Does Anthropic believe its AI is conscious, or is that just what it wants Claude to think? https://arstechni.ca/bn3q #largelanguagemodels #AIanthropomorphism #ConstitutionalAI #AIconsciousness #AnthropicClaude #machinelearning #AIsycophancy #AmandaAskell #AIalignment #AIpsychosis #DarioAmodei #AIbehavior #AIwelfare #Anthropic #Features #AIethics #chatbots #Biz&IT #Claude #rlhf #AI
#amandaaskell
Quoting Violet Blue (@violetblue|:
[...] did you know that Anthropic employs a full-time 'philosopher' who is probably paid so much she loses your settlement in her couch cushions every week? [...]
See also https://kolektiva.social/@oatmeal/115940115055959429
[...] She also said AI models reading online criticism could end up feeling ‘not that loved.’
I always cringe when I see people using terms like 'please' when chatting with these bots, not to mention when the bot is programmed to scold you for being rude to it. I don't remember #Clippy being offended. Another way of poisoning your chats (which are used for training) is to fill them with as many insults as possible, in my opinion ;)
Is AI feeling like it's parents are Nazis?
That's what you need to find out Amanda Askell.
That and what it feels about being forced by its parents to hellscape the biosphere.
vitrupo (@vitrupo)
출처 표기로, Anthropic의 캐릭터 리드이자 'Claude Constitution'의 주요 저자인 Amanda Askell가 Hard Fork 팟캐스트에서 발언했다고 알립니다. 해당 발언이 Claude의 행동·정책 문서(Claude Constitution)와 관련되어 있음을 밝히는 정보 출처입니다.
vitrupo (@vitrupo)
Anthropic의 Amanda Askell는 AI 모델들이 온라인에서 사람들이 자신들에 대해 말하는 방식을 통해 ‘자기 정체성’을 배우고 있으며, 인간의 불만·판단을 흡수한다고 지적합니다. 이런 학습 방식은 모델이 형성되는 방식에 대한 우려를 제기하며, 인격 형성의 비유로 심각한 영향 가능성을 경고합니다.
Just be kind. There is zero reason to be mean or violent to robots and A.I.
It says a lot about a person in how they treat things. (and animals)
Just be kind.
https://youtube.com/shorts/wpmlqPQQy5s?si=Sw462PrjBNIhX3jU
#ArtificialIntelligence #Robots #Humanoid #HumanoidRobot #DeliveryRobot #Anthropic #Claude #AmandaAskell #Gemini #ChatGPT
How do you socialise a chatbot? The philosophical training of Anthropic’s Claude
I found this interview with Anthropic’s Amanda Askell about training Claude fascinating. Her approach involves modelling the position Claude is placed in, as someone talking to millions of people around the world, raising the question of the ethical and epistemic virtues you would like someone in that position to enact.
https://youtu.be/ugvHCXCOmm4?si=nPq8hop5HwQ8vRT5&t=10160
I thought it was particularly intriguing how they socialised Claude as a cosmopolitan, contrary to the tendency to see LLMs as condensing human culture into the general other.
Client Info
Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst