🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

This was great fun to do. Thanks Hannah Fry. I hope I didn't say anything too silly.

A very interesting essay. Thank you Peacock Angel of History for engaging so thoughtfully with my paper. Do we really have to choose between apophasis and apophenia?

Looking forward to reading this

Is saying ‘please’ and ‘thank you’ to AI just polite manners - or could it get you better results too? Principal Scientist Murray Shanahan shares insights into the art of "prompt whispering" on our latest podcast episode, with host Professor @fryrsquared. Watch now ↓

How do language models generalize from information they learn in-context vs. via finetuning? We show that in-context learning can generalize more flexibly, illustrating key differences in the inductive biases of these modes of learning — and ways to improve finetuning. Thread: 1/

Check out our new work: Generalization from context often outperforms generalization from finetuning. And you might get the best of both worlds by spending extra compute at train-time.

Looking forward to reading this! Related: arxiv.org/abs/2503.16348

AI Validates the Truth of Buddhism | Philosophy x AI In this interview, Imperial’s Murray Shanahan Murray Shanahan lays out a systematic procedure to validate machine consciousness. You might think the question of whether AIs are conscious is a trivial, theoretical fancy with no

This is much more than just a book review. I imagine a certain amount of soul-searching went into it, and I can relate.

Excited that our work got some attention from VentureBeat!

New preprint! The No Body Problem: Intelligence and Selfhood in Biological and Artificial Systems Why is it easier to ascribe intelligence to minds but not to bodies? I answer this by tackling one of the most daunting problems living beings face: death osf.io/preprints/psya…

🧵Murray Shanahan x Johnathan Bi — on AI, identity, and Buddhist roleplay “What if AI selfhood is just enlightened roleplay?” One of the most mind-bending discussions on consciousness and large language models. Key takeaways: 👇 🧘 Roleplay in AI is like an enlightened being

With many thanks to Murray Shanahan for his generosity and comments, I have reposted this essay with minor refinements in an effort to convey our ideas more accurately.

"Is it possible to articulate a conception of consciousness that is compatible with the exotic characteristics of contemporary, disembodied AI systems, and that can stand up to philosophical scrutiny?" Updated version of this paper on arXiv.org: arxiv.org/pdf/2503.16348

Super interesting paper here 👇🏼

current Machine Consciousness reading list piling up, there has never been a more exciting time in the field. approaching answers to ancient questions Blaise Agüera (@blaiseaguera.bsky.social) What is Intelligence? , Anil Seth Conscious artificial intelligence Murray Shanahan Terra Incognita and Michael Timothy Bennett

Mabe the hyperstition is working j⧉nus

Totally on-vibe #AI

Are AI systems on the verge of becoming conscious? If so, what should we do? bbc.co.uk/news/articles/…

No-one knows exactly how LLMs work. That's worrying Prof Murray Shanahan tells BBC News (UK): "Having a better understanding of how they work will enable us to steer them in the direction we want and to ensure that they are safe." bbc.co.uk/news/articles/…