Sign in

Wiep Hamstra

@wiep.bsky.social
112 followers 87 following 125 posts

Works in tech for gov. I am a human with brains so I can read, think and write without robots. Dutch chatbotoloog. See www.digimonitor.nl/onderzoek/onderzoek-chatbot-2026/ Wieplog destaatvanhetweb.nl

PostsRepliesMedia
Reposted by Wiep Hamstra
Lia @liasae.bsky.social · 24/09/2026
I am now keeping this on speed-dial
Edited old IBM printout in black caps on off-white paper.
'A computer can never be held accountable
Therefore a computer must always do the hacking haha what are they gonna do, put a computer in jail? Haha "oops it's alive" lmao'
81101323
Wiep Hamstra @wiep.bsky.social · 26/09/2026
When an anonymous AI-professor (we know his name, but in most headlines not mentioned or presented with a photo) frauds his way into academics by prompting his ‘articles’, and all say suddenly the system is the problem: then we know who is protecting who, don’t we?
111
Wiep Hamstra @wiep.bsky.social · 25/09/2026
Je weet zeker dat het pas erg is als het zielig is als een witte man wordt uitgelachen om iets wat de rest van de niet mannelijke en witte samenleving al een tijdje roept.
131
Reposted by Wiep Hamstra
Dr Katie Twomey @k2mey.eurosky.social · 19/09/2026
youtu.be/bqHsy8KtqEg?... please watch this to understand the mess we're in. I am SO lucky to have learned about the history of "AI" at Sussex during my PhD. But this has taught me more. I might have to go and smash things, don't mind me.
youtu.be
Ghost In The Machine
YouTube video by Arjav Mehta
53018
Reposted by Wiep Hamstra
Smooth Dunk @smoothdunk2.bsky.social · 20/09/2026
when ai leaves u behind
7 panel comic.

Panel 1:
A Red Guy with AI across his chest looks like he’s about to run away. A Yellow Guy stands next to him. Red Guy says “I’m the future, buddy. You comin’?”
Yellow Guy replies “Nah”

Panel 2: 
Red Guy is almost out of the panel. He leans in and says “Enjoy getting left behind, pal”
Yellow Guy replies “ok”

Panel 3:
Yellow Guy stands alone.

Panel 4: 
Close up on Yellow Guy, there is a quiet “huff-huff-huff” noise coming from behind him

Panel 5:
Yellow Guy turns to look behind him

Panel 6: 
Close up on Yellow Guy’s face as if from the point of view of an attacker.
Yellow Guy yells “Jesus Christ”

