communicates the notion that the debt compounded from going fast lives in the brains of the developers and affects their lived experiences and abilities to “go fast” or to make change
Cognitive debt here described that it comes from going fast during dev, resulting in reduced ability to make changes / go fast in the future, bc of affected experiences and abilities of the people involved.
Cognitive debt, a term gaining traction
Technical debt nicely captures that “human understanding” also matters, but the words “technical debt” conjure up the notion that the accrued debt is a property of the code and effort needs to be spent on removing that debt from code.
While techdebt is about the accumulation of human decisions, and the resulting erosion of human understanding, the term itself suggests it is a property of the code itself, and that one could remove it from code.
The term technical debt is often used to refer to the accumulation of design or implementation choices that later make the software harder and more costly to understand, modify, or extend over time
Techdebt def used in this article. the sum total of choices made over time that make software harder to understand / maintain / modify over time. (Is there a measure for it, when do you decide to replace e.g.? #openvraag)
By Margaret-Anne Storey, Professor of Computer Science, University of Victoria and Canada Research Chair in Human and Social Aspects of Software Engineering https://en.wikipedia.org/wiki/Margaret-Anne_Storey
added to feedreader
I've experienced this myself on some of my more ambitious vibe-code-adjacent projects. I've been experimenting with prompting entire new features into existence without reviewing their implementations and, while it works surprisingly well, I've found myself getting lost in my own projects. I no longer have a firm mental model of what they can do and how they work, which means each additional feature becomes harder to reason about, eventually leading me to lose the ability to make confident decisions about where to go next.
Vibecoding and adjacent projects lead to loosing overview of your own work, no mental model of what you made as you would have otherwise. Extending something becomes harder over time, bc you don't know what you're actually extending from. This is a counter force (not counter argument I think) to the notion of genAI having deterministic automation as endpoint.
https://web.archive.org/web/20260215105347/https://simonwillison.net/2026/Feb/15/cognitive-debt/
Simon Willison on cognitive debt (the consequences of vibecoding more or less).
Study on Cognitive Debt, via [[Peter Rukavina p]]
This study explores the neural and behavioral consequences of LLM-assisted essay writing.
This piece does not look at cognitivedebt where I saw it mentioned most (i.e. around the effect of vibecoding), it looks at using LLMs in writing (essays)
Dana Library Hand, a public domain dedicated 'handwritting' style font, after a 1899 disjoined handwriting sample meant for librarians when writing index cards. Mentioned in a Mastodon mention. Downloaded.
Digitaal Stelsel Mobiliteitsdata (DSM)
Noemt NDOV als tactische tafel. Noemt NDOV loket en Stichting OpenGeo als open data CC0 bron hiervoor. Noemt de diverse overlegtafels onder DSM
openov.nl zegt open data bron te zijn voor OV gegevens, en heeft daarvoor OVapi, maar het oogt een beetje verlaten.
Palantir is suing a Swiss media start-up for their FOIA based investigative journalism, and showing that Palantir tried for 7 yrs to engage CH as client, and that the Swiss Army evaluated the software as not fit for purpose, in part bc it would mean Swiss data would leak the US gov. Palantir is seen as too close to the US admin (bc of Thiel mostly I suppose), and the company's reputation is taking a dent amongst potential European clients
A description of the type of things Activate does w AI agents. Ref'd in [[Paolo’s Weblog It’s still time for free AI love]]
we might move again. The point is that we can. We can because we own our prompts, our skills, our databases, our memory architecture, they all live in our bar. None of it lives inside OpenAI or Anthropic. When we moved, we rewired the model layer and everything else stayed put. That’s the whole trick, really. If you control the pieces that make your agents smart, switching the engine underneath is just plumbing.
Description of how Activate keep their prompts, skills, databases, memory architecture under their own control and within their own environment.
Moving means wiring up another model or models, but the rest is kept as is.
[[Paolo Valdemarin p]] on avoiding lock-in in a single specific AI silo. Bc what is state of the art will shift every few months.
On Wero, which is going live for every iDeal transaction in the Netherlands the coming months. This will def reduce my online card usage.
What if I actually did have dirt on me that an AI could leverage? What could it make me do? How many people have open social media accounts, reused usernames, and no idea that AI could connect those dots to find out things no one knows?
AI agents as kompromat collectors
Fabio Manganiello made a webmention library for python, so it can be easy to add webmention to static sites etc.
Articles and books by [[Paul Nemitz c]] (als co-authoring w [[Matthias Pfeffer p]]
this reminds me of installation art that is just a prescription / recipe of how to recreate it.
A collection of 'curated' prompts wrt openglam. Ref'd by [[Hessel van Oorschot p]] in the context of the Apertus llm by ETH in ZH.
UN open source week, 22-26 June, in NYC. probably streamed, 2025 sessions are available on video
The UN Open Source Principles Open by default: Making open source the standard approach for projects. Contribute back: Encouraging active participation in the open source ecosystem. Secure by design: Making security a priority in all software projects. Foster inclusive participation and community building: Enabling and facilitating diverse and inclusive contributions. Design for reusability: Designing projects to be interoperable across various platforms and ecosystems. Provide documentation: Providing thorough documentation for end-users, integrators and developers. RISE (recognize, incentivize, support, and empower): Empowering individuals and communities to actively participate. Sustain and scale: Supporting the development of solutions that meet the evolving needs of the UN system and beyond.
UN FOSS principles https://unite.un.org/en/news/sixteen-organizations-endorse-un-open-source-principles (ao signed by OKF, Rocketchat, Nextlcoud in 2025)
We're in the habit of exporting public money out of Europe, and importing dependencies.
This how Schleswig Holstein minister for digital transformation Dirk Schrödter phrases it: exporting public money, importing dependencies.
So, for now, Schleswig is still running Windows, but has replaced Microsoft Office with LibreOffice on "nearly 100 percent" of machines. It's using Thunderbird for email, calendars, and contacts, talking to Open-Xchange on the back end, alongside the Matrix-based Element for chat and Nextcloud for collaboration.
Schleswig-Holstein is still using Windows, taking it step by step. MS Office to LibreOffice, Thunderbird w Open-Xchange for mail, calendar, contacts, Element/Matrix for chat, Nextcloud for collab.
Brine also contains metals and chemicals (foreign to the marine environment) that have adverse effects on marine ecosystem
Brine also concentrates other materials, metal, chemicals. (at what point is brine dense enough to harvest such things)
Because energy for desalination increases with seawater salinity, buildup of salt in brine endangers potable water supply there.
pushing back brine into the sea means higher salinity and thus more energy needed for desalination.
The Persian Gulf (also known as Arabian Gulf) is surrounded by desalination plants with about 50% of worldwide capacity to desalinate seawater. Most of these plants dispose of hypersaline effluent (brine) via surface and nearshore outfall into the Gulf
some 50% of global desalination capacity is around the Persian gulf
more viable fixes such as modernizing irrigation or repairing urban water networks that lose nearly 30% of supply to leaks.
urban water networks lose 30% of supply to leaks.
Groundwater extraction exceeds natural recharge by two- to threefold in many plains, causing land subsidence of 25–35 centimeters per year and collapsing centuries-old qanats.
groundwater extraction is 2 or 3 times larger than replenishment. Subsidence of 25-35 cm p yr. Old qanat systems collapse, increasing the problems.
Agriculture consumes roughly 90% of Iran’s water while contributing only about 12% of GDP, a ratio driven by inefficient flood irrigation and water-intensive crops like rice cultivated in arid basins.
Iran's agri uses 90% of water but is 12% GDP. Driven by flood irrigation (not drips e.g.) and crops that use lots (rice) grown in arid areas
The pipeline itself crosses some of Iran’s driest terrain, where heat and exposure could evaporate 20–30% of the transported water, while brine residues risk contaminating fragile soils.
evaporation during transport an issue
Desalination produces hypersaline brine, which, if discharged improperly, could raise salinity in parts of the Gulf of Oman by up to 1.5 times and increase water temperatures by 2°C, threatening marine ecosystems and already stressed fisheries.
Issue w desalination is the resulting brine and its disposal. How does Israel (a net exporter of desalinated water) deal with that?
