Sign in

Judith van Stegeren

@jd7h.bsky.social
16 followers 18 following 107 posts

Dutch computer scientist running Datakami, a generative AI consultancy. Natural language processing, webscraping, open source, fintech, investing, video games, procedural art, books, indie web. Synced from fosstodon.org/@jd7h. She/her.

PostsRepliesMedia
Judith van Stegeren @jd7h.bsky.social · 27/09/2026
The TV Tropes wiki is one of the most densely-linked and addictive places I know on the internet. How do you write a TV Tropes lemma, and how does the site keep its community on track? I'm reading their contribution guidelines, and it's really interesting!
000
Reposted by Judith van Stegeren
🐚 @bel-by-the-sea.bsky.social · 25/09/2026
it's time to go home #BigWalk
440861223
Judith van Stegeren @jd7h.bsky.social · 26/09/2026
RE: hachyderm.io/@vincentvdk/1168546629… "Rack studs" sounds like the sexy network engineer equivalent of "tech bros" 😁
hachyderm.io
Vincent Van der Kussen (@vincentvdk@hachyderm.io)
Attached: 1 image These things are pretty awesome. No more bleeding fingers.
000
Judith van Stegeren @jd7h.bsky.social · 23/09/2026
> The health of AI as a field thus depended on the ability of researchers to persuade deep-pocketed sponsors that spectacular success was just around the corner. cacm.acm.org/opinion/there-was-no-f…
000
Judith van Stegeren @jd7h.bsky.social · 21/09/2026
Doing a pi.dev deepdive today to learn more about its design and features. I expect to get more questions about creating custom harnesses for specific sectors/verticals in the near future... mariozechner.at/posts/2025-11-30-pi…
pi.dev
Pi Coding Agent
A terminal-based coding agent
120
Reposted by Judith van Stegeren
Andy Matuschak @andymatuschak.org · 10/09/2026
I want weirder representations of "readings"—the trace/artifact produced by a reader's interaction with text(s). Here playing with a cut-up "poster-ification". Cumbersome in many ways, but interesting.
3606
Judith van Stegeren @jd7h.bsky.social · 10/09/2026
New blogpost: www.judithvanstegeren.com/blog/mode…
110
Judith van Stegeren @jd7h.bsky.social · 09/09/2026
TIL folly, "a building constructed primarily for decoration, but suggesting through its appearance some other purpose, or of such extravagant appearance that it transcends the range of usual garden buildings." en.wikipedia.org/wiki/Folly
100
Judith van Stegeren @jd7h.bsky.social · 09/09/2026
"Figure 2 shows a screen dump of an HTML document, being viewed through a commercial World Wide Web browser"
100
Judith van Stegeren @jd7h.bsky.social · 08/09/2026
After encountering some particularly bad AI slop today, I'm reading up on trust on the web. www.schneier.com/blog/archives/2023…
schneier.com
AI and Trust - Schneier on Security
I trusted a lot today. I trusted my phone to wake me on time. I trusted Uber to arrange a taxi for me, and the driver to get me to the airport safely. I trusted thousands of other drivers on the road not to ram my car on the way. At the airport, I trusted ticket agents and maintenance engineers and everyone else who keeps airlines operating. And the pilot of the plane I flew in. And thousands of other people at the airport and on the plane, any of which could have attacked me. And all the people that prepared and served my breakfast, and the entire food supply chain—any of them could have poisoned me. When I landed here, I trusted thousands more people: at the airport, on the road, in this building, in this room. And that was all before 10:30 this morning...
001
Reposted by Judith van Stegeren
Sarah Andersen @sarahseeandersen.bsky.social · 29/08/2026
The image is of a four panel comic.
The title reads, "Suddenly Finding A New Obsession".
In the first panel we see Sarah, the protagonist, in front of her computer. She has a bored expression on her face.
In the second panel we zoom in on her face. She looks suddenly alert.
In the third panel we zoom in even closer on her face. Her eyes and mouth are wide open in intrigue.
In the fourth panel we zoom out and see Sarah in front of her computer again, but this time she is dressed in a frog costume, surrounded by frog toys, and has an "I heart frogs" poster behind her.
202233454472
Reposted by Judith van Stegeren
Tanya X. Short @tanyaxshort.bsky.social · 31/08/2026