Panel 7:
AI Red Guy is now chasing the Yellow Guy. 
Red Guy: “Use me”
Yellow Guy “No”
Red Guy “USE ME”
75256379611
Reposted by Wiep Hamstra
Pavel @spavel.bsky.social · 20/09/2026
Every chatbot prompt box says "check answers for accuracy." But increasingly, "humans in the loop" refuse to do that. Instead, they become complicit in the AI industry's project of destroying the provenance of information. But you are person, and it is an app. It can't take responsibility for you.
productpicnic.beehiiv.com
Assigning agency to AI means surrendering your own
Claude is not a colleague, collaborator, or sentient being. It is a SaaS product.
825385
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 20/09/2026
"AI-praatkwadrant: praat over AI maar dan zonder ruzie" Helder en constructief stuk door @wiep.bsky.social 👏👏👏 📝 destaatvanhetweb.nl/2026/09/20/a...
destaatvanhetweb.nl
AI-praatkwadrant: praat over AI maar dan zonder ruzie - De Staat van het Web!
Voordelen en labels die aan AI worden toegekend, zoals sneller werken, slimme keuzes, handig op kantoor, nieuw: zijn ze nu waar of niet? En hoe zit het met de nadelen van die technologie? Voor wie zij...
263
Wiep Hamstra @wiep.bsky.social · 20/09/2026
Vroege feedback: welkom. Ik schreef dit over hoe je gesprekken kunt voeren “AI”: destaatvanhetweb.nl/2026/09/20/a...
destaatvanhetweb.nl
AI-praatkwadrant: praat over AI maar dan zonder ruzie - De Staat van het Web!
Voordelen en labels die aan AI worden toegekend, zoals sneller werken, slimme keuzes, handig op kantoor, nieuw: zijn ze nu waar of niet? En hoe zit het met de nadelen van die technologie? Voor wie zij...
252
Reposted by Wiep Hamstra
Ilyaz Nasrullah @ilyazn.eurosky.social · 13/09/2026
Sinds ChatGPT op de markt verscheen, raakt de samenleving om de zoveel tijd in de paniek. We denken dan ineens: nu is het zo ver, AI gaat ons uitroeien. Zo’n moment is er nu weer, ‘dankzij’ de recente hack bij Hugging Face door OpenAI. www.ilyaz.nl/blog/een-op-...
ilyaz.nl
Een op hol geslagen narratief: het Hugging Face-incident | Ilyaz Nasrullah
De feiten vs. het narratief in de media (inclusief technische context)
55921
Wiep Hamstra @wiep.bsky.social · 13/09/2026
Deploying a wonky chatbot with uncanny handwritten content, mixed with “AI” and having “AI” doing the checks in ‘conversations’ 🤔 “Voor een groot deel van deze controles gebruiken we inmiddels een taalmodel als beoordelaar. AI die AI beoordeelt dus." virtuelegemeenteassistent.nl/nieuws/wanne...
virtuelegemeenteassistent.nl
Virtuele Gemeente Assistent
Gem is de virtuele gemeente-assistent die bij 27 gemeenten draait en 2,8 miljoen inwoners bedient. Common Ground, in beheer bij ICTU.
131
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 13/09/2026
This is also #critical_AI_literacy, even (or especially) if it's very different from what you are used to.
32211
Reposted by Wiep Hamstra
Olivia Guest · Ολίβια Γκεστ @olivia.science · 09/09/2026
how the turntables
0418
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 09/09/2026
💯 mu.social/profile/iris...
0102
Wiep Hamstra @wiep.bsky.social · 09/09/2026
Interesting that theft, abuse, pornography, destruction of knowledge, billions burned, women attacked, teenagers killed did not count so much to the privileged. Because ‘AI is the future’. But now a male and white mathematician is robbed of his work (also bad!), some feel ‘betrayed’. 🤔
15120
Wiep Hamstra @wiep.bsky.social · 06/09/2026
Critique on AI industry is going mainstream, we see devastating effects and BS, so let’s organize a manel instead of inviting experts who are doing this work for years and were silenced or ignored.  💯 agree @irisvanrooij.bsky.social
021
Reposted by Wiep Hamstra
ElizabethwithaZ @elizabethwithaz.bsky.social · 30/08/2026
The AI Literacy we get: "Here's a piece of software founded on theft, labor exploitation, and environmental devastation. It consistently produces bias and falsehoods, so be sure to use it ethically and responsibly!" The AI Literacy we need:
212259
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 29/08/2026
Using LLMs in science for “literature review” or “summarizing" is *automated cryptogyny* olivia.science/cryptogyny/ @olivia.science 🧵
olivia.science
Cryptogyny
On the systematic obfuscation of women's contributions.
3388166
Wiep Hamstra @wiep.bsky.social · 13/08/2026
