fanf42 (@fanf42@social.treehouse.systems)
@alexelcu
I happen to have been on Twitter lately, several times. Once, I looked for what people were posting.
Holy shit. I think you're almost right (it's a big AI generated world), but clearly they missed the very last part of the prompt...
Show Original Post
PetitPixel29 (@PetitPixel29@social.sciences.re)
đ§âđ» #IA / #Hallucinations
â« #ChatGPT : la proposition dâ #OpenAI pour Ă©viter les hallucinations pourrait tuer son propre chatbot » @france.theconversation.com â Faire Ă©valuer Ă lâIA la confiance quâil attribue Ă sa rĂ©ponse ? Un principe incompatible avec lâ expĂ©rience utilisateur #UX qui pourrait signer lâarrĂȘt de mort des agents conversationnels
Show Original Post
remixtures (@remixtures@tldr.nettime.org)
"AI search tools and chatbots are already known to fabricate links and references. But answers based on the material from actual papers can mislead as well if those papers have been retracted. The chatbot is âusing a real paper, real material, to tell you something,â says Weikuan Gu, a medical researcher at the University of Tennessee in Memphis and an author of one of the recent studies. But, he says, if people only look at the content of the answer and do not click through to the paper and see that itâs been retracted, thatâs really a problem.
Gu and his team asked OpenAIâs ChatGPT, running on the GPT-4o model, questions based on information from 21 retracted papers on medical imaging. The chatbotâs answers referenced retracted papers in five cases but advised caution in only three. While it cited non-retracted papers for other questions, the authors note it may not have recognized the retraction status of the articles. In a study from August, a different group of researchers used ChatGPT-4o mini to evaluate the quality of 217 retracted and low-quality papers from different scientific fields; they found that none of the chatbotâs responses mentioned retractions or other concerns. (No similar studies have been released on GPT-5, which came out this August.)
(...)
The problem is not limited to ChatGPT. In June, MIT Technology Review tested AI tools specifically advertised for research work, such as Elicit, Ai2 ScholarQA (now part of the Allen Institute for Artificial Intelligenceâs Asta tool), Perplexity, and Consensus, using questions based on the 21 retracted papers in Guâs study. Elicit referenced five of the retracted papers in its answers, while Ai2 ScholarQA referenced 17, Perplexity 11, and Consensus 18âall without noting the retractions."
#AI #GenerativeAI #ChatGPT #Health #Science
Show Original Post
cyberkendra (@cyberkendra@techhub.social)
Microsoft just made a huge move in the AI space! They're now letting businesses choose between different AI models (OpenAI vs. Anthropic's Claude) in their Copilot tools. Think of it like having multiple search engines instead of being stuck with just one. This gives companies more options and better tools for research and automation. Competition is heating up! đ„
Read Story - https://www.cyberkendra.com/2025/09/microsoft-365-copilot-adds-claude-ai.html
#AI #microsoft #copilot #chatgpt #aitrends
Show Original Post
itnewsbot (@itnewsbot@schleuss.online)
Why does OpenAI need six giant data centers? - On Tuesday, OpenAI, Oracle, and SoftBank announced plans for... - https://arstechnica.com/ai/2025/09/why-does-openai-need-six-giant-data-centers/ #aidevelopmenttools #aiinfrastructure #machinelearning #claymagouyrk #masayoshison #datacenters #coreweave #samaltman #sbenergy #softbank #aichips #chatgpt #bizâą #nvidia #openai #oracle #ai
Show Original Post
jeffhorton (@jeffhorton@mstdn.ca)
I put "Dictated but not read" in my last poll and notice that https://en.wikipedia.org/wiki/Dictated_but_not_read wonders if it is still needed.
I'm thinking it is still valid but also we need a new version like
"Chatted but not read"
To indicate that you just passed along llm drivel as your words without any review either.
Show Original Post
dashremover (@dashremover@mastodon.social)
PR guy: "Let's make the AI sound more human."
AI: "We're thrilled to embark on this incredible journey of innovation, empathy, and unprecedented synergy."
Me: hey who let LinkedIn write the robot
#AI #ChatGPT #TechHumor đ€đ«
Show Original Post
cyberveille (@cyberveille@mastobot.ping.moi)
đą ShadowLeak : exfiltration zeroâclick cĂŽtĂ© service via lâagent Deep Research de ChatGPT
đ Selon Radware (radware.com), des chercheurs ont mis au jour **ShadowLeak**, une **attaque zeroâclick** exploitant lâagent **Deep Research de ChatG...
đ cyberveille : https://cyberveille.ch/posts/2025-09-24-shadowleak-exfiltration-zero-click-cote-service-via-lagent-deep-research-de-chatgpt/
đ source : https://www.radware.com/blog/threat-intelligence/shadowleak/
#ChatGPT #IOC #Cyberveille
Show Original Post
shish (@shish@techhub.social)
#ChatGPT has emerged as a top driver of referral traffic to #retailersâ websites, giving companies like Walmart, Target, Etsy and eBay a fresh source of online #shoppers
Show Original Post
arstechnica (@arstechnica@c.im)
Why does OpenAI need six giant data centers? https://arstechni.ca/VuhX #AIdevelopmenttools #AIinfrastructure #machinelearning #ClayMagouyrk #MasayoshiSon #datacenters #CoreWeave #samaltman #SBEnergy #SoftBank #AIchips #ChatGPT #Biz&IT #NVIDIA #openai #oracle #AI
Show Original Post
ppasseri (@ppasseri@infosec.exchange)
The Netskope Threat Labs Report for #Retail 2025 is out! đą
âïž 95% of organizations use #genAI apps
âïž 57% of DLP violations are for regulated data
âïž 81% #chatGPT is the most used genAI app
âïž 11% of #malware downloads come from OneDrive
https://www.netskope.com/resources/threat-labs-reports/threat-labs-report-retail-2025
Show Original Post
cryptadamist (@cryptadamist@universeodon.com)
Bill Simmons made a clear statement supporting #JimmyKimmel and calling out government censorship as "stupid shit".
now the #AI summary claims he said the opposite.
we're reaching the singularity sooner than i thought.
#Google #Gemini #chatgpt #grok #uspol #uspolitics #kimmel #1stamendment #freespeech
Show Original Post