4 I propose that consciousness isrequired for creativity in projects with aesthetic goals. If an AI lacks consciousness, then it is incapable of aesthetic experi-ence, and without aesthetic experience, it cannot engage in aesthetic creative projects.
3 I argue that there is nocompelling case for thinking that consciousness is generally necessary for creativity. However, lessons learned from thisdiscussion suggest that a more localised claim about aesthetic creativity has greater promise.
2. Human creativity typically involves some conscious experience of the creative project. If consciousness is necessary for creativity,then a case could be made that these (presumably) unconscious machines are not really creative.
1 - AI has displayed notable originality across the worlds of art, science and gaming. But is it right to say that such machines are creative? This question is bound up with other challenging questions about the capacities of artificial systems.
I have a new article out in *AI & Society* on AI creativity and AI consciousness. It's been fun developing this argument. Thanks to all those I've discussed it with! Link and abstract in the comments 🙂
My new paper on artificial consciousness is out today! There's some news coverage of this today too. Links in comments.
Can we ever know if AI is conscious? 🤖
Cambridge philosopher @tom-mcclelland.bsky.social argues that the only rational stance is agnosticism.
Find out why this matters for ethics and tech hype 👇
https://bit.ly/4j0MSaY
The boring explanation is that there are fancy cameras for hybrid in-person/online meetings called OWLs owllabs.co.uk/products/mee...
Was genuinely in a staff meeting this week where the chair said 'We can begin when the owl has arrived'. I think we've reached peak- Cambridge.
New paper with @tom-mcclelland.bsky.social! What is it to see someone as an object – to perceptually objectify them? We argue that seeing is not just a metaphor. We see others afford doing things with and to.
doi.org/10.1111/ejop...
The journal wanted to charge £2400 to make this 2-page piece open access 🤨. I declined, but there's a read-only version with this link rdcu.be/epvJy
Short piece out in *AI & Society* called "The winds of culture: AI art generators and the Aeolian harp". I suggest that making music with an aeolian harp has interesting similarities to making music with AI (link in comments)
Quite right! Biscoff is clearly a coffee accompaniment.
Makes me proud to be British!
For some reason this made me think of @tom-mcclelland.bsky.social
Cambridge Neuroscience are hosting a panel for the Cambridge Festival on 'Consciousness at the Edge'. Hear top neuroscientists (and me) discuss sleep, disorders of consciousness and AI consciousness. Link in comments.
6:00pm-7:30pm on Thu 27 Mar
McCrum Lecture Theatre, Corpus Christi College
The book's here and is designed as an entry-level guide www.politybooks.com/bookdetail?b...
Had a letter on AI published in @theguardian.com this weekend (link in comments). Lots of interesting stuff happening right now on the prospects of conscious AI and what to do about it. #AI #artificialconsciousness
Something along this lines is probably true. But the difficult part is the transition from modelling to experiencing? Why couldn't the (very useful) modelling of your environment happen without you experiencing the environment?
Yeah, that's a very important question. The former seems more likely but the latter is a live possibility.
There's a recent open letter that calls for greater precautions around conscious AI. The Metro interviewed me for an article about it here metro.co.uk/.../fears-ai... Some of the quotes are taken out of context but I think it retains the gist of what I said.
No, the argument in the paper is that it doesn't extend
That's annoying - it seems to work for me arxiv.org/pdf/2412.13145
Thanks. The paper argues that we should be agnostic about artificial consciousness but not about other kinds of consciousness. Probabilistic evidence of consciousness
is fine, but is still hard to come by with advanced AI.