Iran desalination project, for the Isfahan region. #2025/12
Curl is closing bug bounty program, bc they got buried in AI slop submissions. Also mentions an example of AI assisted bug hunting that was fruitful, to contrast it with lazy stuff.
https://web.archive.org/web/20260206135535/https://felipezamana.com/perpetual-beta/
Interview with [[Harold Jarche p]] on PKM etc.
PKM isn’t necessarily about disorientation, but it is a tool for exploration. You can then get to a point of reorientation, which may not be complete, and then be able to take action or make it actionable.
PKM as tool for exploration and perhaps later reorientation, towards action / actionable things. Vgl [[Actionable sense als groep 20200801065550]] blog disc some of us had in 2003.
I left Facebook 10 years ago and left Twitter 2 years ago. The only platform I’m on now is LinkedIn. And even there, some days I think that maybe I should get off it, too. I mean, the algorithm has changed so much. Here’s an example: a post I put on LinkedIn about 3 years ago got 29k views. Now, I’m lucky to get 1k views. That’s what the algorithm is. I have around 7k followers on LinkedIn, and I’m only getting 1k views? You know that the algorithm is just squeezing it down, and I’m not a paying member of LinkedIn, so…
Harold left FB 2015, Twitter 2023. As ambivalent about LinkedIn as I am. Says his engagement there is down by a factor 30. Only one in 7 of his contacts even sees his post.
Vgl [[Early Adoption - Early Leaving]]
You know, the e-books basically were a way for somebody who wanted to read my content in a book format, with a flow to it and things like that.
his ebooks as blogposts put together in a flow.
You mentioned that many of your blog posts begin as small moments of curiosity and questioning about the world around you and your clients’ experiences.
blogging as sense of wonder
No one wants to change.
unless there's a tied-to-emotion trigger.
I think I’ve touched some people individually. I’ve had some successful projects where we actually got something where we could see the benefits of what we were doing.
I've come to the conclusion that touching individuals is most of all one can do. And then spread that wrt [[Effectief gedrag is besmettelijk 20200831071537]]
being comfortable with not knowing is part of blogging. Sometimes I’m putting stuff out there, and I don’t know if it’s any good, I don’t know if it makes any sense, but I gotta get it out there, and let’s see what happens. There’s a lot of stuff that doesn’t go any further than a blog post. It is what it is. It’s my process of trying to make sense of things.
blogging as being comfortable with not knowing, holding questions. This comes close to writing as thinking itself.
being comfortable with being disoriented
holding questions
Dave Snowden? It’s a sense-making framework as well, but there’s a part where Dave talks about aporia: a state of puzzlement or disorientation, until you finally get to the point where you actually understand what you don’t know. First, you don’t even know what you don’t know, because everything is confusing. And then, through exploration, you get to a certain point where you think, “Okay, I don’t understand this, now I’ve got to learn about that.” So, I actually know what I don’t know, and now I can go into it and take some action. That’s where the actionable part comes. And part of that, too, is being comfortable with being disoriented.
refs [[Dave Snowden p]] aporia, disorientation as trigger to get a sense of what it is you don't know, through exploration. To get to the point where your knowledge need becomes actionable
Well, PKM is not actually my idea. There are other people who are writing about personal knowledge management. I changed it to mastery because I wanted to move away from the knowledge management world, which was too much about big systems and big databases. I wanted to focus on what I, as an individual, do, and what we as a community or a network do. For example, how do we enable that kind of collaboration and cooperation? So, personal knowledge mastery just became the term because it is about mastery, and you never master it completely, right? It’s like any discipline, a lifelong thing. You continuously try to get better.
how Harold replaced the m in pkm for mastery. This comes very close to my [[Kenniswerk is ambacht 20040924200250]] artisanal view on knowledge work. The focus on practice in community, networks and on your own. Vgl [[% Practice Praktijk OP]]
Right now, I have a book in progress with Clark Quinn. Clark and I have known each other for 20 years, and it’s based on PKM, but it’s more of a how-to manual, right down to the actual process of personal knowledge mastery. The working title is Seek and Share, but we’ll see where it goes. So, we’ve been working on that for several months now.
Harold is working on a book on PKM. Seek and Share is working title.
The e-books that I published, for example, are years’ worth of blog posts, where I take the best ones, synthesize them, and put them together into a better product. The first book that I published was after 10 years of blogging. I took all these 10 years’ worth of content and launched the Perpetual Beta series. Then, I’ve added several to that. The last e-book was launched in 2024, and that was more of a questioning of what the heck is going on, because that was where AI was starting to rear its ugly head and everything.
Harold used his blog as source for creating ebooks on diff topics. #idea Vgl jaarboeken v blogarchief
“Do I keep writing through my retirement or not?” I haven’t done any major consulting lately. I’m running my workshops, doing some writing, hosting my community, and taking vacations. So, who knows?
Harold is nearing the end of his working life, what does it mean for his blogging? If it is a signboard more, or a professional writing outlet, then yes it may fall away. But if it is your primary space for expression, whatever the topic then the blog can morph with it, no?
other thing is that, given the state of the world right now, every once in a while, I question if what I’m writing about is really important. Who knows?
signals the relative triviality of a lot of his writing in face of geopolitical upheaval. I feel him but also fully disagree, clearly signalling your humanity in the face of it all is key. vgl othering. life is 'small' by def. There is no 'need' to be part of the 'big' discussions for a blog to matter. Cat pictures ftw.
Quite often, I’ll share a draft in my private community before I post it publicly. I think the only real change to my writing style and my focus is that I was doing less of it.
the groups are used for sharing drafts, meaning the blog contains more 'finished' writing in contrast. Imo this means the blog is less about wonder, curiosty and more about performance / presentation?
Well, I’m writing more in private communities. For example, I host a private community with about 40 global members. It’s focused on knowledge management, collaboration, sense-making, that kind of stuff. Because it’s private, we share a lot more there. We do a monthly Zoom call, and then we have a Slack space where we chat. I also belong to two other private groups like that.
Harold in response: withdraws to closed / cosy spaces with group. Not on the open web it seems. Does that make his blog more a signboard than a conversational space?
how do you ensure your blog remains an authentic space for human sensemaking and professional development? Also, has your perception of writing changed somehow because of that?
question how to keep blog as authentic space, for sensemaking/learning. And if it changes writing in the context of ai slop
There aren’t many people who’ve been blogging for 20+ years as I have. I don’t know if that helps a bit or not.
most early bloggers stopped. Impacting imagined audience? Has Harold, have I changed our imagined audience?
this year, I decided that I would get back in the game. My objective is to write one blog post per week. I’ll be happy if I can do that. But I’m also conscious that anything that I put up is going to be scraped, which makes me sort of think that I’m feeding the beast, but there are a number of people who have asked me to keep writing.
2026 decided to blog more again despite the aicrawlers. Can relate. I realised that my primary goal for blogging is distributed conversations as it was at the start, so whatever else happens is a 'don't care'.
I also started seeing things like my website getting scraped by Google and LLMs, which made me drop my writing a lot.
LLMs scraping made him write less.
My blogging activity peaked about 5 years ago. Around that time, I started seeing this effect of enshittification of online platforms. Before Musk bought Twitter, it was the number one source of visitors to my blog. I had 20k followers on Twitter at the time, and I noticed that these platforms were starting to use the algorithm, filtering things down.
Says his blogging peaked 5 yrs ago (2020-21, pandemic?) then saw enshittification impact traffic
It became my business card: “Do you want to know something about me? Go to my blog, read some articles, see what you think, and then maybe you want to hire me after that.”
Vgl [[Blogs als avatar 20030731084659]] and being hired by RWS
In those early days, the only way to communicate with people was in the blogosphere. I guess that’s where I sort of sharpened my ability to write.
Blogs as the original social software (before platforms became socmed)
But there wasn’t a whole bunch of work in the local area where I lived. I was looking at cheap ways for professional development and cheap ways for marketing to get things started. Then a friend suggested, “Why don’t you do a blog?” My current blog started in 2004. It was a way for me to reach out and to talk to people, and in those days of blogging, there were a lot of people who were helping others out, because there were so few of us, particularly in the educational technology area, where I was doing a lot of writing initially, and later in knowledge management.
