thill.bsky.social
Director of the Donders Centre for Cognition at www.ru.nl/donders
Interested, amongst others, in natural and artificial cognition, how humans and intelligent systems interact, and in theories of embodiment.
303 posts
933 followers
907 following
Regular Contributor
Active Commenter
comment in response to
post
I have difficulties accepting this hypothetical because it feels like accepting that a thought would have an agreed-upon look seems to go against the definition of thought. Seems like this is creating an analogy or metaphor, not literally defining a look.
comment in response to
post
as opposed to questions like “can something look like a thought?”
comment in response to
post
It’s a concept with a relatively clear agreed-upon definition in terms of concrete components, so you can assess the degree to which something matches that definition?
comment in response to
post
I have no solutions, but I saw this floating around on BSky not too long ago and it seems appropriate bsky.app/profile/stux...
comment in response to
post
Not sure that has a meaningful answer though.
comment in response to
post
I understood it as being an assumption that people are either more likely to want or more likely to not want, and the puzzle being figuring out which one it is from the answers given.
(It's clear that "want AND not-want" is the most likely state, but the Q is how much of that is e.g. "want")
comment in response to
post
That brings back memories 😊
comment in response to
post
Yay! Congrats and well deserved! 😊
comment in response to
post
Ah man what a missed opportunity to impose a NaN tariff! Or Inf, automated trading algos would have loved that.
comment in response to
post
Using a random number generator for tariffs makes conceptually more sense than this and you could at least keep some sanity by defining a not-batshit-crazy range to sample from.
Just sad that there is no country they import from but not export to. Would have loved to see the divide by zero.
comment in response to
post
I’m literally trying and failing to come up with something that would make less sense than what they are actually doing.
comment in response to
post
This is absolutely wild. Everyone knows you need to normalise by population size before converting trade deficits to tariffs.
comment in response to
post
Well there’s the book, of course. For a paper, maybe this? www.cell.com/trends/cogni...
John Spencer also wrote a bunch of review-ish things aimed at devel. psychologists.
comment in response to
post
Agreeing to something that is due in six months, then, three months later to something that is due in three months and then, after another month, to something that is due in two months…. that’s what gets you in the end.
comment in response to
post
I find it sad that on the one hand, this is such a 90s statement and on the other, it still needs saying.
comment in response to
post
ClearlyNvrSentAnSMSInThe90s 😉
comment in response to
post
Wait, does he mean to say that Musks days are numbered? Does Musk know?
comment in response to
post
I'm pretty sure there is nothing socratic about how an LLM interacts with users 😉
comment in response to
post
I wrote this thing with David Windridge in 2018 - basically arguing that even if you want to be an all-in conputationalist (which all of current AI seems to be), you still need a body to be more than a data source/sink:
www.sciencedirect.com/science/arti...
comment in response to
post
ouais, les semla, c’est un délice. T’imagines qu’au début, c’était qu’un petit pain trempé dans du lait… ça a bien évolué au fil des temps 😉
comment in response to
post
Je ne suis pas fort en pâtisserie, mais là, il me manque les feuilles d’une pâte feuilletée 😉
En suédois, on appelle ce genre de semla ´wienersemla’, ein voici une recette (en suédois, mais bon, il n’y a rien de surprenant):
www.coop.se/recept/wiene...
comment in response to
post
En Suède, on trouve aussi des semla à la pâte feuilletée - bon si on aime pas, on aime pas, mais pour le reste, c le meilleur des mondes 😋
comment in response to
post
So, Hungarian notation, but also applied to types. Careful, someone might actually think this would be a good idea 😁
comment in response to
post
I would argue that the extra work is focussing more specifically on the aspects that humans find difficult to explain as exclusively reactive behaviours, e.g. behaviours we consider “intentional”
comment in response to
post
I think around 2007/8 or so. It was at the end of the first euCognition network (before EuCogII started).
comment in response to
post
Yeah and some are easier to agree with than others, but it depends on the reader which ones they are 😉 What was nice about euCognition was that it was a cognitive systems network, so had people interested in natural and/or artificial cognition. The set of definitions is accordingly diverse.
comment in response to
post
A long time ago, the euCognition network asked this. David Vernon still has the answers up here: www.vernon.eu/euCognition/.... I don't think this is something people will ever agree on.
comment in response to
post
For a forensic analysis of Truss’s letter, see: davidallengreen.com/2025/01/a-cl... It’s even worse than it seems to be. 😄
comment in response to
post
Yeah, I think it's basically that.
1. This thing is even more popular than in our wildest dreams. We are so gonna make 100B $$$ by 2029, never mind our current losses (invest now).
2. The price right now is a bargain (subscribe now)
comment in response to
post
You can get the alphabetically sorted list in the search field of your app library:
support.apple.com/en-gb/guide/...
comment in response to
post
Also, you can tell how important this consideration is to Tesla by the fact that the locking mechanism is apparently designed to remain functional and of added value even if the vehicle BLOWS UP.
comment in response to
post
I still maintain that we live in the bad alternate timeline that branched of from the prime universe in 2016 due to the actions of a time traveller who replaced Elon Musk. Wonky time is just a side effect of the universe unravelling as a result 😉