Mark Dingemanse
@dingemansemark.bsky.social
Language, interaction, tech • Here with doubts about the next monetizable monopoly... • papers https://markdingemanse.net • blog https://ideophone.org • fedi https://scholar.social/@dingemansemark/ • POSSE: Publish on Own Site, Syndicate Everywhere
Pinned
Welcome new followers! You may find me posting about linguistics, language technology, "AI", open research, epistemic injustice, home computing, and whatever captures my fancy. I blog at ideophone.org (since 2007) and my academic site —with PDFs of papers etc.— is markdingemanse.net
Reposted by Mark Dingemanse
I have a co-authored, peer-reviewed chapter in this newly released book! With Laura Chahda, and drawing on my research for BYE BYE I LOVE YOU, I wrote about linguistic considerations at the end of life, aiming to show how speech-language pathologists can support communication. It's important!
November 22, 2025 at 7:05 PM
I have a co-authored, peer-reviewed chapter in this newly released book! With Laura Chahda, and drawing on my research for BYE BYE I LOVE YOU, I wrote about linguistic considerations at the end of life, aiming to show how speech-language pathologists can support communication. It's important!
the whack-a-mole problem: every few months we see a new creative bypass of this kind (that will be half-patched by ad hoc metaprompts & later by finetuning)
when will it dawn that LLMs are fundamentally vulnerable to endless new attacks like this, and therefore fundamentally unsafe?
when will it dawn that LLMs are fundamentally vulnerable to endless new attacks like this, and therefore fundamentally unsafe?
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."
https://arxiv.org/html/2511.15304v1
https://arxiv.org/html/2511.15304v1
November 20, 2025 at 5:49 PM
the whack-a-mole problem: every few months we see a new creative bypass of this kind (that will be half-patched by ad hoc metaprompts & later by finetuning)
when will it dawn that LLMs are fundamentally vulnerable to endless new attacks like this, and therefore fundamentally unsafe?
when will it dawn that LLMs are fundamentally vulnerable to endless new attacks like this, and therefore fundamentally unsafe?
Reposted by Mark Dingemanse
I was in too much haste, and now have no time left.
November 16, 2025 at 8:23 AM
I was in too much haste, and now have no time left.
I'm seeing quite some citations of our 2021 paper doi.org/10.1177/0023830920914339 that seem to have somehow missed
not just the evidence and arguments from the body of the paper
not just the abstract
but the subtitle 🫠
not just the evidence and arguments from the body of the paper
not just the abstract
but the subtitle 🫠
November 13, 2025 at 9:31 PM
I'm seeing quite some citations of our 2021 paper doi.org/10.1177/0023830920914339 that seem to have somehow missed
not just the evidence and arguments from the body of the paper
not just the abstract
but the subtitle 🫠
not just the evidence and arguments from the body of the paper
not just the abstract
but the subtitle 🫠
Taking notes as I'm reading a draft policy piece that ominously lauds AI's possibilities to make processes more efficient 👀
Due to mass hype and operational ambiguity, executives can use "AI" as a justification for job cuts, increased worker surveillance, speed-up, reducing hiring—with little accountability or repercussion. The *logic* of AI has often been as damaging to labor as the tech itself (see: DOGE.)
November 13, 2025 at 7:35 PM
Taking notes as I'm reading a draft policy piece that ominously lauds AI's possibilities to make processes more efficient 👀
Language as Particle, Wave and Field, Pike 1959
November 13, 2025 at 7:16 PM
Language as Particle, Wave and Field, Pike 1959
Pleidooi voor nutteloze beschaving doi.org/10.1075/nb.0...
Met voorzitters van Stutterheim tot Leufkens, sprekers van Jakobson tot Nissim, en onderwerpen van fonologie tot AI: de AVT heeft een roerige geschiedenis en een mooie toekomst — als we "nutteloze beschaving" op waarde blijven schatten
Met voorzitters van Stutterheim tot Leufkens, sprekers van Jakobson tot Nissim, en onderwerpen van fonologie tot AI: de AVT heeft een roerige geschiedenis en een mooie toekomst — als we "nutteloze beschaving" op waarde blijven schatten
Pleidooi voor nutteloze beschaving | John Benjamins
Abstract
In het vijfenzeventigste jaar van het bestaan van de Algemene Vereniging voor Taalwetenschap (AVT) waait er een
fascistische wind in Nederland. Universiteiten gehoorzamen bij voorbaat door ge...
doi.org
November 11, 2025 at 10:47 AM
Pleidooi voor nutteloze beschaving doi.org/10.1075/nb.0...