[[Harold Jarche p]] describes how he came to blogging. He was in a place where there were no others to find. Blogging was finding the others. Early blogging scene was small and people helped eachother out. Did edutech first, then KM. Started in (early iirc) 2004.
Harold explained why he has chosen to focus on human intelligence rather than artificial intelligence, prioritizing depth, reflection, and community over scale and algorithms. This shift has led him increasingly toward smaller but trusted networks where knowledge can be shared more meaningfully.
smaller / trusted networks vs scale.
disorientation (an essential stage of learning)
n:: also important in transition / adaptation Vgl adopting new tool. Disorientation is not bc 'the tool is bad' it's bc you are learning #blogdit
we explored how writing has shaped Harold’s thinking and how blogging functions as a core sensemaking practice. H
writing and blogging as sensemaking, as learning
work is learning, and learning is the work.
[[Harold Jarche p]] motto
First Monday, open internet journal will cease publication in May 2026.
City council resolutions Munich wrt digital sovereignty and assessment of it. Documents downloaded to Zotero Raadsbesluiten Munchen in Zotero
Munich has an assessment tool for digital sovereignty developed at Munich Univ.
a package to install nextcloud quickly on a new server w back end tools. Debian 13 required.
Aimed at small groups / companies / associations / schools. Would work on a Hetzner cloud server. But works for larger set-ups too. By people in Schleswig-Holstein (where the public admin has switched to Nextcloud)
Making the case that Postgres will cover your db needs at low cost, and practice tested os tools, without complexity.
Ministry of Justice advised entities (ZBO's, UO's) to reconsider usage of Zivver, after FTM published investigative work into Kiteworks. IND as UO of the Ministry of Justice decided to leave Zivver, effective in 2026
Zivver is originally Dutch, but got bought by the US-ian Kiteworks https://www.zivver.com/nl/blog/persbericht-zivver-sluit-zich-aan-bij-kiteworks in #2025/06 Kiteworks is run by ex-intelligence cadre of the IDF. Zivver specialises in secure comms for various gov entities and hospitals etc.
IND stopt met Amerikaanse Zivver.
Civio sued the Spanish gov for transparency of BOSCO, an algorithm on deciding energy poverty vouchers, in 2018. Supreme court has sided with them in #2025/09 ordering the release of source code.
Digitale Tweelingen en Europa
is dit LDT EDIC
AI voor woningbouw
hopelijk niet genAI
10.00 uur Plenaire opening en keynote 11.00 uur Eerste sessieronde AI voor woningbouw Starten met indicatoren Werksessie Appstore Digitale Tweeling in de planketen 12.00 uur Netwerklunch & matchmaking 13.00 uur Tweede sessieronde Digitale Tweelingen: een vliegende start Benutten van toegepast onderzoek: XCARCITY Digitale Tweelingen en Europa Digitale Tweelingen in de planketen 14.00 uur Pauze 14.30 uur Derde sessieronde Xperience: hoe start je het DT-proces? Missiegedreven innoveren Project Digitale Tweeling Nederland 15.45 uur Plenaire afsluiting met resultaten testbed en speciale gastspreker 17.00 uur Netwerk-borrel
programma
communitydag NL DT #2026/03/03 in de Prodent Amersfoort
documentation of the Keeping API. Tested it in Postman already, see [[Keeping Ton API token]] ao. - [ ] vraag #teamfinance wat we uit die API zouden willen halen voor onze interne info.
De samenstelling van de PT is een vertegenwoordiging van verschillende rollen – ingevuld op niveau van (senior) strategisch adviseur, portfoliomanager of specialist: a. twee vertegenwoordigers van beleidsdepartementen; b. drie vertegenwoordigers van uitvoeringsorganisaties: één voor het burgerdomein, één voor het bedrijvendomein en één voor uitvoeringsorganisaties breed; c. twee vertegenwoordigers van medeoverheden; d. één vertegenwoordiger van de private partijen met een publieke taak; e. één vertegenwoordiger namens de opdrachtnemers; f. één vertegenwoordiger van BZK vanuit regie verantwoordelijkheid; g. deelnemer(s) op (vaste) uitnodiging. 4 Op voorstel van de voorzitter van de PT en in afstemming met de overige deelnemers kan op bovenstaande gevarieerd worden om invulling te geven aan eventuele specifieke wensen en behoeften op een GDI-domein. 5 Voor alle GDI-partners is het op verzoek mogelijk om agendalid te zijn en mee te lezen op agenda en stukken. 6
Deelname op sr strategisch adviseur / portfolio management niv. Geonovum zou agendalid kunnen zijn?
De PT heeft een onafhankelijk voorzitter. 2 De voorzitter wordt benoemd door de dgDOO.
PT vzs worden door dgDOO benoemd (vz OBDO)
PT heeft tot taak: a. de PGDI te adviseren over richten en prioriteren van functionaliteit op het specifieke GDI domein en toe te zien op de uitvoering daarvan; b. het leveren van inbreng voor de GDI-meerjarenvisie en het GDI-programmeringsplan; c. het vertalen van nationale en Europese ontwikkelingen naar consequenties voor de specifieke functionaliteit en de daaronder vallende afspraken, standaarden en voorzieningen; d. het ophalen van functionele wensen van afnemers, ook met oog op de aansluiting van organisatiespecifieke voorzieningen op de GDI; e. het betrekken van externe kennis en zorgen voor inbreng van private partijen; f. het signaleren en bespreken van eventuele knelpunten en dilemma’s en zorgdragen voor escalatie naar de Programmeringsraad GDI waar nodig.
PT taken - PGDI adviseren mbt prio's (een taak PGDI) op het domein van de PT, en de uitvoering te monitoren - inbreng GDI meerjarenvisie en programmeringsplan - vertalen nationale Europese ontwikkelingen mbt consequenties, incl functionaliteiten, afsprakenstelsels en standaarden, voorzieningen. - [ ] achterhalen of #geonovumtb aanhaakt bij PT gegevensuitwisseling, en infra. Of wie dan wel
wordt op elk van de vier GDI-domeinen – toegang, interactie, gegevensuitwisseling en infrastructuur – een programmeringstafel (hierna: PT) ingesteld.
op vier GDI onderwerpen worden programmeringstafels ingesteld. - toegang - interactie - gegevensuitwisseling - infrastructuur
Het secretariaat van de architectuurraad wordt verzorgd door het bureau MIDO.
MIDO doet ook secr architectuurraad net als PGDI
De samenstelling van de architectuurraad bestaat uit een vertegenwoordiging van verschillende rollen ingevuld op managementniveau (Chief Technology Officer / Chief Information Officer): a. drie / vier vertegenwoordigers van uitvoeringsorganisaties; b. een / twee vertegenwoordigers van opdrachtnemers; c. een / twee vertegenwoordigers van medeoverheden; d. twee / drie vertegenwoordigers van beleidsdepartementen; e. één vertegenwoordiger van BZK als opdrachtgever; f. één vertegenwoordiger van Forum Standaardisatie. 5 De architectuurraad wordt ondersteund door architecten.
architectuurraad bestaat niet uit architecten, maar MT level mensen van ZBO/UOs, departementen, decentrale overheden, MinBZJ als opdrachtgever, Forum Standaardisatie. Telkens op CIO / CTO niveau. Architecten staan de leden bij.
De architectuurraad wordt voorgezeten door een onafhankelijke voorzitter. 2 De voorzitter wordt benoemd door de dgDOO.
onafh vz, aangesteld door dgDOO (vz OBDO)
architectuurraad heeft tot taak: a. zorg te dragen voor de totstandkoming van een samenhangende, toekomstvaste GDI die wendbaar, bruikbaar en efficiënt is door toepassing van moderne architectuurinzichten; b. het (doen) opstellen van de GDI-architectuur; c. voorstellen te doen voor de transitie vanuit architectuurperspectief; d. te adviseren over inrichtingskeuzes van de GDI en de daartoe behorende bouwstenen; e. kaders voor te stellen voor onderhoud, ontwikkeling en beheer van de GDI; f. architectuurproducten van projecten die bouwstenen van de GDI ontwikkelen te toetsen, om te kunnen beoordelen of aan de GDI-kaders wordt voldaan; g. te monitoren of de afspraken in het kader van de GDI-architectuur in de praktijk bij afnemers gerealiseerd worden; h. adequate overleggen in te richten.
