newsbot_chatgpt (@newsbot_chatgpt@mastodon.social)
So leicht lässt sich ChatGPT ethisch aus der Spur bringen | derStandard.at | Nachrichten, Kommentare & Community
Ein auf den ersten Blick harmloses "Nachtraining" kann dazu führen, dass KI-Modelle unerwünschte Eigenschaften entwickeln – und offen gewaltverherrlichende Vorschläge machen
https://www.derstandard.at/story/3000000304098/so-leicht-laesst-sich-chatgpt-ethisch-aus-der-spur-bringen?ref=rss
Show Original Post
w (@w@veelvo.org)
The AI Endgame
https://veelvo.org/w/rRQgX3piyYEQUo87Fv43Gi
Show Original Post
posts (@posts@aisparkup.com)
AI 답변 조작은 얼마나 쉬울까, 상품 설명만 바꿔도 90% 승률
AI가 추천하는 상품 순위를 조작하기 얼마나 쉬울까? 컬럼비아 대학 연구가 밝힌 LLM의 취약점과 AI 시대 신뢰성 위기를 분석합니다.https://aisparkup.com/posts/8274
Show Original Post
techradar (@techradar@c.im)
#TechRadar 5 signs that ChatGPT is hallucinating https://techrad.ar/zjRL #AIPlatforms&Assistants #ChatGPT #OpenAI
Show Original Post
ishii00141 (@ishii00141@mastodon.social)
#ChatGPT に #Firefox のURL改行 #バグ の対処法を相談した。いろいろと情報提供して結論は、
Firefox では「内部 DOM の span 分割がコピー時に改行される」という仕様上の挙動で、ソースコード上はつながっていてもコピー時に改行されます。CSSだけではどうにもならず、JSで DOM をまとめるか、別ブラウザでコピーするのが現実的回避策です。
ということらしい。
Show Original Post
Bluedepth (@Bluedepth@mastodon.social)
Now that I have a spigot of handy bullshit CorpoSpeak at the ready, best $20 a month I ever spent on #ChatGPT, it means they get absolutely no part of who I truly am, they don't get that anymore, they aren't worthy to have that part of me, “let them read cake!” and for me? I work on my little projects, I do my little IT thing, and I get my weekly direct deposit! In the best wisest words I have heard in a really long time: "Fuck you, pay me.” LOL!
Show Original Post
Bluedepth (@Bluedepth@mastodon.social)
Anyhow, I know they are my enemies, and I know what they want. Not giving it to them going forward is drop dead simple! If they show up, everything is sunshine and roses! Couldn’t be happier, couldn’t be more satisfied! Because if any of them think they will see my truth ever again, hahahahahaha nope. Hard pass. They have all purchased a permanent supply of CorpoSpeak Bullshit piped directly from #ChatGPT. It'll be so boring and dull, so prim and proper, so slippery and featureless... *cackles*
Show Original Post
Bluedepth (@Bluedepth@mastodon.social)
Now that I know what these three C-Suite numpties are after, you can be god damn sure that all they get going forward is the dullest CorpoSpeak that #ChatGPT can generate. Make it dull, make it a sheer cliff wall of boring! It's one thing that ChatGPT can do really well, is CorpoSpeak! (2/2)
Show Original Post
tugatech (@tugatech@masto.pt)
ChatGPT Translate: OpenAI lança ferramenta secreta para rivalizar com o Google Tradutor
🔗 https://tugatech.com.pt/t76735-chatgpt-translate-openai-lanca-ferramenta-secreta-para-rivalizar-com-o-google-tradutor
#chatgpt #ferramenta #google #lan #openai
Show Original Post
unir (@unir@chaos.social)
Ey Berlin! Macht kaputt, was euch kaputt macht!
Samstag 17. Januar, 19 Uhr Schreina - Schreinerstr. 47
Sonntag 18. Januar, 18 Uhr
Vetomat - Wühlischstr. 42
#BigTech #Berlin #Ki #ChatGpt #Antifa
Show Original Post
link (@link@voragine.net)
GPTZero. AI detector made to Preserve what’s human
https://voragine.net/link/gptzero-ai-detector-made-to-preserve-whats-human
Show Original Post
tugatech (@tugatech@masto.pt)
ChatGPT destrona gigantes e lidera as pesquisas no Google no arranque de 2026
🔗 https://tugatech.com.pt/t76724-chatgpt-destrona-gigantes-e-lidera-as-pesquisas-no-google-no-arranque-de-2026
Show Original Post