Met voorzitters van Stutterheim tot Leufkens, sprekers van Jakobson tot Nissim, en onderwerpen van fonologie tot AI: de AVT heeft een roerige geschiedenis en een mooie toekomst — als we "nutteloze beschaving" op waarde blijven schatten
Met voorzitters van Stutterheim tot Leufkens, sprekers van Jakobson tot Nissim, en onderwerpen van fonologie tot AI: de AVT heeft een roerige geschiedenis en een mooie toekomst — als we "nutteloze beschaving" op waarde blijven schatten
One year ago, with Stella Punselie and Bonnie McLean: The Anatomy of Iconicity, in Open Mind doi.org/10.1162/opmi...
In which we combine iconicity ratings, experimental evidence, and structure mapping to identify & explain crosslinguistic patterns of iconic words
#iconicity #cognitivescience
In which we combine iconicity ratings, experimental evidence, and structure mapping to identify & explain crosslinguistic patterns of iconic words
#iconicity #cognitivescience
November 8, 2025 at 2:24 PM
One year ago, with Stella Punselie and Bonnie McLean: The Anatomy of Iconicity, in Open Mind doi.org/10.1162/opmi...
In which we combine iconicity ratings, experimental evidence, and structure mapping to identify & explain crosslinguistic patterns of iconic words
#iconicity #cognitivescience
In which we combine iconicity ratings, experimental evidence, and structure mapping to identify & explain crosslinguistic patterns of iconic words
#iconicity #cognitivescience
As I was saying, if you give up thinking for yourself, the singularity is already here ideophone.org/bringing-abo...
(re FT story with "AI pioneers" claiming "human-level intelligence is already here" www.ft.com/content/5f2f... )
(re FT story with "AI pioneers" claiming "human-level intelligence is already here" www.ft.com/content/5f2f... )
AI pioneers claim human-level general intelligence is already here
Tech leaders say systems now rival human intelligence in key tasks, further fuelling the superintelligence debate
www.ft.com
November 6, 2025 at 7:30 PM
As I was saying, if you give up thinking for yourself, the singularity is already here ideophone.org/bringing-abo...
(re FT story with "AI pioneers" claiming "human-level intelligence is already here" www.ft.com/content/5f2f... )
(re FT story with "AI pioneers" claiming "human-level intelligence is already here" www.ft.com/content/5f2f... )
Climbing the scaffolds and replastering this earthen building together is a beautiful sight, and a powerful metaphor for
language, as I wrote in @annualreviews.bsky.social doi.org/10.1146/annu... (h/t @maartenkossmann.bsky.social)
language, as I wrote in @annualreviews.bsky.social doi.org/10.1146/annu... (h/t @maartenkossmann.bsky.social)
October 24, 2025 at 8:48 PM
Climbing the scaffolds and replastering this earthen building together is a beautiful sight, and a powerful metaphor for
language, as I wrote in @annualreviews.bsky.social doi.org/10.1146/annu... (h/t @maartenkossmann.bsky.social)
language, as I wrote in @annualreviews.bsky.social doi.org/10.1146/annu... (h/t @maartenkossmann.bsky.social)
Two bits of music as the weekend nears:
1. Jacob Collier's solo (but not quite solo) show at North Sea Jazz, with special guest Julian Lage around the 40m mark www.youtube.com/watch?v=mjCS...
2. New Netherlands Bach Society recording just dropped, w/ Maude Gratton: www.youtube.com/watch?v=JmTz...
1. Jacob Collier's solo (but not quite solo) show at North Sea Jazz, with special guest Julian Lage around the 40m mark www.youtube.com/watch?v=mjCS...
2. New Netherlands Bach Society recording just dropped, w/ Maude Gratton: www.youtube.com/watch?v=JmTz...
Bach - Toccata in G minor BWV 915 - Maude Gratton | Netherlands Bach Society
YouTube video by Netherlands Bach Society
www.youtube.com
October 23, 2025 at 6:56 PM
Two bits of music as the weekend nears:
1. Jacob Collier's solo (but not quite solo) show at North Sea Jazz, with special guest Julian Lage around the 40m mark www.youtube.com/watch?v=mjCS...
2. New Netherlands Bach Society recording just dropped, w/ Maude Gratton: www.youtube.com/watch?v=JmTz...
1. Jacob Collier's solo (but not quite solo) show at North Sea Jazz, with special guest Julian Lage around the 40m mark www.youtube.com/watch?v=mjCS...