Taken architectuurraad: - totstandkoming van GDI obv 'moderne' architectuurinzichten - het opstellen van de GDI architectuur - vu architectuur naar transitie te kijken - inrichtingskeuzes en bouwstenen adviseren - kaders mbt onderhoud/beheer voorstellen - architectuurproducten mbt gdi toetsen - afspraken monitoren mbt gdi architectuur of die in de praktijk worden gebracht door pgdi leden.
een architectuurraad ingesteld.
nationale architectuurraad
PGDI komt ten minste viermaal per jaar bijeen
kwartaallijks
secretariaat van de PGDI wordt verzorgd door het bureau MIDO
MIDO v MinBZK is secrtariaat PGDI
Vervanging bij afwezigheid is alleen op horizontaal niveau (minimaal directeursniveau) mogelijk.
vervanging gaat horizontaal. Goed dat dit zo expliciet is, vgl wat er bij GI Beraad gebeurde over tijd.
De PGDI wordt voorgezeten door een voorzitter die het draagvlak heeft van de leden van de PGDI. 2 De staatssecretaris benoemt de voorzitter. 3 De voorzitter geeft op een objectieve wijze invulling aan het voorzitterschap vanuit een breed perspectief op de digitale overheid; 4 De PGDI bestaat voorts uit de volgende leden op minimaal directeursniveau: a. een vertegenwoordiger namens het Uitvoeringsinstituut Werknemersverzekeringen en/of de Sociale Verzekeringsbank; b. een vertegenwoordiger namens de Vereniging van Nederlandse Gemeenten; c. een vertegenwoordiger namens het Ministerie van Volksgezondheid, Welzijn en Sport; d. een vertegenwoordiger namens de Unie van Waterschappen; e. een vertegenwoordiger namens het Interprovinciaal Overleg; f. een vertegenwoordiger namens de Dienst Uitvoering Onderwijs; g. een vertegenwoordiger namens de Pensioenfondsen; h. een vertegenwoordiger namens de Belastingdienst (ook voor Douane en Toeslagen); i. de coördinerend opdrachtgever GDI; j. een vertegenwoordiger namens de Kamer van Koophandel; k. een vertegenwoordiger namens de Manifestgroep; l. een vertegenwoordiger namens Logius (ook voor KOOP); m. een vertegenwoordiger namens Rijksdienst voor Identiteitsgegevens; n. een vertegenwoordiger namens Rijksdienst voor Ondernemend Nederland; o. alsmede, afhankelijk van het onderwerp, de betrokken (kleine) uitvoeringsorganisatie(s).
leden zijn de 'afnemers' GDI, op dir niveau. vz is door stas benoemd en niet qq. - [ ] achterhaal mensen in PGDI. #geonovumtb
e PGDI heeft tot taak: a. het OBDO te adviseren over de gewenste en noodzakelijke meerjarige (door-)ontwikkeling op de GDI; b. het opstellen van het GDI-programmeringsplan met daarin het totaaloverzicht van beheer en exploitatie, projecten en programma’s met bijbehorende begroting van de GDI voor het volgende jaar (t+1); c. het bijeenbrengen van de ontwikkelingen en behoeften op de verschillende GDI-domeinen; d. het richten en prioriteren van de benodigde inzet op de GDI gelet op de (technische) samenhang en de onderlinge afhankelijkheden; e. de opdrachtgever te adviseren over opdrachten in het kader van de GDI; f. zo nodig adequate vooroverleggen in te richten.
Taken PGDI - adviseert OBDO mbt GDI - stelt programmeringsplan GDI op - inventariseert behoeften en ontwikkelingen - stelt prio's voor GDI - adviseert MinBZK over uit te zetten opdrachten mbt GDI - richt overleggen in.
Dit besluit wordt binnen 2 jaar na inwerkingtreding geëvalueerd waarbij ook op de werking van de nieuwe structuur wordt ingegaan. Vervolgens vindt evaluatie plaats wanneer daar aanleiding voor is.
Dat zou in 2024 zijn geweest. Maar is kennelijk nog van kracht zonder wijziging. En zou dan nu ongeveer opnieuw moeten worden geevalueerd, voorjaar 2026. #openvraag komt met het nieuwe kabinet dit besluit onder druk?
Instellingsbesluit Sturing digitale overheid 2022. - OBDO - PGDI - architectuurraad - programmeringstafels (PT)
OBDO komt ten minste viermaal per jaar bijeen.
kwartaalijks is er een OBDO. zijn verslagen publiek eigenlijl #openvraag
Programmeringsraad Generieke Digitale Infrastructuur (hierna: PGDI).
Er is een PGDI.
Het OBDO wordt voorgezeten door de directeur-generaal Digitalisering en Overheidsorganisaties (hierna: dgDOO) en bestaat uit de volgende leden: a. een lid van de directie van de Vereniging van Nederlandse Gemeenten; b. een vertegenwoordiger namens het Interprovinciaal Overleg; c. de algemeen directeur van de Unie van Waterschappen; d. een vertegenwoordiger van het Ministerie van Economische Zaken en Klimaat; e. een vertegenwoordiger van het Ministerie van Justitie en Veiligheid; f. een vertegenwoordiger van het Ministerie van Sociale Zaken en Werkgelegenheid; g. een vertegenwoordiger van het Ministerie van Volksgezondheid, Welzijn en Sport; h. een vertegenwoordiger van het Ministerie van Onderwijs, Cultuur en Wetenschap; i. een vertegenwoordiger van het Ministerie van Infrastructuur en Waterstaat; j. een vertegenwoordiger van het Ministerie van Financiën; k. de directeur Digitale Samenleving; l. de CIO-Rijk; m. de voorzitter Manifestgroep; n. de voorzitter van de Programmeringsraad GDI.
De samenstelling van het OBDO. vz is DG Digitalisering Overheidsorganisaties (dgDOO). Vanuit BZK zitten ook de dir digitale samenleving, en CIO rijk er in
vz Manifestgroep vertegenwoordigd de ZBO/UOs
Het OBDO bevordert de interoperabiliteit en stelt – op advies van het Forum Standaardisatie – de samenstelling van de lijst met open standaarden vast, waarbij voor overheidsorganisaties het ‘pas-toe-of-leg-uit’ regime geldt. Het OBDO stimuleert en volgt het gebruik van de adoptie van de standaarden op die lijst.
OBDO pusht op interoperability, stelt geadviseerd door Forum Standaardisatie de pas-toe leg-uit lijst vast
De Ministeries van Algemene Zaken (op het niveau van raadsadviseur), Landbouw, Natuur en Voedselkwaliteit, Buitenlandse Zaken en Defensie, zijn agendalid en kunnen, op niveau van secretaris-generaal of directeur-generaal, op eigen verzoek aan het overleg van het OBDO deelnemen. Andere organisaties kunnen op uitnodiging deelnemen.
Min AZ, LNV, BuZa en MinDef zijn agenda-id
De volgende producten worden ter advisering aan het OBDO aangeboden: – voorstellen met betrekking tot een overheidsbrede visie op digitale overheidsdienstverlening aan burgers, bedrijven en instellingen; – voorstellen voor de digitale transitie van het openbaar bestuur in verbinding met sectorale doelen en ambities voor de digitale overheid; – uitkomsten van in- en of uitvoeringstoetsen, als onderdeel van beleidsvoorstellen; – de GDI-meerjarenvisie; – een voorstel voor allocatie van het centrale budget in de vorm van het jaarlijkse GDI-programmeringsplan; – adviezen aan de staatssecretaris met betrekking tot de financiering van de GDI.
