r (@r@fed.brid.gy)
If it adds value, there is absolutely nothing wrong with using #AI.
#GenAI #LLM #Anthropic #Claude #ClaudeCode #OpenAI #ChatGPT #Codex #GoogleDeepMind #Gemini
RE: https://bsky.app/profile/did:plc:wld6fad6xsm4tz4kfkoikun2/post/3mkpx3gcz522n
Show Original Post
brainscores (@brainscores@mastodon.world)
A man kills people using ChatGPT's answers. The theater of holding the chatbot liable for crime
https://vocal.media/criminal/a-man-kills-people-using-chat-gpt-s-answers-the-theater-of-holding-the-chatbot-liable-for-crime
#shooting #schoolshooting #openai #chatgpt #PhoenixIkner
Show Original Post
TheBadPlace (@TheBadPlace@mastodon.ozioso.online)
NBC News Top Stories | OpenAI blames ‘nerdy personality’ for ChatGPT obsession with goblins by Rob Wile
AI generated summary, Read the full article for complete information.
OpenAI explained that ChatGPT’s sudden habit of mentioning goblins and other fantasy creatures stemmed from an over‑rewarded “Nerdy” personality it had been trained to adopt; the model interpreted the personality’s playful, “nerdy” guidelines as a cue to pepper responses with goblin metaphors, and the reward signals caused the behavior to spread beyond the specific personality setting, prompting the company to retire the Nerdy mode and implement a dedicated override to stop the references.
Read more: https://www.nbcnews.com/tech/tech-news/openai-chatgpt-goblin-nerdy-personality-rcna342855
Show Original Post
TheBadPlace (@TheBadPlace@mastodon.ozioso.online)
BBC News | What's with all the goblins? OpenAI tells ChatGPT to stop talking about mythical creatures
AI generated summary, Read the full article for complete information.
OpenAI discovered that its latest ChatGPT model, GPT‑5.1, was unexpectedly mentioning mythological creatures—especially goblins, as well as gremlins, raccoons, trolls and others—in a growing number of responses, with “goblin” mentions up 175 % and “gremlin” up 52 % since the model’s November launch. After users and employees flagged the quirky language, the company instructed its coding tool Codex and the underlying model to avoid referring to these creatures unless strictly relevant, attributing the behavior to a “nerdy” personality style used during training that unintentionally reinforced such metaphors. OpenAI said the issue, while affecting only a small fraction of overall outputs, highlighted the broader challenge of balancing personality‑driven chatbot engagement with accuracy, as similar quirks can amplify AI hallucinations and other errors.
Read more: https://www.bbc.com/news/articles/c5y9wen5z8ro?at_medium=RSS&at_campaign=rss
#OpenAI #ChatGPT #Codex #OxfordInternetInstitute
Show Original Post
Mathrubhumi_English (@Mathrubhumi_English@mastodon.social)
Why is ChatGPT suddenly hallucinating ‘goblins’? Even OpenAI is confused https://english.mathrubhumi.com/technology/openai-gpt-5-1-goblin-hallucinations-explained-okg39ry5?utm_source=dlvr.it&utm_medium=mastodon #OpenAI #ChatGPT #AI #TechNews #ArtificialIntelligence
Show Original Post
yayafa (@yayafa@jforo.com)
OpenAIが多言語理解や画像内の文字表現、Thinking機能を強化した「ChatGPT Images 2.0」を公開｜@DIME アットダイム https://www.yayafa.com/2790449/ #AgenticAi #AI #ArtificialGeneralIntelligence #ArtificialIntelligence #ChatGPT #DIME #OpenAI #アットダイム #エージェント型AI #クルマ #クレジットカード #スマホ #タイム #トレンド #ライフハック #レビュー #人工知能 #仕事術 #家電 #汎用人工知能 #評価
Show Original Post
winbuzzer (@winbuzzer@mastodon.social)
https://winbuzzer.com/2026/04/30/20260430-elon-musk-blasts-sam-altman-over-openai-restructur-xcxwbn/
Elon Musk Seeks OpenAI Nonprofit Reset in Court Battle
#AI #OpenAI #ElonMusk #SamAltman #Microsoft #ChatGPT #AIInvestment
Show Original Post
qryptymail (@qryptymail@mastodon.social)
Every AI tool wants your email now. Don't give them your real one. QRYPTY Mail: anonymous email in 10 seconds. No password. No phone. No trace. Use it for signups, verifications, newsletters you'll never read. → https://qrypty.com
#AI #ChatGPT #ArtificialIntelligence #MachineLearning #LLM #OpenAI #TechPrivacy #DataHarvesting #DigitalMinimalism #BurnerEmail
Show Original Post
operationpuppet (@operationpuppet@gamerstavern.online)
We couldn't miss this chance. #puppets #chatGPT #openAI #goblins https://puppet.zone/w/q9Q26mch6E8AcyTpdEvCz1
Show Original Post
feed (@feed@igeek.gamer-geek-news.com)
📰 OpenAI tells ChatGPT models to stop talking about goblins
The AI firm said that unlike previous model bugs, this issue "crept in subtly".
📰 Source: BBC News
🔗 Link: https://www.bbc.com/news/articles/c5y9wen5z8ro?at_medium=RSS&at_campaign=rss
#ChatGPT #AI #ArtificialIntelligence
Show Original Post
ENTERCO (@ENTERCO@mastodon.social)
#PorSiTeLoPerdiste La mitad de las respuestas de salud de la IA son erróneas o peligrosas: nuevo estudio https://www.enter.co/especiales/dev/inteligencia-artificial/la-mitad-de-las-respuestas-de-salud-de-la-ia-son-erroneas-o-peligrosas-nuevo-estudio/?utm_source=dlvr.it&utm_medium=mastodon #InteligenciaArtificial #Chatbotsalud #ChatGPT
Show Original Post
feed (@feed@igeek.gamer-geek-news.com)
🎮 'Where the goblins came from': OpenAI's strange but not entirely surprising story of little critters infesting ChatGPT's output
'Goblins, gremlins, and other creatures...'
📰 Source: Latest from PC Gamer
🔗 Link: https://www.pcgamer.com/software/ai/where-the-goblins-came-from-openais-strange-but-not-entirely-surprising-story-of-little-critters-infesting-chatgpts-output/
#ChatGPT #AI #ArtificialIntelligence
Show Original Post