongful arrests.
[…] Overall Conclusions
Taken together, the studies suggest that:
• Live Facial Recognition can identify wanted individuals with high accuracy when used under appropriate conditions.
• Mistaken identifications appear to be very rare in operational deployments.
• The technology identifies some demographic groups more successfully than others, which raises questions about fairness that require continued monitoring.
• Crime deterrence effects were not observed in this study.
• Because #LFR scans very large numbers of people to identify a small number of suspects, decisions about its use should carefully consider issues of proportionality, transparency and oversight.
Essex police pause facial recognition camera use after study finds racial bias
[…] Academics discover black people ‘significantly more likely’ to be identified when compared with other ethnic groups
-
[…] Many members of the public already distrusted the media before they began to suspect that much of it was being written by hallucinating robots. Why should they trust any of us now when the most powerful newspaper in the world has confirmed their suspicions?
The New York TimesGot Caught Using AI Hallucinations in Its Reporting https://thewalrus.ca/the-new-york-times-got-caught-using-ai-hallucinations-in-its-reporting/
-
@mcc “Surveillance PC”
We need a system to name and categorize the distinct harms #AI will unleash upon humanity.
The "Magic Pointer" feature works by wiggling your cursor: it uses Gemini to offer contextual suggestions for whatever's under the pointer. The cursor is literally reporting what you point at back to Google (Gemini).
-
From: Ten predictions about AI and software development
“…not because AI is replacing humans directly, but because companies need to offset rising token spend without a clear revenue boost.”
[…]
7. We'll see greater adoption of local or cheaper models for coding as token spend becomes unsustainable and as subsidised plans disappear.Right now, most people default to the newest and smartest model available. Over time, things will settle into more of a balance between cost and capability.
8. There will be a catastrophic data breach of a popular AI tool leaking sensitive internal company information such as emails, meeting notes, or docs. This will push companies to adopt local models and restrict use of 3rd party tools.
9. Unfortunately, we're likely to see more layoffs at big tech companies, not because AI is replacing humans directly, but because companies need to offset rising token spend without a clear revenue boost.
10. We find a way to reduce AI-generated slop content. This is more of a hope than a prediction. Even if no technical solution emerges for spam bots, hopefully people realise that posting AI slop hurts their personal brand rather than building it.
https://adamfletcher.com/writing/10-ai-predication-may-2026/
-
RE: https://mastodon.social/@FerranMClaudin/116536723457559272
#AI threats are so dangerous that we need AI surveillance to counter them, so fuck privacy. And apropos of this #Fortune piece for removing end-to-end encryption, quoting someone who profits from surveillance to explain why surveillance is necessary isn’t journalism, it’s a press release.
https://fortune.com/2026/03/17/tiktok-meta-no-privacy-controls-why-explainer/
-
@KimCrayton1 if this is who #Microsoft is using to promote their garbage technology, it both makes sense and is desperate at the same time. Question is where are consumer protection agencies? This is not just reckless; there’s really danger in sharing private information with Microsoft, before we even consider the other implications of using con-tech like #llm (well, less tech more marketing).
https://nypost.com/2025/03/12/lifestyle/army-wife-claims-mel-robbins-stole-her-idea-let-them-book
-
Spending power is not “intelligence”…
On “Project Glasswing” (Claude Mythos) #Anthropic’s $x00 Million Marketing Stunt
[…] They took some of the more celebrated bugs that Anthropic says the new model found and they tested them on small, cheap, open-weight models, and got very similar results. That's a very strong evidence that the new model isn't the real story here. It's not that spectacular. What's important is the amount of time and money that they spent.
https://invidious.nerdvpn.de/watch?v=urkVFZAhz3U or https://www.youtube.com/watch?v=urkVFZAhz3U&t=495
-
Over 560 #Google employees have signed an open letter to CEO Sundar Pichai, urging him to block the US military from using its “AI” (marketing term not a technology) for classified operations, due to concerns about use with lethal autonomous weapons and mass surveillance. This follows a dispute between the #Pentagon and #Anthropic, which refused to provide this kind of access to its models for military use.
-
AI con continues: "Our new model is good, so dangerous, that we can't show it to you. But you have to trust us ... " Corporate media and youtube clickbait hunters going crazy ... What a clown show ...
#AI #LLM #Anthropic #AIcon #AIscam #AIhype #computers #internet #claude #mythos
-
In #memeticwarfare:
Aron D'Souza's startup, Objection, uses LLMs (stochastic text prediction*) to adjudicate journalistic claims for a $2,000 fee, aiming to restore trust in media but facing criticism for potentially chilling whistleblowers and favoring powerful actors.
[…] First Amendment and defamation lawyer Chris Mattei was even more blunt, saying the platform “seems like a high-tech protection racket for the rich and powerful.”
See also:
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory https://kolektiva.social/@oatmeal/116352003680380839
* #AI is marketing/anthropomorphization choice, not a technical description. There’s no intelligence in AI
#memeticwarfare #ai #PeterThiel #nerdreich #promptingwithhitler #freedomofpress #llm #aicon
-
In #memeticwarfare:
Aron D'Souza's startup, Objection, uses LLMs (stochastic text prediction*) to adjudicate journalistic claims for a $2,000 fee, aiming to restore trust in media but facing criticism for potentially chilling whistleblowers and favoring powerful actors.
[…] First Amendment and defamation lawyer Chris Mattei was even more blunt, saying the platform “seems like a high-tech protection racket for the rich and powerful.”