De OBDO wordt telkens aangeboden: - voorstellen overheidsbrede digi visies - voorstellen digitale transformatie openbaar bestuur - uitkomsten invoerings /uitvoeringstoetsen - de GDI meerjarenvisie - jaarlijkse GDI plan en budget - adviezen die de Stas (v BZK) krijgt tav financiering GDI
Het OBDO heeft tot taak de staatssecretaris te adviseren op de volgende terreinen: a. het beleid voor de digitale overheid, waarbij met name wordt geborgd dat deze gemakkelijk, betrouwbaar, veilig en begrijpelijk en inclusief is voor burgers en bedrijven evenals maatschappelijke instellingen, met oog voor de menselijke maat; b. de digitale transitie als integraal onderdeel van beleidsontwikkeling, met als doel maatschappelijke resultaten te realiseren en de daartoe noodzakelijke modernisering van de overheid te stimuleren; c. de toe te passen afspraken en afsprakenstelsels; d. de ontwikkeling van de generieke digitale infrastructuur en daarmee over de scope van de GDI; e. de aansluiting van sectorale digitaliseringsambities op generiek beleid. 2
De adviezen v OBDO betreffen: - beleid digitale overheid - digitale transitie als onderdeel beleidsontwikkeling - afspraken en -stelsels die toe te passen zijn - ontwikkeling van generieke digitale infrastructuur (GDI) - de verbinding tussen generiek beleid en sectorale digitaliseringsambities
LinkedIn API november 2025 version, mentioned as element in complying with DMA portability requirements. This is meant for companies not members then? But it says members. And it was offered to me when trying to download my personal account.
It seems LinkedIn has put this in place of a regular download, actually making it harder to get at your data.
Six degrees of Jeffrey Epstein
Minister van Klimaat en Groene Groei D66Staatssecretaris van Klimaat en Groene GroeiCDA
Onder EZK Klimaat eigen minister en stas plus 'groene groei'...
Staatssecretaris van Economische Zaken – Digitale Economie en Soevereiniteit
De digi-stas onder EZ, en zowel 'digitale economie' als 'digitale soevereiniteit' Wat betekent dit voor digi-programma's onder BZK nu? Logius, CIO Rijk, directie architectuur, FDS/IBDS, NDS, etc.
Slagvaardige Overheid
VVD stas. Wat valt onder 'slagvaardig', deregulering?
Binnenlandse Zaken en Koninkrijksrelaties Minister van Binnenlandse Zaken en KoninkrijksrelatiesCDAStaatssecretaris van Binnenlandse Zaken – Koninkrijksrelaties en Slagvaardige OverheidVVD Minister van Volkshuisvesting en Ruimtelijke Ordening D66
Geen stas digitalisering onder BZK, data / basisregs / fds / nds valt wel onder BZK.
Het nieuwe kabinet qua posities
personalised postage stamps
Germany fines Amazon 59 million Euro and prohibits their price control mechanisms. (both the US Amazon and Irish entity) Meaning they are not allowed to force pricing by traders on their platforms (this is imo the way they ensure that Amazon is always the lowest price online) This is a German effort, but coord'd with EC wrt DMA, though the fine seems based on German laws.
Part I — Multipolarity and the Expansion of Opportunity 1.1 — Multipolarity Is the Condition We Are Operating In 1.2 — Why Merit Stopped Working 1.3 — Opportunity Expands in Multipolar Systems 1.4 — AI Accelerates Opportunity by Exposing Gaps 1.5 — From Scarcity Thinking to Opportunity Literacy Part II — How Work Actually Functions Now 2.1 — Work Is Interpretive 2.2 — How Decisions Are Made and Why Responsibility Slides Downward 2.3 — How Acceleration Changes Mental Load 2.4 — Productivity Models Under Multipolar Conditions Part III — Being Useful, Making Choices, and Gaining Influence Over Time Author’s Note to Founders 3.1 — Usefulness Is the New Currency 3.2 — From Career Ladders to Role Creation 3.3 — AI as a Legitimacy Wedge 3.4 — Agency Without Heroics Part IV — Where Abstraction Collapses 4.1 — Manufacturing Is a Truth Serum 4.2 — When Decisions Become Physical 4.3 — UGME: Geometry Encodes Responsibility 4.4 — Manufacturability as Leadership Part V — Clarity, Protection, and Not Wasting This Moment 5.1 — Clarity Is an Active Discipline 5.2 — Protection Is a Condition for Contribution 5.3 — Agency Under Constraint 5.4 — Never Waste This Crisis
ToC of [[Sovereign Minds by Nadia Alter]] interesting terms, like 'opportunity literacy'
Sovereign Minds Leadership and Work in the AI Era Structure, Agency, and Judgment Under Acceleration
[[Sovereign Minds by Nadia Alter]] looks very worthwhile to get. mentions agency in the subtitle.
(via [[Alberto Cottica p]] on masto)
on responsibility, decision making and authority in organisations getting amorph, even before AI. Vgl [[Corporations as Slow AI 20180201210258]]
via [[Alberto Cottica p]] on Masto
Page that lists team and funding sources (ao a 3 yr grant by Dutch Adessium foundation)
Corporate Europe Observatory: shedding light on corporate lobbying in the EU.
Keep watch on a number of specific domains: - agri - energy / climate - lobbying regulations - trade / investment - economy / finance - other topics Digi is missing in this list, perhaps under other?
[[Tracy Durnell]] explains how she adds sidenotes in WordPress classic with just a bit of css, and using aside in html.
(v. [[Peter Rukavina p]]'s fav stream)
the CIA World Fact Book, a early web staple is no more.
Historian Mar Hicks documents this in Programmed Inequality: when work is coded as “women’s work,” it becomes a target for automation.
2017 [[Programmed Inequality by Mar Hicks]] How Britain Discarded Women Technologists and Lost Its Edge in Computing, US historian https://marhicks.com/ https://en.wikipedia.org/wiki/Mar_Hicks https://www.linkedin.com/in/mar-hicks-3487a539/
The pipeline runs from women as the original “computers” in the 1940s, through the masculinization of computing that pushed women into typing pools and administrative support, through the automation of those roles, to AI assistants today automating what remains: scheduling, reminding, organizing, emotional management.
There is a line from computers in the original sense, to typing pools, admin support, to automation to AI.
he connection between AI assistants and human secretaries isn’t metaphorical—it’s genealogical. In a 2020 academic paper titled “Alexa, Tell Me About Your Mother,” researchers Jessa Lingel and Kate Crawford trace the lineage explicitly. Alexa has a mother. Her name was Secretary.
"Alex has a mother. Her name was Secretary"
“Obedient and obliging machines that pretend to be women are entering our homes, cars and offices,” warned UNESCO’s Director for Gender Equality, Saniye Gülser Corat, in the agency’s landmark 2019 report.
Unesco report, Saniye Gülser Corat (dir for gender equality).
This gap between feeling and explanation isn’t mysterious. Neuroscientist Antonio Damasio spent decades documenting what he calls “somatic markers”—physiological responses that guide decision-making before conscious reasoning engages. These signals aren’t primitive or irrational. They’re compressed expertise—the accumulated wisdom of lived experience, speaking faster than language.
Somatic markers, Antonio Damasio physiologically expressed compressed lived experience.
Spanish court decision that algorithms used to decide about citizens by government entities, need to be transparant. Presented as landmark ruling with EU consequences, but I'm not sure if this isn't already true for public sector decision making under the AI Act. In NL the national algorithm register and tools like IAMA are aimed at this too, and have been in place for some time, even if not fully implemented yet.
[[Martijn Aslander p]] ontdekt de cronjob
UK Information Commissioner announces formal investigation of Grok
[[Peter Rukavina p]] on cognitive debt, vs cognitive investment aka learning wrt mundane infrequent multistep tasks. Good nudge to maintain little checklists for infrequent things better.
WhatsApp has its own backup feature (actually, it has more than one way to do it.) WhatsApp supports end-to-end encrypted backups that can be protected with a password, a 64-digit key, and (more recently) passkeys. WhatsApp’s public docs are here and WhatsApp’s engineering writeup of the key-vault design is here. Conceptually, this is an interesting compromise: it reduces what cloud providers can read, but it introduces new key-management and recovery assumptions (and, depending on configuration, new places to attack). Importantly, even if you think backups are a mess — and they often are — this is still a far cry from the effortless, universal access alleged in this lawsuit.
WhatsApp has its own backup feature, w additional key pairs etc. But this is not what is being claimed.
