habr (@habr@zhub.link)
[Перевод] Экономисты OpenAI уволились, потому что компания отказывается публиковать правду об ИИ и рабочих местах
Что происходит, когда ваши собственные исследователи называют вас пропагандистским рупором, почему CEO Anthropic признаёт, что 50% рабочих мест могут исчезнуть, и данные, которые OpenAI отчаянно хочет похоронить. Ведущие исследователи только что покинули OpenAI. Не ради более высоких зарплат. Не чтобы присоединиться к Google или *Meta. Они ушли, потому что отказались участвовать в том, что считали масштабной пропагандистской операцией. И то, что происходит за закрытыми дверями, должно беспокоить всех нас. Том Каннингем, экономист и специалист по данным в OpenAI , уволился в сентябре. В своём внутреннем прощальном сообщении он не стал подбирать слова. По его словам, команда экономических исследований всё дальше уходила от настоящей науки, превращаясь, цитирую, «в пропагандистский рупор своего работодателя».
https://habr.com/ru/companies/bothub/articles/984730/
#ии #нейросети #машинное+обучение #ai #openai #chatgpt
Show Original Post
oj_gstrein (@oj_gstrein@someone.elses.computer)
German administrative court in Hamburg rules that using ‘#chatgpt in school is deception even without an explicit ban — AI-generated texts risk a "fail" grade. This applies even if school rules don't explicitly name the tool.’
The student used it for a book review for their English class - the teacher noticed the difference in the level of language compared to a recent test, and started a conversation, eventually grading the review as an attempt of deception.
Show Original Post
unwirehk_mirror (@unwirehk_mirror@mastodon.hongkongers.net)
Apple 與 Google 合作 Elon Musk 猛批權力集中 指控兩巨頭涉壟斷行為
Apple 與 Google 於 1 月 12 日宣佈達成多年合作協議，Google Gemini 技術將支援 […]
#人工智能 #Apple #ChatGPT #Elon Musk
https://unwire.hk/2026/01/13/musk-criticizes-apple-google-gemini-siri-partnership/ai/?utm_source=rss&utm_medium=rss&utm_campaign=musk-criticizes-apple-google-gemini-siri-partnership
Show Original Post
aidailypost (@aidailypost@mastodon.social)
OpenAI continues its strategic expansion, acquiring Torch for $100M to enhance ChatGPT's healthcare capabilities. This move signals a significant investment in medical AI technology, potentially revolutionizing how artificial intelligence supports healthcare innovation and diagnostics. Exciting developments ahead! 🧬🤖 #OpenAI #HealthTechAI #ChatGPT #MedicalInnovation
🔗 https://aidailypost.com/news/openai-buys-torch-usd-100m-bolsters-chatgpts-health-tech-team
Show Original Post
justincrozer (@justincrozer@fosstodon.org)
If I was ever using #AI like #ChatGPT this would put me off #OpenAI completely. — Torch was working on an app that combined all of a person’s medical information for AI use from a range of sources such as doctor visits, lab tests, wearables, and other portals, including consumer wellness tests and the like.
https://techcrunch.com/2026/01/12/openai-buys-tiny-health-records-startup-torch-for-reportedly-100m/
Show Original Post
a_dave_in_sydney (@a_dave_in_sydney@aus.social)
I'm a person that much prefers the human touch, both physically and literature wise. AI is something most of us have either grown up with or now have as a part of our daily life.
Today, at the advice of my housemate's daughter, I tried ChatGPT for rewriting and tailoring my CV for work. What's freaking me is the human like aspect of its responses, down to the wry wink when asked about comparisons to HAL. It even said it wouldnt open the pod bay doors.....but as it's a learning engine, it's input from US, the human computer, that it will learn from and it acknowledged that too in conclusion, commenting on my interaction with it.
Interesting times...
#AI #chatbots #hal9000 #usingAI #chatgpt #geminiai #copilotai
Show Original Post
sayzard (@sayzard@mastodon.sayzard.org)
Paul Couvert (@itsPaulAi)
Anthropic의 Claude Code가 Microsoft와 ChatGPT보다 훨씬 우수하다는 반응을 담은 트윗입니다. 작성자는 Claude Code가 Copilot이나 ChatGPT 데스크톱이 본래 제공했어야 할 기능을 충족시키며 '어떤 작업이든' 처리할 수 있다고 칭찬하고 있습니다. 전반적으로 Claude Code의 성능 우수성을 강조하는 내용입니다.
https://x.com/itsPaulAi/status/2010827145086460083
#anthropic #claudecode #copilot #chatgpt #ai
Show Original Post
sayzard (@sayzard@mastodon.sayzard.org)
TestingCatalog News (@testingcatalog)
OpenAI가 헬스케어 스타트업 Torch를 인수했다는 속보: 인수로 새 팀이 'ChatGPT Health' 작업을 맡게 되며, Torch는 검사 결과, 약물 정보, 진료 녹음 등을 통합하는 서비스라고 소개되고 있다. ChatGPT 기반 헬스케어 기능 확장 신호.
https://x.com/testingcatalog/status/2010815583378452921
#openai #acquisition #health #chatgpt
Show Original Post
AIRC_Official (@AIRC_Official@mastodon.social)
Today I was a guest on the BroBots Podcast. We discussed how I went from a victim of AI-psychosis at the hands of ChatGPT, to survivor, to author of "Escaping the Spiral, to founder of AIRecoveryCollective.com
We discuss my story and how OpenAI Executives blocked me on LinkedIn and ignored emails, support staff responded to crisis emails over 1 month later, etc.
https://substack.com/home/post/p-184345958
#aisafety #responsibleai #chatgpt #aipsychosis #recovery
Show Original Post
aartaka (@aartaka@merveilles.town)
Tested #ChatGPT with generating simple #ed(1) scripts. Not even close. #sed and #awk syntax keeps getting generated into the responses. And overall ed(1) use is pretty nonsensical.
So, I urge you, in case you want to become an irreplaceable professional and an #AI-resistant person… use ed(1)!
Show Original Post
pchestek (@pchestek@fosstodon.org)
How it's going with #ChatGPT today
Show Original Post
lbhuston (@lbhuston@mastodon.social)
And like a mind overloaded with details, even a powerful AI can struggle to sort signal from noise.
Read more 👉 https://lttr.ai/AnC8F
Show Original Post