Asad Sayeed (also @asayeed[@zirk.us])
@asayeed.bsky.social
Thought leader, whose main act of thought leadership is to declare myself thought leader. Computational psycholinguist at the University of Gothenburg.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
the egyptians used huge sponges to build the pyramids. think about it!
February 12, 2026 at 4:31 AM
the egyptians used huge sponges to build the pyramids. think about it!
I have no idea what this is about, but it's great game theory.
Listen, the El Paso situation isn’t as enigmatic or confusing as people think.
Time for some game theory 1/ 🧵
Time for some game theory 1/ 🧵
February 11, 2026 at 7:25 PM
I have no idea what this is about, but it's great game theory.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
You’ll know the Chomsky hierarchy* if you’re a computer scientist.
You’ll know the Chomsky hierarchy if you’re a linguist.
But you’ll only know the subregular language hierarchy if you’re a sublinguist.**
*) regular, context-free languages, etc.
**) phonologist***
***) #sorry
You’ll know the Chomsky hierarchy if you’re a linguist.
But you’ll only know the subregular language hierarchy if you’re a sublinguist.**
*) regular, context-free languages, etc.
**) phonologist***
***) #sorry
February 11, 2026 at 6:20 AM
You’ll know the Chomsky hierarchy* if you’re a computer scientist.
You’ll know the Chomsky hierarchy if you’re a linguist.
But you’ll only know the subregular language hierarchy if you’re a sublinguist.**
*) regular, context-free languages, etc.
**) phonologist***
***) #sorry
You’ll know the Chomsky hierarchy if you’re a linguist.
But you’ll only know the subregular language hierarchy if you’re a sublinguist.**
*) regular, context-free languages, etc.
**) phonologist***
***) #sorry
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Laura Dern for president
February 9, 2026 at 7:36 PM
Laura Dern for president
It's oddly under-discussed that there are actually enough people who want to read and pay for this that it's a viable business model. I knew it before ChatGPT was a thing. A lot of genre fiction readers' preferences DO NOT ALIGN with those of writers or the publishing industry for some reason.
The whole article is repulsive but this passage is at the core of it. She has no interest in writing books, she wants to win some kind of competition that nobody entered
February 8, 2026 at 6:42 PM
It's oddly under-discussed that there are actually enough people who want to read and pay for this that it's a viable business model. I knew it before ChatGPT was a thing. A lot of genre fiction readers' preferences DO NOT ALIGN with those of writers or the publishing industry for some reason.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
but what about girlsenberries
February 8, 2026 at 5:55 PM
but what about girlsenberries
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
I've been singing this to myself all day
February 7, 2026 at 6:39 PM
I've been singing this to myself all day
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Let’s be honest: in many of these winter Olympic “sports”, the snow is doing most of the work. An unexpected thaw would expose these so-called athletes for the frauds they are
February 6, 2026 at 8:38 PM
Let’s be honest: in many of these winter Olympic “sports”, the snow is doing most of the work. An unexpected thaw would expose these so-called athletes for the frauds they are
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
now that UG is unpopular can proponents (me included) be called UGgos?
February 2, 2026 at 9:00 PM
now that UG is unpopular can proponents (me included) be called UGgos?
writing an unauthorized sequel to game of thrones, a prance with princes
February 2, 2026 at 8:59 PM
writing an unauthorized sequel to game of thrones, a prance with princes
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Its sort of a shame that big long form magazine journalism is dying simply because we now frown on the authors simply making stuff up for a better story
February 2, 2026 at 4:45 PM
Its sort of a shame that big long form magazine journalism is dying simply because we now frown on the authors simply making stuff up for a better story
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Nobody ever tries to bribe me. I feel left out in the modern era
February 2, 2026 at 8:52 PM
Nobody ever tries to bribe me. I feel left out in the modern era
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
(And now AI LLMs do something eerily similar - reflect the user back to themselves but packaged as an outside person, with fondness and care. There are already whole subreddits full of people who have fallen in love with their chatbot…)
February 2, 2026 at 4:54 AM
(And now AI LLMs do something eerily similar - reflect the user back to themselves but packaged as an outside person, with fondness and care. There are already whole subreddits full of people who have fallen in love with their chatbot…)
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
bluesky this, blacksky that, what about chomsky, the colorless green social network?
January 31, 2026 at 4:35 AM
bluesky this, blacksky that, what about chomsky, the colorless green social network?
Apropos of nothing, but there's a lot of people -- spectators -- who *might* be "reached" in the "debate me, bro" format who *won't* be reached if you don't engage in that. I'm sorry, but there's a big part of the population who doesn't respect the source otherwise.
January 30, 2026 at 5:15 PM
Apropos of nothing, but there's a lot of people -- spectators -- who *might* be "reached" in the "debate me, bro" format who *won't* be reached if you don't engage in that. I'm sorry, but there's a big part of the population who doesn't respect the source otherwise.
All the Swedish funders very kindly have deadlines all within the same like 2 months, it's nuts.