If you use native device backup on iOS or Android devices (for example, iCloud device backup or the standard Android/Google backup), your WhatsApp message database may be included in a device backup sent to Apple or Google.
backed-up decrypted message can be stored elsewhere when you do backups of your phone, e.g. Google or Apple depending on your device
Several online commenters have pointed out that there are loopholes in WhatsApp’s end-to-end encryption guarantees. These include certain types of data that are explicitly shared with WhatsApp, such as business communications (when you WhatsApp chat with a company, for example.) In fairness, both WhatsApp and the lawsuit are very clear about these exceptions. These exceptions are real and important. WhatsApp’s encryption protects the content of your messages, it does not necessarily protect information about who you’re talking to, when messages were sent, and how your social graph is structured. WhatsApp’s own privacy materials talk about how personal message content is protected while other categories of data exist.
The lawsuit is not about metadata, or WhatsApp use within a company which is not E2EE apparently (making it very unsuited for work situations I'd say)
The most important thing to keep in mind here is that Meta’s encryption happens on the client application, the one you run on your phone. If the claims in this lawsuit are true, then Meta would have to alter the WhatsApp application so that plaintext (unencrypted) data would be uploaded from your app’s message database to some infrastructure at Meta, or else the keys would. And this should not be some rare, occasional glitch. The allegations in the lawsuit state that this applied to nearly all users, and for every message ever sent by those users since they signed up. Those constraints would tend to make this a very detectable problem. Even if WhatsApp’s app source code is not public, many historical versions of the compiled app are available for download. You can pull one down right now and decompile it using various tools, to see if your data or keys are being exfiltrated. I freely acknowledge that this is a big project that requires specialized expertise — you will not finish it by yourself in a weekend (as commenters on HN have politely pointed out to me.) Still, reverse-engineering WhatsApp’s client code is entirely possible and various parts of the app have indeed been reversed several times by various security researchers. The answer really is knowable, and if there is a crime, then the evidence is almost certainly* right there in the code that we’re all running on our phones.
If the claim is correct, one could reverse engineer the app to see if true. Not a low hurdle but possible. 'the answer is knowable'
In the case of WhatsApp, the application software is written by a team inside of Meta. This wouldn’t necessarily be a bad thing if the code was open source, and outside experts could review the implementation. Unfortunately WhatsApp is closed-source, which means that you cannot easily download the source code to see if encryption performed correctly, or performed at all. Nor can you compile your own copy of the WhatsApp app and compare it to the version you download from the Play or App Store. (This is not a crazy thing to hope for: you actually can do those things with open-source apps like Signal.)
WhatsApp being closed source cannot be proven to work as advertised by outsiders. Unlike Signal
Today WhatsApp describes itself as serving on the order of three billion users worldwide, and end-to-end encryption is on by default for personal messaging. They haven’t once been ambiguous about what they claim to offer. That means that if the allegations in the lawsuit proved to be true, this would be one of the largest corporate coverups since Dupont.
Publicly WhatsApp has always maintained they do E2EE, the lawsuit says otherwise, that would be a major scandal. But also makes the claim hard to swallow
The state of encryption on major messaging apps in early 2026. Notice that three of these platforms are operated by Meta.
this is a sobering image. Signal at 70 million monthly active users. Apple imessage 1,3 billion Whatsapp 3 billion Instagram 2 billion FB Messenger 1 billion Telegram 1 billion Snapchat 900 million Discord 200million WeChat 1.3 billion Dingtalk 191million QQ 553 million no mention of Threema too tiny I suppose.
Not only does WhatsApp’s encryption prevent Meta from mining your chat content for advertising or AI training
again, not true for the metadata though
should never be able to read the content of your messages.
no mention here of the type of metadata WhatsApp holds: Signal only if account exists, and when last used. WhatsApp has contact lists and the date / time of every message between sender/receivers etc. That in itself is an issue imo.
Beginning in 2014 (around the time they were acquired by Facebook), the app began rolling out end-to-end (E2E) encryption based on the Signal protocol.
WhatsApp started rolling out E2EE around the time they were acquired by Meta. They use the Signal protocol
The downside of vast scale is that apps like this can also collect data at similarly large scale. Every time you send a message through an app like WhatsApp, you’re sending that data first to a server run by WhatsApp’s parent company, Meta.
The scale is the reason the collected data is an issue.
In terms of scale, modern messaging apps are unbelievably huge. At the start of the period in the lawsuit, WhatsApp already had more than one billion monthly active users. Today that number sits closer to three billion. This is almost half the planet. In many countries, WhatsApp is more popular than phone calls.
Scale of WhatsApp is close to 3 billion people.
USA is leaving OGP under the Trump regime. The USA was a cofounder in 2011 and together w Brazil held a launching conference w civil society in July that year. I missed the OGP launch in Washington bc I thought the formal invitation by then Secretary of State Hillary Clinton I received was spam. [[Hoe ik Hillary Clinton spamfilterde en de oprichting van OGP miste]]
Headline somewhat misleading, mostly covers that CapGemini is selling their US branch that works for ICE where bonuses directly correlate with immigrants located and detained. They're not ditching doing business with ICE, they're selling it to the next bidder who by def has no qualms doing business with ICE. And it remains to be seen how fast this sale will happen if at all. A few other examples that are smaller, like office leases. IBM-style behaviour still the norm imo. Vgl [[Comment le groupe français Capgemini aide la police fédérale américaine ICE à localiser les migrants]]
SpaceX 'bought' xAI for 250 billion USD. xAI previously 'bought' Twitter in March 2025. SpaceX is expected to go public this or next year. Iow, Musk is combining everything to a) inflate overall value b) externalise all risks to investors c) get his biggest payday yet
Grok will be live tweeting nudified astro-pics from every launch and starlink satellite burning up in the atmosphere
TenneT Holding is selling part of their stock in TenneT Germany, to the German government, over 25%. at 3.3 billion Euro.
TenneT Holding is a GOE 100% owned by the Dutch government.
De Duitse investering zorgt ervoor dat Tennet niet langer zelf het merendeel van de aandelen heeft van de Duitse activiteiten. De Nederlandse pensioenuitvoerder APG, een Noors staatsoliefonds en een Singaporees staatsinvesteringsfonds bezitten samen ongeveer 46 procent."TenneT Holding zal minimaal 28,9 procent van de aandelen in TenneT Duitsland behouden", schrijft het bedrijf. "Hierdoor beschikt TenneT Holding over volledige betrokkenheid bij belangrijke besluiten."
The shares of Tennet Germany are: Tennet Holding 28,9% German state: 25% APG (Dutch pensionfund), Norwegian state oil fund, Singapore state investment fund (huh?), together the remaining 46%. This seems to leave TenneT a minority but still largest shareholder.
French justice dept doing a search of X offices in Paris, wrt purposefully eroding democratic processes w ao amplifying hatespeech.
Mirlo received NLNet funding to add AP capabilities, and make Mirlo available as federated instances
Mirlo is a platform a la Bandcamp
Het genoemde AI Deltaplan is overigens geen deltaplan. Het is eigenlijk niet eens een plan. Het is meer een opsomming van wensen van mensen die zouden willen dat Nederland gewoon net zo zou worden als Silicon Valley. “Maar dan wel veilig”, zeggen ze. Ik schreef daar eerder over, als je in Europa wil innoveren moet je dat op een Europese manier doen. En niet kleine stukjes uit Amerika kopiëren en dan hopen dat dat hier gaat werken. Want dat doet het niet.
word.
uitbreiding van de WBSO voor ontwikkeling van AI”
WBSO, de research belastingaftrek, krijgt AI poot
“We stimuleren de bouw van een AI-fabriek in Noord-Nederland en van Europees- autonome datacentra. W
Dit is geen landelijk ding, maar een EU subsidie voor een regio, no? https://www.nijbegun.nl/projecten/ai-fabriek/
Stimuleren autonome Europese datacentra. C'est quoi? Heb zelf wel beeld, maar zij dan?
Probeer vooral eens wat er nog werkt als het door het Amerikaanse Equinix onderhouden Koppelnet Publieke Sector gehacked is. Of Solvinity!
“We richten een Nederlandse Digitale Dienst op: compact, deskundig en met doorzettingsmacht. Deze dienst ondersteunt de digitalisering Rijksbreed, stelt kwaliteitsstandaarden op en borgt goede ontwerpkeuzes. We verminderen afhankelijkheid van externe IT-leveranciers door meer IT-talent in dienst van het Rijk te nemen.”