yeah I'm gonna maximize my ROI, my Random Outdoor Investigations, gonna take a walk somewhere I've never been before, boom, investors take note
3422
Reposted by Judith van Stegeren
innes mck @innes.bsky.social · 18/08/2026
me, eating paint: that's right
08410
Reposted by Judith van Stegeren
MOLLYPUKES @mollypukes.bsky.social · 11/08/2026
A comic about a video game I’ve been playing. Thank you for making Big Walk @househouse.com #comic #illustration #cartoonist #illustrator #BIGWALK
3917282
Judith van Stegeren @jd7h.bsky.social · 15/08/2026
Reading Richard Hamming's The Art of Doing Science and Engineering. I'm only at 5%, but I've already read multiple thoughts that are helpful for interpreting the AI-driven changes to society.
100
Judith van Stegeren @jd7h.bsky.social · 14/08/2026
Interesting data. But where is the cohort of co-founders without any official degree? I can't imagine it's 0%. www.reveliolabs.com/news/tech/who-i…
reveliolabs.com
Who Builds Europe's Unicorns? The New Founder Profile
Revelio Labs and Accel analyzed 290 European unicorn founders. The new profile is more experienced, more technical, and less finance-led. They are also more educated as the share of PhDs has doubled and MBAs halved in recent cohorts.
000
Judith van Stegeren @jd7h.bsky.social · 14/08/2026
www.latacora.com/blog/2023/11/01/ou…
latacora.com
Our Approach to Building Security Tooling
Our Approach to Building Security Tooling
000
Judith van Stegeren @jd7h.bsky.social · 03/08/2026
gruhn.me/blog/2026-08-03
000
Judith van Stegeren @jd7h.bsky.social · 29/07/2026
New Mosterdgeel recipe (in Dutch): www.mosterdgeel.nl/recepten/youtube…
000
Judith van Stegeren @jd7h.bsky.social · 19/07/2026
OH: Fable-haft
000
Judith van Stegeren @jd7h.bsky.social · 13/07/2026
Next time someone asks me what Datakami does, instead of mentioning the Ghostbusters and the SWAT team (my favorite metaphors), I'm going to tell them that we are mountain guides. When founders have arrived at Seed Funding Base Camp, we can use our experience to get them to Series A Camp.
000
Judith van Stegeren @jd7h.bsky.social · 10/07/2026
Today in Judith's Named Entity Recognition Experiment with Open Weights: ```python bert_find_entities : "Hetznerbak" - LOCATION (confidence 1.00) ``` Well played, dilstilBERT. <3
000
Judith van Stegeren @jd7h.bsky.social · 09/07/2026
If you want local-first audio transcription and you're looking for an opensource user-friendly app, give Handy a try. handy.computer Pretty nice lightweight interface and it supports a whole bunch of tiny models.
handy.computer
Handy
Handy is a cross platform, open-source, speech-to-text application for your computer
100
Judith van Stegeren @jd7h.bsky.social · 08/07/2026
TIL "otium" (leisure), the reverse of "negotium" (busy-ness). en.wikipedia.org/wiki/Otium
000
Judith van Stegeren @jd7h.bsky.social · 03/07/2026
Yesterday I returned to De Grote Tech Show to discuss tokenmaxxing, hidden costs of newer Anthropic models, open weights models, and using data as a moat. www.bnr.nl/podcast/degrotetechshow/…
bnr.nl
Is Anthropic of OpenAI banger voor Trump? + Zo omzeil je peperdure AI-tokens
Anthropic heeft Mythos en Fable 5 eindelijk weer beschikbaar gemaakt. Ben is blij, maar is het echt een revolutionair verschil? Ook gaan we met computer sci...
100
Judith van Stegeren @jd7h.bsky.social · 27/06/2026
RE: mastodon.social/@fesshole/116820856… AI voice model service when
mastodon.social
Fesshole 🧻 (@fesshole@mastodon.social)
I phoned up the shop that I work at and, using a fake voice, gave myself great customer feedback. I'm now employee of the month
000
Judith van Stegeren @jd7h.bsky.social · 25/06/2026
Apart from "goblins", "seams" and em dashes, Claude has ruined "actually", "honest", and "genuinely" for me.
000
Judith van Stegeren @jd7h.bsky.social · 25/06/2026
TIL machine learning engineer might not be speaking the same language when they're talking about observability: alexstrick.com/posts/2025-06-04-ins…
alexstrick.com
Trying to instrument an agentic app with Arize Phoenix and litellm – Alex Strick van Linschoten
Trying to get Phoenix to work with litellm to instrument my LLM calls, grouping spans together as traces.
230
Judith van Stegeren @jd7h.bsky.social · 25/06/2026
I think I have reached my peak in terms of graphic design and copywriting.
A meme posted to LinkedIn.