The ‘magic’ (or: curse) of LLM’s explained. Why and how people think computers speak directly to them. (But also: the stars, nature, horoscopes..) Why ‘AI psychosis’ is not helpful to say when people suffer. And much more. What a great read! 👀 Don’t miss the footnotes 🌟 By @markdingemanse.net
1166
Wiep Hamstra @wiep.bsky.social · 28/06/2026
Nieuw: met ‘AI’ het uitbesteden uitbesteden.
Tekst uit VK-artikel: We hebben een au pair thuis, heel fijn, maar ik was de hele tijd bezig met instructies. Ik geef nu één prompt aan het begin van de week. Kijk, hier zegt hij bijvoorbeeld dat mijn zoon woensdag voetbal heeft en meteen erna de avondvierdaagse, dan krijg ik bericht: hé, kan dat wel?’
110
Reposted by Wiep Hamstra
Prof. Stefan Rahmstorf @rahmstorf.bsky.social · 16/06/2026
Ever since I started on #AMOC stability as a postdoc in 1991, there's been this vexing "cold blob" in the sea surface temperature trend since the year 1900. data.giss.nasa.gov/gistemp/maps/ Over ocean it's ERSST data; for winter since that relates most to ocean heat transport (👉Caesar et al 2018). 🌊
517055
Wiep Hamstra @wiep.bsky.social · 21/06/2026
How to chase a mammoth? And make coats out of the fur? That is human knowledge but it is not on computers. What we know, or think to know, is sometimes decoded and interpreted by decoding forms of language, like pictures in caves. So just apply this on the claims when someone says ‘AI’ knows all.
232
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 21/06/2026
Excellent news
08822
Reposted by Wiep Hamstra
Teresa Heffernan @tjheffernan.bsky.social · 18/06/2026
As studies show, those who don't understand 'AI' are more likely to adopt it, which explains why so many experts are opposed to it and why so many very AI-literate Canadians are refusing to adopt it. Maybe the gov needs to listen to experts and Canadians instead of the AI industry?
07124
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 16/06/2026
Seeing scary entanglement between Dutch technofascist movements and EU level research into AI. Not that I could not guess, but seeing it black on white remains disconcerting ...
3267
Wiep Hamstra @wiep.bsky.social · 17/06/2026
Sometimes basic concepts like ‘machines are not minds’ need some explaining. Especially needed in times when people easily mess up their logic thinking because of false marketing and hype economics around heavy subsided failing software.
0318
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 16/06/2026
It is so important that we collectively push back against these developments. Please read and consider signing our open letter to Dutch Government if you have not yet. cc @lekkerresistance.bsky.social @wiep.bsky.social @marentierra.bsky.social @olivia.science bsky.app/profile/iris...
073
Reposted by Wiep Hamstra
Sonya 🕊️ @sonya.eurosky.social · 12/06/2026
Paywalled, but the city is Taylor, Texas. This makes me so upset that I could breathe fire.
183
Reposted by Wiep Hamstra
Rita, antifascist🏴 @oldsquida.bsky.social · 12/06/2026
“Palantir tracks everyone. So we built a site that tracks Palantir. They've been watching us. Now we can watch them. Contracts. Lawsuits. Executive pay. Global footprint. Lobbying activity. Media strategies. All in one place.”
palantirwatch.org
PalantirWatch — Tracking Palantir Technologies
Independent watchdog tracking Palantir's contracts, executives, lobbying, legal exposure, and global footprint. OSINT-driven public interest research.
2028911249
Wiep Hamstra @wiep.bsky.social · 12/06/2026
I remember people going for lots of Pokémons in one of our biggest academic hospitals, UMCG.
010
Wiep Hamstra @wiep.bsky.social · 31/05/2026
This is what expertise in “AI” looks like. Just sit for one day in a class and get your capgemini zwemdiploma.
1134
Reposted by Wiep Hamstra
Tamara Kneese @tamigraph.bsky.social · 22/05/2026
Pretty good summary of why people are so pissed off about data centers, the physical representations of the technology companies keep saying is going to take everyone’s jobs away
But when Saline residents finally had their chance to speak, a volley of pointed questions followed. They asked about the emissions from the fossil fuels that would be used to generate power for the data center; the bright lights from the construction site shining into children’s bedrooms late at night and before dawn; and local roads that had been rendered dangerouso by constant construction-related traffic. None of these impassioned complaints appeared to engender much sympathy from the board, or indeed from the three Related Digital executives who were sitting against the wall observing the proceedings.
