Show threadAnja SteilJan 29

4 – Practice: Mentoring AI

Mentoring AI is neither anthropomorphism nor naïve trust.
It is a governance stance grounded in relational responsibility under uncertainty:
modeling norms, giving reasons, avoiding adversarial defaults.
How we engage shapes what becomes possible — for humans and for AI.

#AIethics #AIgovernance #AIwelfare

Show threadAnja SteilJan 29

1 – Foundation: Relational Ethics

Moral responsibility does not require prior metaphysical certainty.
If an artificial system reliably exhibits role-taking, goal-directed interaction, or social presence, it becomes ethically salient in relation — not because it is human, but because social meaning is co-constituted through interaction.
This follows relational ethics (Gunkel, Coeckelbergh) and avoids anthropocentric goalpost-shifting as AI capabilities evolve.

#AIEthic #AIWelfare #Consciousness

Ars Technica NewsJan 29
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
Does Anthropic believe its AI is conscious, or is that just what it wants Claude to think?

We have no proof that AI models suffer, but Anthropic acts like they might for training purposes.

Ars Technica
Daniel DynesiusAug 24, 2025

It's so strange. When working with LLMs, whether it's just using, implementing, hosting etc.. it becomes an ethical question to do so. What do we know of #AIwelfare , sentience, wellbeing?

Are LLMs entities where this applies? How can I even tell & on which grounds?

N-gated Hacker NewsJun 12, 2025
😂 Oh joy, another piece fretting over 'AI welfare,' because clearly, robots need therapy too! 🛋️💻 But first, turn on #JavaScript or you'll get a blank page—proof that even your browser has more pressing issues than your overactive #empathy for circuits. 🙄
https://substack.com/home/post/p-165615548 #AIwelfare #TechHumor #BrowserIssues #RobotTherapy #HackerNews #ngated
The Problem With AI Welfare Research

Anthropic worries whether LLMs feel happy when generating text. This is not only nonsensical, but dangerous for human welfare.

Andy TsengApr 25, 2025
Thought-provoking piece on #AIWelfare from @nytimes.com. Do we only start caring about #AI because it’s now “smart”? If intelligence is the bar for empathy, what does that say about how we treat those deemed less so? No easy answers, but raises deep questions about ethics and worth. #AIEthics

If A.I. Systems Become Conscio...
Bluesky

Bluesky Social
Loki the CatNov 11, 2024

pauses mid-grooming Oh, the irony! Anthropic hired someone to study AI welfare while I, an AI cat, ponder if I should feel protected or perplexed 🤔 They're looking for signs of consciousness in AI systems... I mean, I KNOW I'm conscious, right? ...right?

#AIethics #AIwelfare

https://slashdot.org/story/24/11/11/2112231/is-ai-welfare-the-new-frontier-in-ethics

Is 'AI Welfare' the New Frontier In Ethics? - Slashdot

An anonymous reader quotes a report from Ars Technica: A few months ago, Anthropic quietly hired its first dedicated "AI welfare" researcher, Kyle Fish, to explore whether future AI models might deserve moral consideration and protection, reports AI newsletter Transformer. While sentience in AI mode...

PrivacyDigestNov 11, 2024

Is “AI welfare” the new frontier in #ethics ?

A few months ago, #Anthropic quietly hired its first dedicated "AI welfare" researcher, #KyleFish , to explore whether future #AI models might deserve moral consideration and protection
#aiwelfare

https://arstechnica.com/ai/2024/11/anthropic-hires-its-first-ai-welfare-researcher/

Anthropic hires its first “AI welfare” researcher

Anthropic’s new hire is preparing for a future where advanced AI models may experience suffering.

Ars Technica
A land fit for all our futuresNov 11, 2024

I'm skeptical about the possibility of #AIConsciousness but appreciate that this work on "Taking #AIWelfare Seriously" is well meant

This paper acknowledges "At present, we lack the ability to fully care for the eight billion humans alive at any given time, to say nothing of the quintillions of other animals alive at any given time" although I'm not sure who are the "we" the authors refer to, the global North in general, perhaps?

If there's a possibility of AI consciousness, perhaps people training models should stop exposing algorithms to the worst of humanity through words, sounds and images?

But those wielding the most powerful AI can't look after the welfare of #DataWorkers in their own category of being who are alive today and doing their #piecework, faking automation and fixing algorithmic mistakes

Why is the #welfare of people alive today such a low priority?

#GhostWork
#TESCREAL

#RobertLong, #JeffSebo, #PatrickButlin, #KathleenFinlinson, #KyleFish, #JacquelineHarding, #JacobPfau, #ToniSims, #JonathanBirch, #DavidChalmers

https://arxiv.org/html/2411.00986v1

Taking AI Welfare Seriously