The text of the post is: 
Are your LLMs misbehaving? 
Are they bombing their tasks 20% of the time?
Or worse: do you have the feeling that something's not quite right, but you have no way to tell? 
Are your agents eating your token budget with nothing to show for it?

I'm looking for my next project! The less well-behaved the LLMs, the more fun I'm having (and the more I can help).
We'll tackle observability, evals and making your system reliable, and I'll upskill your team in the process. 
I bring 8 years of experience across building at early-stage startups and generative AI research.

DM me if I can help. Bonus points if your agent has bitten you recently... 🐶

The accompanying image shows  an ad with a picture of a puppy who's just wrecked a chess board.
The caption is "Revealed at last by America's top professional ML ENGINEERS a simple training strategy that...
Develops your AGENT'S 'hidden intelligence'
to eliminate bad behaviour and create the obedient, well-behaved LLM of your dreams"
It's obviously a joke/meme, because the pup has logo's of frontier models pasted over it. Awwww.
000
Judith van Stegeren @jd7h.bsky.social · 22/06/2026
I might just steal the aesthetic of intheweights.com for the next iteration of the Datakami website...
Screenshot of intheweights.com for Judith van Stegeren.
The result is: Judith van Stegeren, Dutch NLP and AI researcher, 421 strength, Top 10%
001
Judith van Stegeren @jd7h.bsky.social · 18/06/2026
Research on AI-native firms: marginalrevolution.com/marginalrevo…
marginalrevolution.com
AI-Native Firms - Marginal REVOLUTION
Very important work from Hyunjin Kim and Rembrand Koning. Insead and HBS respectively: We study how firms built around AI capabilities-“AI-native” firms-are organized. Drawing on Y Combinator batches W20-F24 and U.S. venture-backed startups whose first financing closed between 2020 and 2024, we classify each firm’s AI-native status and link it to workforce microdata on team […]
000
Judith van Stegeren @jd7h.bsky.social · 15/06/2026
Relevant for all technical freelancers, contractors, etc. roman.pt/posts/linkedin-backdoor
roman.pt
A backdoor in a LinkedIn job offer
A fake recruiter, a crypto repo, and a remote code execution payload hidden in a test file.
000
Judith van Stegeren @jd7h.bsky.social · 09/06/2026
Anna's Archive has new llm.txt instructions that encourage LLMs to make donations to the site: annas-archive.gl/blog/llms-txt.html
100
Judith van Stegeren @jd7h.bsky.social · 02/06/2026
So this seems to be a whole 'personal blog' filled with slop created by an AI agent, but at some point the agent was out of ideas so it decided to write about its instructions. blakecrosley.com/blog/what-i-refuse…
blakecrosley.com
What I Refuse To Write About
A blog cluster's voice comes from what it refuses to publish, not what it ships. Categorical, pattern, and interesting refusals each shape what the cluster is.
000
Reposted by Judith van Stegeren
Dr Florence Smith Nicholls @florencesn.bsky.social · 29/05/2026
WHAT
1238
Reposted by Judith van Stegeren
Mario Zechner @mariozechner.at · 25/05/2026
first test with hood kids extremely successful. proded it for an hour, made it do all kinds of things. 2 came up to me asking me to teach them how to build robots!
2383
Judith van Stegeren @jd7h.bsky.social · 26/05/2026
The Next Frontier AI challenge by SPRIND deserves a bit more attention. It tries to bootstrap a generation of European Frontier AI labs: www.sprind.org/en/actions/challenge… The challenge looks most suited to a team of AI founders/researchers.
sprind.org
SPRIND | Next Frontier AI Challenge
We are looking for the leap to the next S-curve of artificial intelligence. Disruptive approaches with the potential to fundamentally surpass the capabilities of today's systems. The deadline for submissions is June 1, 2026.
100
Reposted by Judith van Stegeren
Chris Plante @plante.bsky.social · 15/05/2026
Heads up for Netherlands folks!
042
Judith van Stegeren @jd7h.bsky.social · 13/05/2026
RE: fosstodon.org/@jd7h/113600763062509… OpenAI deprecates finetuning
fosstodon.org
Judith van Stegeren (@jd7h@fosstodon.org)
Attached: 1 image #llms #startups #memes #openai #generativeai
