Odd piece, where Ben Werdmüller says he is a bad note taker, and then describes what is actually ok note taking. Seems assumptions get in the way vgl https://www.zylstra.org/blog/2025/11/warning-tacit-assumptions-may-derail-pkm-conversations/
1,804 Matching Annotations
- Dec 2025
-
Tags
Annotators
URL
-
-
arxiv.org arxiv.org
-
The paper announcing Apertus.
Saved Apertus: Democratizing Open and Compliant LLMs for Global Language Environments in Zotero
-
-
huggingface.co huggingface.co
-
The Apertus models also expand multilingual coverage, training on 15T tokens from over 1800 languages, with ~40% of pretraining data allocated to non-English content. Released at 8B and 70B scales, Apertus approaches state-of-the-art results among fully open models on multilingual benchmarks, rivalling or surpassing open-weight counterparts
Apertus is trained on over 1800 languages (!?) with 40% non English content, meaning many of them can only have had 1/100 or 1/1000 of a procent (1/10k, 1/100k) 60/1799 is 0,033%
Tags
Annotators
URL
-
-
www.swiss-ai.org www.swiss-ai.org
-
Multimodal World Foundation Models for Physical AI
-
FlexLM: Efficient Targeted LLM Compression
llm compression, is reduction of computational and memory footprints (both in creation and usage it seems)
-
Project list of the Swiss AI initiative
Tags
Annotators
URL
-
-
academic.oup.com academic.oup.com
-
1963 paper by Edmund L Gettier, wrt 'Gettier problems' counter examples that Justified True Belief is not sufficient def of K.
-
-
en.wikipedia.org en.wikipedia.org
-
Gettier claims to have shown that the JTB account is inadequate because it does not account for all of the necessary and sufficient conditions for knowledge.
Edmund Gettier, Gettier problem, epistemology. 1963 paper.
holds that the K def of 'justified true belief' is inadequate. Encountered in Is Understanding A Species Of Knowledge? in Zotero which posits that understanding is a species of knowledge that 'can be Gettiered'
-
-
link.springer.com link.springer.com
-
Explanation as Orgasm* in Zotero
Mentioned in https://philpapers.org/archive/BOUAPT.pdf v [[A Phenomenal Theory of Grasping and Understanding 20240418132232]]
-
-
map.meshcore.dev map.meshcore.dev
-
map of meshcore nodes in NL
-
-
meereco.nl meereco.nlMeer ECO7
-
Binnen MeshCore is het niet alleen het apparaat, maar vooral de software die een belangrijke rol speelt. Bij het inrichten van een repeater dienen we dan ook bij de flasher te kiezen voor de rol repeater.
A repeater is only different in software from a terminal.
-
Meshcore repeater
-
Meshtastic heeft heartbeats om continu binnen het netwerk aan te geven wie er is. MeshCore doet dit niet,
meschcore has less overhead in data traffic, no heartbeat pings.
-
Meshtastic gebruikt bredere bandbreedtes (125 kHz of 250 kHz), waarbij MeshCore bewust lager zit (62.5 kHz). Hierdoor heeft MeshCore voor bebouwde omgevingen een aantal voordelen:
Meshcore uses half / a quarter of the Meshtastic bandwidth.
-
Eind 2025 is er een flinke verhuizing gaande van Meshtastic naar MeshCore. Dat heeft er o.a. mee te maken dat MeshCore een betere ontvangst lijkt te hebben en zijn de eerste signalen dus ook positief. Steeds meer eigenaren van repeaters zetten er dan ook MeshCore-firmware op en stappen daarmee over naar MeshCore.
Implies there's a migration wave from Mehstastic to Meshcore
-
Meshtastic bestaat sinds 2020 en heeft de afgelopen jaren wat bekendheid gegeven aan het opzetten van mesh-netwerken. In 2025 is MeshCore beschikbaar gekome
Meshcore is from this year, Meshtastic started in 2020.
-
Meshcore as alternative to Meshtastic. Seems less US centric.
-
-
huggingface.co huggingface.co
-
Apertus LLM on Huggingface
-
-
www.swiss-ai.org www.swiss-ai.org
-
The model is named Apertus – Latin for “open” – highlighting its distinctive feature: the entire development process, including its architecture, model weights, and training data and recipes, is openly accessible and fully documented.
Apertus committed to openness wrt all its aspects. Is it in the overview yet?