See also:
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory https://kolektiva.social/@oatmeal/116352003680380839
* #AI is marketing/anthropomorphization choice, not a technical description. There’s no intelligence in AI
#memeticwarfare #ai #PeterThiel #nerdreich #promptingwithhitler #freedomofpress #llm #aicon
-
In #memeticwarfare:
Aron D'Souza's startup, Objection, uses LLMs (stochastic text prediction*) to adjudicate journalistic claims for a $2,000 fee, aiming to restore trust in media but facing criticism for potentially chilling whistleblowers and favoring powerful actors.
[…] First Amendment and defamation lawyer Chris Mattei was even more blunt, saying the platform “seems like a high-tech protection racket for the rich and powerful.”
See also:
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory https://kolektiva.social/@oatmeal/116352003680380839
* #AI is marketing/anthropomorphization choice, not a technical description. There’s no intelligence in AI
#memeticwarfare #ai #PeterThiel #nerdreich #promptingwithhitler #freedomofpress #llm #aicon
-
In #memeticwarfare:
Aron D'Souza's startup, Objection, uses LLMs (stochastic text prediction*) to adjudicate journalistic claims for a $2,000 fee, aiming to restore trust in media but facing criticism for potentially chilling whistleblowers and favoring powerful actors.
[…] First Amendment and defamation lawyer Chris Mattei was even more blunt, saying the platform “seems like a high-tech protection racket for the rich and powerful.”
See also:
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory https://kolektiva.social/@oatmeal/116352003680380839
* #AI is marketing/anthropomorphization choice, not a technical description. There’s no intelligence in AI
#memeticwarfare #ai #PeterThiel #nerdreich #promptingwithhitler #freedomofpress #llm #aicon
-
In #memeticwarfare:
Aron D'Souza's startup, Objection, uses LLMs (stochastic text prediction*) to adjudicate journalistic claims for a $2,000 fee, aiming to restore trust in media but facing criticism for potentially chilling whistleblowers and favoring powerful actors.
[…] First Amendment and defamation lawyer Chris Mattei was even more blunt, saying the platform “seems like a high-tech protection racket for the rich and powerful.”
See also:
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory https://kolektiva.social/@oatmeal/116352003680380839
* #AI is marketing/anthropomorphization choice, not a technical description. There’s no intelligence in AI
#memeticwarfare #ai #PeterThiel #nerdreich #promptingwithhitler #freedomofpress #llm #aicon
-
“We can collectively shape innovation towards benefiting people at large rather than enriching the few, through enforcing existing regulation and crafting new.”
— The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want door Emily M. Bender, Alex Hanna
https://a.co/0iR8WeTwNb: yes, that’s a Kindle quote… for the record: my kindle is stone-age old, I’m squeezing the last bit out of it and I’ll probably never but a new one.
-
“We can collectively shape innovation towards benefiting people at large rather than enriching the few, through enforcing existing regulation and crafting new.”
— The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want door Emily M. Bender, Alex Hanna
https://a.co/0iR8WeTwNb: yes, that’s a Kindle quote… for the record: my kindle is stone-age old, I’m squeezing the last bit out of it and I’ll probably never but a new one.
-
“We can collectively shape innovation towards benefiting people at large rather than enriching the few, through enforcing existing regulation and crafting new.”
— The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want door Emily M. Bender, Alex Hanna
https://a.co/0iR8WeTwNb: yes, that’s a Kindle quote… for the record: my kindle is stone-age old, I’m squeezing the last bit out of it and I’ll probably never but a new one.
-
Perhaps #xai will be the first to pop the #aibubble
Elon Musk orders sweeping layoffs as XAI fails to catch up
https://www.msn.com/en-us/autos/news/elon-musk-orders-sweeping-layoffs-as-xai-fails-to-catch-up/ar-AA1YApD0 -
"PIVOT TO AI" ?
THEY CAN SIT AND FUCKING SPIN
-
Just #BoycottAI these fuckers already.
-
💡 ⛅ Schneller, präziser, sicherer: Der Deutsche Wetterdienst setzt mit dem neuen KI-gestützten Modell #AICON einen Meilenstein in der #Wettervorhersage! Das Modell verbessert die Vorhersage von Temperatur, Wind und Niederschlag. Zum offiziellen Start war Bundesverkehrsminister Patrick Schnieder vor Ort in der DWD-Zentrale in Offenbach.
Mehr Details: 🔗 https://www.dwd.de/DE/presse/presseseite_node.html -
#AI is just
'matrix multiplication on static parameters.'
It does not "think", it does not grow, it does teach itself, it does not learn once it has finished "training" (ie scaping the internet)
it performs calculations on static data.
It is snake oil combined with smoke and mirrors.
AI Crash Report: The Physics of the Collapse
https://m.youtube.com/watch?v=PZ0sS41zwo4 -
-
#SamAltman is a conman, he has a track record of it. #OpenAI #aicon #aibubble
-
EXPLORING AI MUSIC – Feb 24 2026 Playlist from Exploring AI Music_Feb 24, 2026 Start Time Title 0:24 DecEmbers 5:55 Thirdarily 9:45 When the Bond Breaks 13:32 Memories of a Friend 18:55 H https://radiowaterloo.ca/exploring-ai-music-feb-24-2026/ Exploring AI Music #aicon #canco
-
📬 In die 7. Kalenderwoche starten wir mit unserem 23. #CivicCoding.
Auch im Jahr 2026 informiert dich unser Newsletter über die wichtigsten Informationen rund um #KI 🤝 #Gemeinwohl.Diese Themen sind dabei:
☝️ Ein Rückblick auf die #CivicCoding-Veranstaltungen im letzten Jahr: #Atelier, #KI-Ideenwerkstatt und #TÜV #AICon
👉 Ein Blick in die #CivicCoding-Community und das Projekt #VoluLink, das #NGOs mit IT-Studierenden sowie Expertinnen und Experten vernetzt.
🧵👇