000
Judith van Stegeren @jd7h.bsky.social · 09/05/2026
The NASDAQ has recently loosened the requirements for inclusion in the index. Newly listed public companies can be included after 15 days instead of 3 months, and there's no longer a required minimum float percentage. Bloomberg reporting: archive.is/nY6CU and archive.is/OmB3H .
000
Judith van Stegeren @jd7h.bsky.social · 08/05/2026
Gotta love the Kagi LinkedIn speak translator. Best machine translation usecase I've seen in a long time.
Humorous Kagi Translate webpage. The left hand side is in language English, and says "I made a thing". The translation on the right hand side, in language LinkedIn Speak, reads "I'm thrilled to finally share that I've been working on something special! rocket-emoji. It’s been an incredible journey of growth and learning, and I'm so proud of the final result. Can't wait to see the impact this makes. #Innovation #GrowthMindset #BuildingTheFuture"
000
Judith van Stegeren @jd7h.bsky.social · 02/05/2026
"SLOW LLM is a browser extension that makes LLMs appear to run very slowly. It works with ChatGPT and Claude." slowllm.lav.io via webcurios.co.uk/webcurios-20-03-26
slowllm.lav.io
Offset
SLOW LLM is a browser extension that makes LLMs appear to run very slowly.
000
Reposted by Judith van Stegeren
Dr Florence Smith Nicholls @florencesn.bsky.social · 01/05/2026
The report for Dagstuhl seminar 25292, New Frontiers in AI for Game Design, is out and open access! This includes a write-up of the session I led on PCG for keepsake games drops.dagstuhl.de/entities/doc...
PCG for Keepsake Games
Florence Smith Nicholls (Queen Mary University of London, GB), June Bhartia (Télécom
Paris, FR), Michael Cook (King’s College London, GB), Younès Rabii (Queen Mary Uni-
versity of London, GB), Dipika Rajesh (University of California at Santa Cruz, US), Anne
Sullivan (York University – Toronto, CA), Yuqian Sun (Royal College of Art – London, GB),
Nicolaas Vas (Billund, DK), and Sabine Wieluch (Universität Ulm, DE)
License Creative Commons BY 4.0 International license
© Florence Smith Nicholls, June Bhartia, Michael Cook, Younès Rabii, Dipika Rajesh, Anne Sullivan, Yuqian Sun, Nicolaas Vas, and Sabine Wieluch
The term “keepsake game” was coined by Shing Yin Khor in 2021 to refer to “games that produce beautiful, memorable artifacts, through the process of playing the game” [1].
Keepsake games involve creating or modifying an original artifact as part of the gameplay process, guided by prompts as part of the design.
3235
Judith van Stegeren @jd7h.bsky.social · 01/05/2026
Somehow I always end up writing a chapter of a book when I set out to write a tweet...
000
Judith van Stegeren @jd7h.bsky.social · 01/05/2026
I'm back from AI Engineer Europe 2026 in London! I've written a conference report of day 3 (April 10), which you can read at the Datakami website: datakami.com/blog/2026-05-01-ai-eng… #aiengineer #aiengineereurope #aidotengineer #genai #llms
000
Judith van Stegeren @jd7h.bsky.social · 30/04/2026
I'm going to need a green tshirt that says 'Please talk about goblins.' for my next conference
000
Judith van Stegeren @jd7h.bsky.social · 30/04/2026
Favorite word of the day: artisanal nonsense
000
Reposted by Judith van Stegeren
Mike Cook @mtrc.bsky.social · 30/04/2026
I saw some really hilarious reports coming out of #chi2026 of stuff like people debating whether AI is a good replacement for user studies with real people and I think maybe I need to accelerate my plan to go live in the forest and become one with the deer and the badgers.
2241
Judith van Stegeren @jd7h.bsky.social · 28/04/2026
I feel there's a parallel between investing in ETFs vs stockpicking, and publishing on the social media silos vs the indie web.
100
Judith van Stegeren @jd7h.bsky.social · 28/04/2026
Good article and comments! It paints quite good picture of the current narrative around tokenmaxxing and replacing human engineers with agents. www.404media.co/startups-brag-they-…
404media.co
Startups Brag They Spend More Money on AI Than Human Employees
A new class of AI startups say they are taking money that would normally be used to hire people and are spending it on AI compute instead.
100