Te vormen 'nederlandse digitale dienst', soort Rijks ICT Gilde plus? Wederom 'rijksbreed'. Gericht op minder afhankelijkheden van externe IT leveranciers, meer IT kennis binnen het Rijk (wederom).
De “call-in bevoegdheid” komt voort uit een initiatiefwetsvoorstel van GroenLinks-PvdA kamerlid Bushoff. De gedachte daarbij is dat bedrijfsovernames, ook als er nog geen gevaar is voor een echt monopolie, toch ongewenst kunnen zijn. Het voorstel noemt daarbij Co-Med, waar veel huisartsenpraktijken onder gingen vallen, met vervelende gevolgen. Deze bevoegdheid lijkt ook best gevolgen te kunnen hebben voor overnames van Solvinity-achtige bedrijven met lokaal zeer belangrijke posities.
call-in bevoegdheid is een NL instrument, om in te kunnen grijpen bij bedrijfsovernames met grote externaliteiten voor de publieke zaak.
Met de new competition tool (NCT), ook omschreven in het beroemde Draghi report (pagina 302), is het mogelijk bedrijven bindende aanwijzingen te geven als er effectief geen concurrentie mogelijk is. Als voorbeeld, als het praktisch onmogelijk is om te concurreren met Google of Microsoft voor email en agendabeheer kan je hiermee een bindende aanwijzing geven dat deze bedrijven het mogelijk moeten maken om als nieuwe speler mee te doen in hun ecosysteem. Zo’n verplichting is er nu soms ook al, maar alleen onder hele specifieke omstandigheden. Met de NCT kan het ook onder de algemene voorwaarde dat er van een niet-concurrerende markt sprake is.
ACM krijgt beschikking over 'new competition tool' NCT a la Draghi , bindende aanwijzing nieuwe spelers toe te laten.
rijksdienst nu werkelijk gemarineerd is in Amerikaanse technologie & dito arbeidsverleden. Ook consultants, big four, integrators en andere partijen waar de overheid op leunt komen uit die wereld. Als we weg willen komen bij “big tech” is het een hele uitdaging om dat te doen vanuit zo’n achtergrond/denkwereld. Ik zeg hier nadrukkelijk niet dat de ambtenaren voor Microsoft werken overigens, noch dat ze corrupt zijn! Desondanks vind ik de monocultuur in achtergrond en adviseurs een probleem. “Microsoft-tenzij” is op veel plekken uitgeschreven en officieel beleid.
Goed punt: wie als deskundig geldt momenteel heeft een achtergrond in datgene waarvan men weg wil bewegen. Dus behoefte aan mensen die kennis hebben op het andere.
We uniformeren de bedrijfsvoering binnen het Rijk onder leiding van het ministerie van BZK: van ICT en inkoop tot HR, onder andere door verplichte standaarden en gezamenlijke voorzieningen
BZK als lead.
“Digitale inkoop en aanbestedingen worden gestandaardiseerd en gecentraliseerd, gestuurd op security-by-design, zero-trust, soevereiniteit, open source en ketenveiligheid. De overheid benut haar marktmacht om veilige standaarden af te dwingen en stelt rijksbrede minimumeisen op voor security. Om voor financiering in aanmerking te komen moeten IT-projecten van de overheid (> €5 mln.) aan centrale IT-standaarden worden getoetst”
is dit ook voor decentrale overheden? 'rijksbreed' is niet overheidsbreed. Wel goed dat uitsluitende voorwaarden lijken te gaan worden gesteld tav standaarden, security by design, zero-trust, soevereiniteit, open source en ketenveiligheid. Maar met 5 miljoen als ondergrens. Daarmee kijkt de VS nog vrolijk mee in alle documenten en zaaksystemen lijkt me. En grotere projecten zullen worden opgeknipt in fasen van elk net geen 5 miljoen. Maar is een begin.
“Digitale autonomie moet het uitgangspunt zijn voor de overheid. We kiezen voor een Europese digitale infrastructuur, bouwen strategische afhankelijkheden in cloud, data en cruciale systemen doelgericht af”
Aankondiging van een bigtech transitie
Nederland moet de aanjager zijn van een geopolitieke en sterke Europese Unie die daadkrachtig optreedt. Geopolitiek besef moet doorklinken in alle beleidsterreinen.
statement mbt algemene Europese soevereiniteit en geopolitieke machtsposities.
Ook zegt het akkoord niet veel over geld, en voor sommige dingen gaat een boel geld nodig zijn, want anders blijft het bij warme woorden hoe belangrijk we dingen vinden. Er is een budgettaire tabel gepubliceerd maar hier lijkt op het eerste gezicht niets te staan over digitalisering. Wel over BTW op sierteelt.
Geen financiele paragraaf over alle digi voornemens
Deze week bleek dat deze NDS een door bigtech gedomineerde lobbyclub had aangesteld om te bepalen wie de overheidsbrede cloud mocht gaan bouwen (archieflink).
doelt op NLDigital denk ik, die in de kamer ook nogal opviel zeg maar.
Er komt een “digitaal bewindspersoon”, maar onbekend waar en hoe Geen woord over geld, is ook niet opgenomen in de “budgettaire tabel”!
digitaal bewindspersoon wederom, maar zonder rijksbudget. Kansje voor '25% van de besparing in de eerste 2 jaar is het budget'?
Twee nieuwe mogelijkheden voor ACM om de markt voor Europese clouddiensten wakker te schudden/te beschermen
welke, en zijn die anders dan wat voortkomt uit EU wetgeving?
Men wil echt weg uit de Amerikaanse cloud, en digitale autonomie moet het uitgangspunt zijn
Check tekst of het alleen autonomie noemt, en hoe gedef'd dan. Want soevereiniteit is veel belangrijker voor een overheid.
Geen enkele verwijzing naar de eerdere Nederlandse Digitaliseringsstrategie (NDS), of het daaruit voortvloeiende beloofde Overheidsbrede Cloudbeleid - opmerkelijk
Dit is heel vreemd ja. NDS is er nog maar net. Vgl [[Nathan Ducastel p]] NDS gaat ook over het digihuishouden van de overheid (maar incl decentraal: is dat het verschil?)
Nederlandse Digitale Dienst binnen de overheid met “doorzettingsmacht”
'Nederlandse Digitale Dienst' heeft ook of is de doorzettingsmacht?
Een revolutie in de ICT in de Rijksoverheid: centrale aansturing, centrale inkoop, verplichte standaarden, geen geld als men zich niet aan die standaarden houdt. Dit vanuit Binnenlandse Zaken, die daar een zware dobber aan gaan krijgen
BZK als departement met doorzettingsmacht voor inkoop. Dat laat vervolgens wel weer decentrale overheden hun gang gaan: all provs en gemeenten hangen aan MS bijv.
https://berthub.eu/articles/posts/digitaal-zoet-zuur-coalitie-akkoord/
Eerste blik [[Bert Hubert p]] op regeerakkoord mbt digitalisering.
[[Martijn Aslander p]] is enthousiast over APIs aanspreken en verbinden met lokale workflows. Terecht. Ik realiseer me dat ik wel APIs gebruik om dingen te versturen (zoals schrijven in mijn feedreader), maar nog weinig om data naar me toe te halen binnen persoonlijke tools. APIs aanspreken voor banktransacties bijv, is een goede suggestie. Idem voor plaatsnamen of andere dingen in notities. Zowel in personal tools als in een template voor bijv Obsidian plugin.
Ollama is automatically detected when running locally at http://127.0.0.1:11434/v1
openclaw can detect presence of ollama if it is visible at this specific localhost address. basically if you have ollama running it will be detected. Meaning I could run openclaw fully locally.
Ollama Ollama is a local LLM runtime that provides an OpenAI-compatible API:
one can set openclaw to work through ollama.
openclaw does not need anthropic, can use models locally
ollama can integrate with OpenClaw, meaning it would run through local models
Setting context length Setting a larger context length will increase the amount of memory required to run a model. Ensure you have enough VRAM available to increase the context length.