-
https://web.archive.org/web/20251207141211/https://www.swiss-ai.org/apertus
Apertus, model developed by ETH and others. 40% non English inputs from 1k languages.
-
-
-
2025 overview of press freedoms around the world. Top 15 all European countries, of which 12 EU MS
-
-
novelfeelings.com novelfeelings.com
-
book review/ summary of [[Alone With You in The Ether by Olivie Blake]] that I picked up in Le Mans summer 2024, and haven't read yet. From a psychologists perspective.
Tags
Annotators
URL
-
-
novelfeelings.com novelfeelings.com
-
However, I also believe that therapy should be about the client; there’s a line between self-disclosure and self-indulgent
when is it a conversation, when are you making it about yourself only?
-
Dr Yalom talks about this as well in the book and argues that self-disclosure can be useful to building connection, the thing that accounts for 80% of success in therapy.
Sessions also contain self-disclosure by Yalom, against therapeutic condition, but here positioned as way to connect. (and I think regular interaction between ppl.)
-
focusing on the ‘here and now’. Often, he picks up on the dynamic between him and the client and raises this with the client, linking it with the issue that brought them to him. Most of the time, this is effective.
Yalom makes the meeting between the original request and the current setting / dynamic in the conversational pair the thing to explore.
-
The book is a collection of stories of these one-off sessions, and they are fascinating to me as a psychologist. How effective can one hour of conversation be? It is true that Dr Yalom (or Irv, as he asks his clients to call him) does not promise to resolve his clients’ needs in one hour; in fact, every consultation is ended with him providing names of other clinicians and urging the client to continue with longer-term therapy.
Yalom during Covid engaged in 1 hour sessions. Not as self-contained therapy, but still self-contained one-off conversations.
-
[[Het uur van het hart by Irvin D. Yalom Benjamin Yalom]] review/summary .
Tags
Annotators
URL
-
-
en.wikipedia.org en.wikipedia.org
-
It was, however, the usage of Franz Brentano (and, as he later acknowledged, Ernst Mach[6]) that would prove definitive for Husserl.[14] From Brentano, Husserl took the conviction that philosophy must commit itself to description of what is "given in direct 'self-evidence'."[15]
Husserl wrt phenomenology based himself on Franz Brentano and Ernst Mach . p7 Moran, Dermot (2000). Introduction to Phenomenology. Routledge.
-
-
philpapers.org philpapers.org
-
A Phenomenal Theory of Grasping and Understanding David Bourget, via [[Stephen Downes]]
In Zotero A Phenomenal Theory of Grasping and Understanding in Zotero
Tags
Annotators
URL
-
-
calibre-ebook.com calibre-ebook.com
-
Support for jumping to and displaying paper edition page numbers [7.17] If an e-book contains embedded information about the pages from the paper edition of the book, the calibre E-book viewer can now jump to a page number via the Go to button in the viewer controls.
Calibre, if it can detect it, can use paper edition page numbers. It's #openvraag how I want to deal with location numbers in ebooks as ref.
-
[8.11.1] Supports hundreds of AI models via Providers such as Google, OpenRouter, GitHub and locally running models via Ollama.
Calibre supported Ollama since 8.11, for Ask AI tab in dictionary panel.
-
New features Allow asking AI questions about any book in your calibre library. Right click the "View" button and choose "Discuss selected book(s) with AI" AI: Allow asking AI what book to read next by right clicking on a book and using the "Similar books" menu AI: Add a new backend for "LM Studio" which allows running various AI models locally
AI features in Calibre. discuss book w C, book suggestions, and LMStudio back-end. I set up Calibre w the LM Studio back-end, so things remain loca.
A posting elsewhere suggested it woud also suggest better metadata through AI. But that article seems generated itself, so disregarded.
Tags
Annotators
URL
-
-
sobrief.com sobrief.com
-
https://web.archive.org/web/20251207114722/https://sobrief.com/books/hour-of-the-heart
[[Het uur van het hart by Irvin D. Yalom Benjamin Yalom]] gift from E, #2025/12
Tags
Annotators
URL
-
-
www.kobo.com www.kobo.com
-
[[Essentialism by Greg McKeown]] 2024 edition as Kobo epub w new foreword and a '21 day challenge'
-
-
-
Hogwarts Legacy Walkthrough guide
-
-
ericsandroni.com ericsandroni.com
-
Another summary of [[Essentialism by Greg McKeown]] the 2014 version
-
-
nielsbohrmann.com nielsbohrmann.com
-
https://web.archive.org/web/20251207094605/https://nielsbohrmann.com/essentialism-book-summary/
summary of [[Essentialism by Greg McKeown]], 2014 book. There's a 2024 2nd edition.