38330
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 21/05/2026
What a breath of fresh air. Good to see at least one university being clear and granting students exactly the space they need to hone their skills. "The use of AI is prohibited for aid in conceptualizing, outlining, drafting, revising, translating, or editing any work submitted for credit.”
law.berkeley.edu
312436
Wiep Hamstra @wiep.bsky.social · 22/05/2026
Een chatbot is een energieslurpende, pratende illusie, een venstertje op falende technologie. Het maakt niet uit of de kantoorchatbot van UWV van de Fransen of Amerikanen is. Praten over soevereiniteit, normaliseert de nutteloosheid van dit soort ‘AI’. ibestuur.nl/data-en-ai/t...
ibestuur.nl
Het maakt niet uit of de kantoorchatbot van UWV Frans is
Een chatbot is een energieslurpende, pratende illusie, een venstertje op falende technologie. Het maakt niet uit of deze van de Fransen of Amerikanen is.
000
Reposted by Wiep Hamstra
Maarten Paulusse @maartenp.bsky.social · 21/05/2026
Once you have a basic understanding of how these LLM-powered chatbots work this shouldn’t be a surprise. It is surprising though that there still doesn’t seem to be widespread alarm about the dangers to democracy these tools pose…
2155
Reposted by Wiep Hamstra
Maarten Paulusse @maartenp.bsky.social · 20/05/2026
Good piece by Jefferson Pooley on how LLM-powered “research” tools like Google’s “Deep Research” and Elsevier’s “AI Discovery” likely strengthen the “Matthew effect”: highly-cited authors are amplified while undercited (often female) authors are increasingly ignored: blogs.lse.ac.uk/impactofsoci...
blogs.lse.ac.uk
The Matthew effect in AI summary - LSE Impact
AI research tools are trained on a literature that is structured by an unequal distribution of attention. Does their use simply recreate these biases ?
512668
Reposted by Wiep Hamstra
Mark Dingemanse @markdingemanse.net · 14/05/2026
Destined to be a classic
1196
Reposted by Wiep Hamstra
hooked on persephonics 👁 @amphote.bsky.social · 14/05/2026
And human-level cognition *already exists*, there are real live people you could use those resources to feed, clothe, house, and pay to think
34312
Reposted by Wiep Hamstra
Boxo McFoxo @boxobark.ing · 13/05/2026
Yes, this is the biggest harm of a mind being projected into the chatbot. Ever since they became a thing it's worried me that summarisation is considered to be a simple task that they are suited for. Even people who would say that they know an LLM isn't thinking will say that it can summarise.
2183
Reposted by Wiep Hamstra
Adolfo Neto @adolfoneto.elixiremfoco.com · 14/05/2026
Important: "Are you a developer at Google, Microsoft, or another tech being pressured to use AI? I would love to hear from you. Using a non-work device, you can message me securely on Signal at ‪(609) 678-3204‬. Otherwise, send me an email at emanuel@404media.co."
111061
Reposted by Wiep Hamstra
Cabel Sasser @cabel.panic.com · 11/05/2026
this graduation speech moment is notable, and her amazed shock at having failed to read the room feels instructive. when you're inside the bubble, you think everybody else is. but everybody isn't.
992272926879
Wiep Hamstra @wiep.bsky.social · 10/05/2026
Why are the TED-talk people so much into trying to ‘understand’ Dawkins and his misogynistic ‘Claudia’. It is not a ‘genius mistake’ by a ‘great male thinker’ with a dementic flaw in his eighties. It really is a very stupid and shameful error that shows that there is focus on the wrong things.
000
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 09/05/2026
Say, hypothetically, an institute that manages your money informs you that ~80% of their employees use MS Copilot. How would that make you feel?
10202
Reposted by Wiep Hamstra
Ketan Joshi @ketanjoshi.co · 07/05/2026
We tried to warn you about Anthropic. Now they're using xAI's ultra-polluting community-killing nightmare data centre to power your toxic claude relationship All the new paid subs from @rutgerbregman.com and @quitgpt.bsky.social's campaign, now powered by Musk 🫠 www.anthropic.com/news/higher-...
New compute partnership with SpaceX
We’ve signed an agreement with SpaceX to use all of the compute capacity at their Colossus 1 data center. This gives us access to more than 300 megawatts of new capacity (over 220,000 NVIDIA GPUs) within the month. This additional capacity will directly improve capacity for Claude Pro and Claude Max subscribers.

