in search of a book gift?
-
in search of a book gift? š§µ
Iāve been going over some of the books I read this year. These are ones I thought were well worth the read (nb, that is not the same as āagreed with positions espousedā), will keep adding over next few days
-
1/n @Noupside
Renee DiRestaās āInvisible Rulers: The People Who Turn Lies into Realityā
This is a great book (even for someone who has read many books on social media..) with compelling analysis to the effect that it is the specific combination of influencer, algorithm, and crowd that is wreaking havoc. And the chapter on DiRestaās own experiences at the hands of Jordanās āWeaponizationā subcommittee an indicator of things to come
-
2/n Cameron Buckner āFrom Deep Learning to Rational Machines
What the History of Philosophy Can Teach Us about the Future of Artificial IntelligenceāA great book! It pairs specific computational advances in deep learning models with specific empiricist philosophers to examine the extent to which challenges to empiricism have been addressed. Itās a great introduction to deep learning/genAI along the way. For anyone interested in current AI systems and their relationship to human cognition.
-
3/n Doyne Farmer āMaking Sense of Chaos: A Better Economics for a Better Worldā
A very readable book on a complex systems approach and perspective to economics, how it compares to traditional, analytic, economic modelling. Itās written around Farmerās own research journey, so itās concrete and accessible, while also giving a sense of how much agent-based modelling is becoming a key tool used by central banks and policy makers.
-
4/n Michael P Lynch (2019) āKnow-it-all-society: Truth and arrogance in political cultureā
Itās on āmoral entanglementā: one becomes committed to belief in a matter of fact because its truth is evidentially related to a moral commitment. Because its falsity would undermine the perceived evidence for that moral commitment, attacks on the belief come to be seen as attacks on values, and become an identity issue, ending openness to evidence.
with examples from both right and left wing discourse
-
@UlrikeHahn Judging by the cover I'm not likely to buy the book. The sub-title is a deal-breaker, which might be the editors fault.
Rational: "What the history of X can teach us about the future of Y" is indicative of inductive reasoning. As a meta-theory, the history of inductive reasoning allows formulating some laws (which can be backed by meta-mathematical considerations). While inductive reasoning may lead to useful heuristics it never warrants formulating laws (that can be taught).
-
@tg9541 2/2 correspondingly the need for analogous components in computational systems. The āfutureā bit is about identifying what is still missing in current genAI systems when compared to human cognition, and positing that those missing components will need to be added to achieve human level performance. Thatās allā¦.
so you might be (inductively!) over-interpreting that subtitle based on past experience
-
@tg9541 Thomas, itās an excellent book. Buckner is a philosopher who also has a degree in computer science, but what really stood out for me is that Iāve never read a book where someone outside of area writes so accurately (imo) about psychological research. Thatās really, really hard. And for what itās worth, I think that subtitle might be giving you a misleading impression: the book is about the role of various āfacultiesā (perception, memory, imagination) in human cognition and 1/2
-
Ulrike Hahnreplied to Ulrike Hahn last edited by [email protected]
@tg9541 hereās an OA version of the first chapter
http://cameronbuckner.net/professional/Buckner%20-%20DL2RM%20-%20CH1.pdf
btw, it was being on Mastodon that alerted me to Cameron Bucknerās work- @dcm pointed me toward it! so thanks Dimitri for that.
-
@UlrikeHahn @dcm OK, you've convinced me - I'll give Buckner the benefit of doubt. This said, the problem of "upgrading genAI to AGI" is, in my opinion, not well posed. Too many broken analogies come to my mind, including some in psychology (e.g., the concept of "component", computational paradigms as technicisms, etc.).
My take: the future of AI is the demotion of professionals who's job it was to create form. Form is now cheap, and mis-application of form will only promote totalitarianism.
-
@tg9541 @dcm for what itās worth, Thomas, Iāve been apprehensive about the implications of developments in AI for a decade and presently see almost no upsides associated with practical, commercial deployments. I also suspect that weāre still underestimating the negative impact that weāll see.
But Iām a cognitive scientist who builds computational models of cognition, and, from that purely scientific vantage point, genAI is not something I could ignore.