2. New Netherlands Bach Society recording just dropped, w/ Maude Gratton: www.youtube.com/watch?v=JmTz...
The learned helplessness that many academics display when it comes to connecting their devices to projectors and displaying slides is a form of negligence
Knowing your tools is part of the job
Knowing your tools is part of the job
Do you have any extremely niche, but serious, ethical stances?
October 20, 2025 at 4:54 AM
The learned helplessness that many academics display when it comes to connecting their devices to projectors and displaying slides is a form of negligence
Knowing your tools is part of the job
Knowing your tools is part of the job
The latest issue of @europeanreview.bsky.social is so gorgeous and interesting
October 11, 2025 at 10:14 AM
The latest issue of @europeanreview.bsky.social is so gorgeous and interesting
Reposted by Mark Dingemanse
New preprint 🌟 Psychology is core to cognitive science, and so it is vital we preserve it from harmful frames. @irisvanrooij.bsky.social & I use our psych and computer science expertise to analyse and craft:
Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...
🧵 1/
Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...
🧵 1/
October 4, 2025 at 5:33 AM
New preprint 🌟 Psychology is core to cognitive science, and so it is vital we preserve it from harmful frames. @irisvanrooij.bsky.social & I use our psych and computer science expertise to analyse and craft:
Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...
🧵 1/
Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...
🧵 1/
Reposted by Mark Dingemanse
ACL terminology:
"Paper" = anything with lots of tables with numbers.
"Position paper" = anything critical or just deeper.
I propose:
"Paper" = intellectual contribution.
"Engineering paper" = anything purely technocratic.
"Paper" = anything with lots of tables with numbers.
"Position paper" = anything critical or just deeper.
I propose:
"Paper" = intellectual contribution.
"Engineering paper" = anything purely technocratic.
March 19, 2025 at 6:17 AM
ACL terminology:
"Paper" = anything with lots of tables with numbers.
"Position paper" = anything critical or just deeper.
I propose:
"Paper" = intellectual contribution.
"Engineering paper" = anything purely technocratic.
"Paper" = anything with lots of tables with numbers.
"Position paper" = anything critical or just deeper.
I propose:
"Paper" = intellectual contribution.
"Engineering paper" = anything purely technocratic.
quelle surprise that the same shitty company is now taking all rights to turn your academic content and likeness into AI sludge in your name, unless you opt out
quelle surprise that academia dot edu now offers AI slopcasts
they were a garbage flattery service ten years ago, and they have turned up the enshittification a few notches ideophone.org/academia-edu...
they were a garbage flattery service ten years ago, and they have turned up the enshittification a few notches ideophone.org/academia-edu...
How Academia.edu promotes poor metadata and plays to our vanity – The Ideophone
ideophone.org
September 22, 2025 at 7:16 PM
quelle surprise that the same shitty company is now taking all rights to turn your academic content and likeness into AI sludge in your name, unless you opt out
Reposted by Mark Dingemanse
Reposted by Mark Dingemanse
This is a fantastic document
I can honestly say that LLMs have made me a better writer, because I see its dreck and write better out of spite
For that reason I think they can be handy in a fear of a blank page way, in that I see the output and the impostor syndrome is gone! I can do better than that
I can honestly say that LLMs have made me a better writer, because I see its dreck and write better out of spite
For that reason I think they can be handy in a fear of a blank page way, in that I see the output and the impostor syndrome is gone! I can do better than that
See e.g. also @dingemansemark.bsky.social ‘s Generative AI and Research Integrity osf.io/preprints/os...