January 30, 2026 at 3:56 PM
All the Swedish funders very kindly have deadlines all within the same like 2 months, it's nuts.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
***Postdoc@Chalmers University***
There is no doubt that generative AI has changed the way we write. But how, and how does that influence scientific writing? Spread the word among recent graduates in (computational) linguistics and related subjects!
www.chalmers.se/en/about-cha...
There is no doubt that generative AI has changed the way we write. But how, and how does that influence scientific writing? Spread the word among recent graduates in (computational) linguistics and related subjects!
www.chalmers.se/en/about-cha...
Vacancies
www.chalmers.se
January 29, 2026 at 7:58 AM
***Postdoc@Chalmers University***
There is no doubt that generative AI has changed the way we write. But how, and how does that influence scientific writing? Spread the word among recent graduates in (computational) linguistics and related subjects!
www.chalmers.se/en/about-cha...
There is no doubt that generative AI has changed the way we write. But how, and how does that influence scientific writing? Spread the word among recent graduates in (computational) linguistics and related subjects!
www.chalmers.se/en/about-cha...
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Ever tried boiled pasta? Not as crunchy but still good
January 26, 2026 at 11:16 AM
Ever tried boiled pasta? Not as crunchy but still good
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
HOW IT WORKS IN MOVIES:
1. Government lies
2. Brave heroes release video showing the truth
3. Press says “the government is lying!”
HOW IT APPARENTLY WORKS IN REAL LIFE
1. Brave heroes release video showing the truth
2. Government lies
3. Press says “Oh no, what now? So confusing! So hard!”
1. Government lies
2. Brave heroes release video showing the truth
3. Press says “the government is lying!”
HOW IT APPARENTLY WORKS IN REAL LIFE
1. Brave heroes release video showing the truth
2. Government lies
3. Press says “Oh no, what now? So confusing! So hard!”
A battle over the truth erupts after deadly Minneapolis shooting
A battle over the truth erupts after deadly Minneapolis shooting
Federal and local leaders are sharply divided over what happened and who is to blame.
dlvr.it
January 26, 2026 at 6:46 AM
HOW IT WORKS IN MOVIES:
1. Government lies
2. Brave heroes release video showing the truth
3. Press says “the government is lying!”
HOW IT APPARENTLY WORKS IN REAL LIFE
1. Brave heroes release video showing the truth
2. Government lies
3. Press says “Oh no, what now? So confusing! So hard!”
1. Government lies
2. Brave heroes release video showing the truth
3. Press says “the government is lying!”
HOW IT APPARENTLY WORKS IN REAL LIFE
1. Brave heroes release video showing the truth
2. Government lies
3. Press says “Oh no, what now? So confusing! So hard!”
As my ability to understand spoken Swedish has gone up, my ability to interpret spoken Danish has gone *down*. I can only think of how uncomfortable it must be to fill your mouth with burning hot potatoes in order to speak.
January 26, 2026 at 9:09 AM
As my ability to understand spoken Swedish has gone up, my ability to interpret spoken Danish has gone *down*. I can only think of how uncomfortable it must be to fill your mouth with burning hot potatoes in order to speak.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
Not a new thing, to be sure, and I actually wouldn't blame the companies for this. LLMs cannot discern source quality, so hard coding against the use of bad sources will always be a post-hoc effort.
January 25, 2026 at 5:49 PM
Not a new thing, to be sure, and I actually wouldn't blame the companies for this. LLMs cannot discern source quality, so hard coding against the use of bad sources will always be a post-hoc effort.
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
I wrote up some thoughts on why the Poverty of the Stimulus Argument - a mainstay in linguistics and other areas of CogSci - is not (really) an argument.
I also argue, controversially, that usage-based reliance on LLMs to refute the POSA are weakened. 🧵
vincentcarchidi.substack.com/p/on-misunde...
I also argue, controversially, that usage-based reliance on LLMs to refute the POSA are weakened. 🧵
vincentcarchidi.substack.com/p/on-misunde...
On Misunderstanding the Poverty of the Stimulus Argument
The Poverty of the Stimulus Argument is not (really) an argument, and usage-based approaches reliant on LLMs to refute it are weakened.
vincentcarchidi.substack.com
January 22, 2026 at 3:48 PM
I wrote up some thoughts on why the Poverty of the Stimulus Argument - a mainstay in linguistics and other areas of CogSci - is not (really) an argument.
I also argue, controversially, that usage-based reliance on LLMs to refute the POSA are weakened. 🧵
vincentcarchidi.substack.com/p/on-misunde...
I also argue, controversially, that usage-based reliance on LLMs to refute the POSA are weakened. 🧵
vincentcarchidi.substack.com/p/on-misunde...
Reposted by Asad Sayeed (also @asayeed[@zirk.us])
I have never subscribed to the ‘great man of History’ as the font of progress and good things but I’m willing to be persuaded by the ‘Stupid Man of History’ as being responsible for some bad ones.
January 23, 2026 at 3:01 PM
I have never subscribed to the ‘great man of History’ as the font of progress and good things but I’m willing to be persuaded by the ‘Stupid Man of History’ as being responsible for some bad ones.