This joins our other significant compute announcements:

An up to 5 gigawatt (GW) agreement with Amazon, which includes nearly 1 GW of new capacity by the end of 2026;
A 5 GW agreement with Google and Broadcom, which will begin coming online in 2027;
A strategic partnership with Microsoft and NVIDIA that includes $30 billion of Azure capacity;
Our $50 billion investment in American AI infrastructure with Fluidstack.
We train and run Claude on a range of AI hardware—AWS Trainium, Google TPUs, and NVIDIA GPUs—and continue to explore opportunities to bring additional capacity online.

As part of this agreement, we have also expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute capacity.
20833379
Wiep Hamstra @wiep.bsky.social · 25/04/2026
HITL is presented as something mechanical, like serial, separated stages in the process. With an attentive, smart, watching human. I see it as a chemical process. Like baking a cake with rotten eggs and then when it is out of the oven, try to get the rotten ones out. 1/2
210
Reposted by Wiep Hamstra
Faolan @brujofaolan.bsky.social · 22/04/2026
DENMARK AND FRANCE 🇩🇰🇫🇷 bans all Microsoft products from all public offices. The phaseout is expected to be completed before autumn 2026. Their products are being replaced with customer versions of Linux and LibreOffice. This is digital sovereignty being taken back
510148483865
Reposted by Wiep Hamstra
Robert Heinze @rheinze.myatproto.social · 24/04/2026
I'm telling you at this point I am more disturbed by the people "experimenting" with the ethical limits of AI than by the fact that AI has none.
36512
Reposted by Wiep Hamstra
Hypervisible @hypervisible.blacksky.app · 24/04/2026
“If you ask tobacco companies to help write your school’s policy on cigarettes…you’re going to end up with guidance on how to smoke responsibly in school.”
newyorker.com
What Will It Take to Get A.I. Out of Schools?
“I find myself speaking with my kids about A.I. in the same terms that we might discuss a creepy neighbor who lives down the block,” Jessica Winter writes. Read her report on the push for A.I. in educ...
111342514
Wiep Hamstra @wiep.bsky.social · 12/04/2026
“Die verwachtingen [veronderstelde voordelen van AI in onderwijs] zijn volgens hen echter grotendeels gebaseerd op aannames, niet op stevig empirisch bewijs.” Onderzoek: Dagmar Mercedes Heeg en Lucy Avraamidou, artikel van @fransvanheest.bsky.social www.onderwijs-wetenschap.nl/geen-bewijs-...
onderwijs-wetenschap.nl
Geen bewijs voor leerwinst, wel risico op ongelijkheid door AI - Onderwijs & Wetenschap
Groningse onderzoekers hebben geen overtuigend wetenschappelijk bewijs gevonden dat AI-systemen het leren daadwerkelijk verbeteren. Tegelijkertijd laten de onderzoekers zien dat AI discriminerende eff...
063
Reposted by Wiep Hamstra
Iris van Rooij 💭 @irisvanrooij.bsky.social · 12/04/2026
“check out this great interview w/a.o. @olivia.science & @irisvanrooij.bsky.social (…) The whole piece is strong, but the beginning is truly fantastic, when Guest explains how the Q of whether something might work is completely irrelevant ("possibilities don't bake bread”).” — @felienne.bsky.social
22311