OSF
osf.io
September 13, 2025 at 7:35 AM
This is a fantastic document
I can honestly say that LLMs have made me a better writer, because I see its dreck and write better out of spite
For that reason I think they can be handy in a fear of a blank page way, in that I see the output and the impostor syndrome is gone! I can do better than that
I can honestly say that LLMs have made me a better writer, because I see its dreck and write better out of spite
For that reason I think they can be handy in a fear of a blank page way, in that I see the output and the impostor syndrome is gone! I can do better than that
happy to be a lazy luddite if it means I don't have to listen to fake conversations no actual human could be bothered to have
September 9, 2025 at 8:48 PM
happy to be a lazy luddite if it means I don't have to listen to fake conversations no actual human could be bothered to have
Reposted by Mark Dingemanse
The unis that insulate students from bad AI will outperform those that embed AI in curricula; the unis that go full luddite, invest in depts that do One Thing Well and hire the world's best scholars of that Thing will outperform this sort of MBA-fueled garbage thinking
September 6, 2025 at 3:11 AM
The unis that insulate students from bad AI will outperform those that embed AI in curricula; the unis that go full luddite, invest in depts that do One Thing Well and hire the world's best scholars of that Thing will outperform this sort of MBA-fueled garbage thinking
Incredible to see that universities using Elsevier's Pure all obediently link to Pure profiles for "research output" but Pure never links back, making the uni's own profile pages useless and unfindable and Pure profiles the premier web presence of their researchers 1/2
A rant about Elsevier Pure – The Ideophone
ideophone.org
September 5, 2025 at 9:53 AM
Incredible to see that universities using Elsevier's Pure all obediently link to Pure profiles for "research output" but Pure never links back, making the uni's own profile pages useless and unfindable and Pure profiles the premier web presence of their researchers 1/2
About that... we audited the open source status of Lumo and found it came in rock bottom in the EU Open Source AI Index osai-index.eu/news/lumo-pr... — consider sharing more details to rise through the openness ranks, @proton.me 🫣
#OpenSource #OpenWashing #lumo
#OpenSource #OpenWashing #lumo
Introducing Lumo, a privacy-first AI built by Proton, where every conversation is confidential
✅ Zero-access encryption
✅ No-logs policy
✅ Open-source and auditable
Try @asklumo.proton.me for free, no sign-up required: lumo.proton.me
✅ Zero-access encryption
✅ No-logs policy
✅ Open-source and auditable
Try @asklumo.proton.me for free, no sign-up required: lumo.proton.me
August 29, 2025 at 8:40 PM
About that... we audited the open source status of Lumo and found it came in rock bottom in the EU Open Source AI Index osai-index.eu/news/lumo-pr... — consider sharing more details to rise through the openness ranks, @proton.me 🫣
#OpenSource #OpenWashing #lumo
#OpenSource #OpenWashing #lumo
Reposted by Mark Dingemanse
I considered writing a long carefully constructed argument laying out the harms and limitations of AI, but instead I wrote about being a hater. Only humans can be haters.
I Am An AI Hater
I am an AI hater. This is considered rude, but I do not care, because I am a hater.
anthonymoser.github.io
August 27, 2025 at 5:04 PM
I considered writing a long carefully constructed argument laying out the harms and limitations of AI, but instead I wrote about being a hater. Only humans can be haters.
Reposted by Mark Dingemanse
Great essay from @frank.computer!
"A chair orders you to sit and sit in a particular way, by its design.
Your being is intended through the tool: you are intended to sit still, face forward, and behave. Artificial intelligence works in exactly the same way."
"A chair orders you to sit and sit in a particular way, by its design.
Your being is intended through the tool: you are intended to sit still, face forward, and behave. Artificial intelligence works in exactly the same way."
Stop saying that AI is just a tool and it only matters how it is used
I’m tired of this phrase and this simple way of thinking about tools. This blog post is a wandering train of thought on the topic of what tools are and why it matters to be even slightly more mature i...
www.frank.computer
August 25, 2025 at 1:52 PM
Great essay from @frank.computer!
"A chair orders you to sit and sit in a particular way, by its design.
Your being is intended through the tool: you are intended to sit still, face forward, and behave. Artificial intelligence works in exactly the same way."
"A chair orders you to sit and sit in a particular way, by its design.
Your being is intended through the tool: you are intended to sit still, face forward, and behave. Artificial intelligence works in exactly the same way."
Reposted by Mark Dingemanse
Prachtig stuk in de @volkskrant.nl van @laurensvhg.bsky.social over “Grote AI-denker Margaret Boden (1936-2025) ...”
Echt het lezen waard!
www.volkskrant.nl/tech/grote-a...
Echt het lezen waard!
www.volkskrant.nl/tech/grote-a...
Grote AI-denker Margaret Boden (1936-2025) herkende al vroeg de creativiteit van computers
Kunnen computers werkelijk creatief zijn? Wie zich over deze vraag buigt, kan niet om Margaret Boden heen. Het multitalent dat filosofie, AI, psychologie en cognitieve wetenschappen verbond, overleed ...
www.volkskrant.nl
August 23, 2025 at 6:24 PM
Prachtig stuk in de @volkskrant.nl van @laurensvhg.bsky.social over “Grote AI-denker Margaret Boden (1936-2025) ...”
Echt het lezen waard!
www.volkskrant.nl/tech/grote-a...
Echt het lezen waard!
www.volkskrant.nl/tech/grote-a...