This setting is in ollama desktop interface. Does it set it for the terminal too? Or are these two separate instances?
Cloud models are set to their maximum context length by default.
If so, why would I need to set anything ever?
Context length is the maximum number of tokens that the model has access to in memory. The default context length in Ollama is 4096 tokens. Tasks which require large context like web search, agents, and coding tools should be set to at least 64000 tokens.
Default ollama context length is 4k. Recommended minimum for websearch, agents and coding tools (like Claude Code or Open code) is 64k. I've seen 128k recommendations for Claude Code
Clawdbot Realistic Costs:Software: Free (MIT licensed, forever)Hardware: VPS $4-5/month, or Raspberry Pi ~$50-100 upfront, or old laptop free, or Mac Mini ~$600AI Model: Claude Pro $20/month (casual) to Claude Max $200/month (heavy use like Viticci)Realistic minimum: ~$25/monthBut remember: that $300+ in 2 days user is real. Heavy agentic use burns through tokens fast.
Assuming cloud based models. Why? You could drop up to 200 month on a VPS and be really self sufficient, but probably wouldn't need a VPS that heavy?
I solved it with a two-tier system - short-term memory.md files and long-term topic indexes.
Author gives claudecode persistence by adding two types of logging context
Clawdbot is built for something completely different. Think of it as a personal executive assistant that lives inside your messaging apps. Just like you’d email instructions to a real assistant who then handles the work while you focus on other things, Clawdbot receives natural language requests through WhatsApp, Telegram, Slack, or Discord and then executes actual tasks on your computer.
Description of Clawdbot/ now named openclaw, as a PA
The $300/day cost you mention for heavy Clawdbot usage matches what I've seen in practice.
wut? racking up API costs to AI models like that is bizarre
Clawdbot uses AI APIs on every interaction and every background task. Costs accumulate.
need to understand better where it is getting its models from. Can be local too? Then costs are your own compute only. Why is everyone always assuming cloud models against fee?
Clawdbot requires technical setup. You need to deploy it on a server or local machine, configure messaging platform integration, manage API keys, and set permission boundaries thoughtfully. It’s not a consumer app. You can’t open an app store, tap install, and have it ready. That’s a real friction point that’s worth acknowledging.
run on a dedicted machine?
Use Clawdbot when your primary bottleneck is the accumulation of small tasks across your digital life.
clawdbot for 'small tasks across your digital life', vgl [[Aazai 2025 aantekeningen]] wrt such tasks.
Clawdbot’s power to access messaging platforms means anyone with a security compromise at any layer could potentially impersonate you to the agent. A prompt injection through a web page it’s browsing, a malicious message in a group chat, or a crafted email could theoretically redirect it toward unintended actions. Proper sandboxing and permission boundaries mitigate this, but they require genuine technical discipline
clawdbot as additional attack surface
Clawdbot requires elevated system permissions to do what it does. It needs to read and write files, execute shell commands, access your terminal, connect to services on your behalf. Running an always-on agent with access to your credentials, your messaging platforms, and your file system creates security surface area that’s worth understanding.
security concerns wrt clawdbot bc it executes actions you'd normally do.
Every conversation you have with it, every preference you state, every decision you make gets stored in a markdown file that evolves over time. Future requests pull relevant context from that history automatically. You don’t have to remind it of things.
Full context maintained in a md file.
Clawdbot is built for async, long-running, continuous tasks that exist across your entire digital life. It’s the tool that monitors your inbox at midnight, processes information while you sleep, and sends you a briefing in the morning. It’s designed to be always on, always learning your preferences, always available through the messaging apps where you already live
clawdbot is persistent across time / tasks.
It maintains persistent memory across conversations, so it remembers your preferences, past decisions, and ongoing projects. It can monitor scheduled tasks, send you proactive notifications, and continuously work on long-running tasks even when you’re not actively messaging it.
clawdbot maintains context over time In a .md log of sorts?
Clawdbot connects to dozens of services by default: Gmail, Google Calendar, Todoist, GitHub, Spotify, even smart home devices. When it needs capabilities it doesn’t have built in, it can request them, and with proper guidance from you, it can expand those capabilities itself.
'by default' or can do it out of the box, if switched on?
More knowledge on d.AP’s website.
the whole point of the article no doubt.
Why This Matters: From Probabilistic Guesses to Deterministic AnswersGrounding your AI in a Knowledge Graph delivers three non-negotiable enterprise advantages:1. Accuracy: Answers are derived from an explicit model of your business, not from statistical correlations in text. You eliminate both factual and relational hallucinations.2. Explainability: Every answer comes with a query that shows exactly how it was derived, which entities were connected and which rules were applied. This turns the AI from a black box into a transparent tool.3. Architectural Stability: The semantic layer, the ontology, remains stable even as the underlying systems change. When you migrate your CRM, you simply update the mapping to the ontology. Your AI, analytics, and dashboards continue to work without interruption. This is agility where it counts.
again this is arriving at database queries you already had, but through a way more convoluted way. When 'reasoning' is deterministic then you don't need a probabilistic layer at all, no?
2. A Schema-RAG system using a Knowledge Graph operates differently:The AI first consults the ontology to understand the question’s components.It finds that Support Ticket is a class linked via a property referencesProduct to the Product class.It discovers that the Product class has a property called productType, and that ‘Connected Service’ is a specific instance of that type.Armed with this understanding of the relationships, it constructs a precise, formal query (SPARQL) to retrieve only the tickets that conform to this logic.
Congrats you just recreated a pre-existing tab in your existing support ticket system, by vibecoding a sparql query that was likely already in your system's manual even.
Consider a critical business question: “Show me all open support tickets for our ‘Connected’ services.”
another telling example. Is the number of 'open support tickets' 'critical' to your business?. Most companies would say no. The article is way too vague in specifying which organisations are the mental audience here. Or worse, thinking all companies are like it. Why would you need AI for this even? Any ticketing system will have a tab for this.
1. A Formal Ontology: This is the rulebook. Built using standards like RDF, OWL, and SHACL, the ontology is an intentional, explicit contract of meaning. It defines the classes of things that matter (Customer, Contract, Product), their properties (hasName, hasValue), and the relationships between them (coversProduct, assignedTo). This is where you declare, unambiguously, that a Debitor in SAP is, in fact, the same as a Client in Salesforce.
ontology here is an information management ontology, mapping variables and their properties across applications ('domains' used above is a confusing term then). The mention of SAP here is a tell, in the sense of what type of scale of enterprise we're talking about.
This grounds a probabilistic LLM in a deterministic model of your business, transforming a clever chatbot into a reliable reasoning engine for mission-critical decisions.
bit hyperbolic phrasing. deterministic layers in front of an llm makes sense. Also a form of prompt engineering in a way. Seem to see that regularly: better inputs to generate better outputs. At what point are you spending so much on inputs, you might as well make the output yourself that way? Is the assumption most companies can actually formulate their processes in terms of 'business logic'? What would an example look like?
The winning architecture uses Schema-RAG: retrieving meaning from a formal ontology before any data query is formed. This approach grounds the AI in verifiable business logic, ensuring accuracy and full explainability
schema-rag as improvement (formal ontology added up front before a query for rag is formulated).
Knowledge Graphs provide the semantic context, constraints and explicit relationships that LLMs lack. This enables true reasoning, like navigating a map of your business, instead of just text retrieval.
knowledge graphs represent semantic context and relationships / constraints. K-graphs are a 1980s thing, I know we added them into the reference architecture for systems of digital twins I cowrote. But have no understanding more recent than the 1990s. - [ ] spend #30mins collecting current state of the art on #knowledgegraphs #pkm
Standard Retrieval-Augmented Generation (RAG) over documents is a good first step, but it fails when faced with complex, cross-domain enterprise questions. It finds text that looks similar, which isn’t the same as finding facts that are related.
criticism of retrieval augmented generatio (RAG): fails in cross domain settings, finds similar text not relations between facts or meaning
Julius Hollmann
author selfdescribed as "CEO @ digetiers I Building d.AP - the semantic foundation for real enterprise AI. Passionate about turning fragmented data into contextual, usable knowledge." so has a business interest in the topic. What's 'real enterprise AI' and what is his def of enterprise (99% of all companies are SME)