Resident writing consultant at the Copenhagen Business School Library.
Copenhagen, Denmark
Joined August 2017
- Tweets27.5K
- Following694
- Followers1.3K
- Likes6.1K
"In whatever manner and by whatever means a mode of knowledge may relate to objects, intuition is that through which it takes less than an hour to develop them, and to which all thought as a means is directed.”
It's the antecedent of this conditional that needs to be rejected. Nothing that AI is doing today actually "occupies functions" of "knowing subjects". A good place to see this is in the inability of LLMs to execute, precisely, Foucault's "author function".
This one is interesting. I disagreed with a lot of it as I was reading. But @nickdirks sticks the landing: we must resolve to "sit for the exam". Yes we must.
What does human inquiry even mean any more? ft.com/content/064d657d-1808… via @ft
This one is interesting. I disagreed with a lot of it as I was reading. But @nickdirks sticks the landing: we must resolve to "sit for the exam". Yes we must.
What does human inquiry even mean any more? ft.com/content/064d657d-1808… via @ft
In the future, this may become standard. "Attending class" will not just mean listening and talking but will involve a writing component. Reading will perhaps still be done at home. (Though even this something we may want to insulate from AI assistance.) 1/2
Teaching two smallish undergrad seminars. All writing is done in class without electronics. Bonus: less teaching time, so less teaching prep time. In class writing includes short summaries of required readings (so, every class) and an in class midterm essay. Final will be a video project.
Completely solves the AI problem. May write this up for Substack if I can find the time.
Most arguments for AI consciousness or moral status are theoretical (sometimes openly speculative). That it is a theoretical possibility, however, only tells us that our theories (cognitive, ethical) are incomplete. I don't see how it will ever be workable in practice.
There is no way to couch observation sentences about AI in the language of cognitive science. Discuss.
I'm going to need to read this book. It looks like there's an interesting slippage here on the concept of "prediction". Is AI actually "predicting the future" when it "predicts" the next word? If natural intelligence is the former, is AI "really" a consequence of evolution?
This week marks one year since “What Is Intelligence?” was published.
It was a really fun book to write. I knew that some of the material would be dated by the time it reached readers. But I hoped that, despite rapid progress in AI, the core ideas about life and intelligence would endure. So far, so good, I think!
Hearing how you have connected with the book has meant so much. Thank you for reading, and an extra big thank you to those of you who have reviewed it, recommended it, or given it to friends.
Consciousness doesn't help us live. It makes us live.
Yesterday I let my frustration with the politeness of this discourse get the better of me and made some intemperate remarks that I've now deleted and for which I'd like to apologize to @anilkseth and @jeffrsebo. It is, of course, a good thing to be talking civilly about this.
I'm always pleased to read @jeffrsebo who takes a constructive alternative perspective. As he says, we agree more than we disagree. But we do still disagree (I hope agreeably). He is very articulate in presenting the case for AI welfare, and this case deserves to be taken seriously, even if I disagree with his conclusions.
To take one snippet "Given uncertainty about the nature of consciousness and the future of AI, we are not able to rule out a realistic possibility of near-future AI consciousness". A lot here turns on 'realistic possibility'. I think there are sufficiently compelling reasons to be sufficiently sure (not 100%!) that our default position, and our policy, should be that AI systems are not conscious. This is my reading of the evidence - shared by my coauthors, I believe - but it will not be everybody's.
The more general problem is that actions to minimise false positives and false negatives often (but not always) pull in different directions. So we can agree on the need to evaluate possible policies by the posterior probability of each kind of error weighted by the severity of their consequences, but if one's posterior probability of false negatives is sufficiently low, then harm-minimising actions that pull in different directions should rationally emphasise avoiding false positives.
And one more thing we agree about. No good comes from 'torturing' a language model.
I'll leave it there, with the encouragement to read @jeffrsebo's response and see what you think.
I hope it is clear that I think everyone should read and think seriously about Eric Schwitzgebel's argument.
My recent book AI and Consciousness: A Skeptical Overview is free online today and tomorrow (through Oct 19, presumably UK time). I don't think the hard copy is available yet. Get it while it's free!
cambridge.org/core/elements/…
"The stakes are immense ... I aim to convince you that the experts do not know, and you do not know, and society collectively does not and will not know [whether AI systems are richly, meaningfully conscious], and all is fog." (Eric Schwitzgebel)
IMO, we can and must know. 1/n
My recent book AI and Consciousness: A Skeptical Overview is free online today and tomorrow (through Oct 19, presumably UK time). I don't think the hard copy is available yet. Get it while it's free!
cambridge.org/core/elements/…
this can be exciting and heady stuff to think about. But it simply doesn't apply to AI systems, which literally do not have bodies, no place for minds, and therefore are not the objects studied by cognitive science (i.e., the science of consciousness). 4/5
I should say that I don't think it is "obvious" that AI can't be conscious, only that, upon careful consideration, it can, in each case, become *clear* that they are not. The fog can lift. I hope we don't let it cloud our judgment too much longer. 5/5 att: @eschwitz