Tags
Annotators
URL
-
-
-
https://github.com/robalexdev/blog-quest
BlogQuest is a browser extension for FF, Chrome , andf a fork from StreetPass (that collects AP profiles), which collects RSS feeds. No alerts, just building the list
Tags
Annotators
URL
-
-
streetpass.social streetpass.social
-
https://streetpass.social browser extension that picks up the relme links to AP profiles while browsing. For FF, safari, chrome, not in Arc. No alerts, just quietly building a list.
Tags
Annotators
URL
-
-
modub.nl modub.nl
-
Het zijn markdown bestanden met een persoonlijkheid, frameworks, en output templates. Die heb ik niet zelf geschreven - ik heb Claude gevraagd om ze te maken. “Maak een Product Owner agent die goed is in prioriteren en impact/effort analyses kan doen.” Claude schrijft dan het volledige bestand, inclusief werkwijze en voorbeelden.Als ik vervolgens zeg “vraag dit aan Tessa”, laadt Claude dat bestand en wordt Tessa.
Seems like these agent .md files contain description of a role that is then included in a prompt.
-
-
modub.nl modub.nl
-
In mijn werkmap heb ik een verzameling “agents” - tekstbestanden die Claude vertellen hoe hij zich moet gedragen. Tessa is er één van. Als ik haar “laad”, denkt Claude vanuit het perspectief van een product owner.
Author has .md files that describe separate 'agents' she involves in her coding work, for each of the roles in a dev team. Would something like that work for K-work? #openvraag E.g. for project management roles, or for facets you're less fond of yourself?
-
-
modub.nl modub.nl
-
https://modub.nl/ blog by Monique Dubbelman , suggested by [[Frank Meeuwsen]]
-
-
github.com github.com
-
bc of Calibre adding AI, this is an AI-less fork, calibre minus a and i, thus clbre.
Tags
Annotators
URL
-
-
www.howtogeek.com www.howtogeek.com
-
Calibre has added AI 'support', mostly to suggest new stuff to read and an option to discuss a book. It has an LM Studio back-end, so I can tie it to my local models.
-
-
www.ernestchiang.com www.ernestchiang.com
-
Luhmann’s card structure:Front: Complete bibliographic information (author, title, publication year, etc.)Back: Page index = “Page x has this concept, page y has that argument”
States that Luhmann's bibliographic cards had his own indexing on the back. Vgl with my own immediate annotations I keep in a book note.
-
-
en.wikipedia.org en.wikipedia.org
-
individuals constantly project into the world around them the interpretive frames that allow them to make sense of it; we only shift frames (or realize that we have habitually applied a frame) when incongruity calls for a frame-shift.
Vgl brain continuously predicting/projecting, not observing out of efficiency. Predictive coding.
-
When we want to explain an event, our understanding is often based on our interpretation (frame). If someone rapidly closes and opens an eye, we react differently based on if we interpret this as a "physical frame" (they blinked) or a "social frame" (they winked). The blink may be due to a speck of dust (resulting in an involuntary and not particularly meaningful reaction). The wink may imply a voluntary and meaningful action (to convey humor to an accomplice, for example). Observers will read events seen as purely physical or within a frame of "nature" differently from those seen as occurring with social frames
Applying different frames yield diff interpretations. Humor often based on frame switch.
-
Framing is an integral part of conveying and processing data daily. Successful framing techniques can be used to reduce the ambiguity of intangible topics by contextualizing the information in such a way that recipients can connect to what they already know.
Vgl [[Social netwerk als filter 20060930194648]] , [[Chunking 20210312215715]
-
https://web.archive.org/web/20251206085443/https://en.wikipedia.org/wiki/Framing_(social_sciences)
Framing, social sciences
-
-
digitallyliterate.net digitallyliterate.net
-
Guardrail: Strict adherence to the Tagging Workflow. Everything defaults to #private. Only notes explicitly tagged #share pass through the publishing filter. Use clear folder structures to separate sensitive data that should never touch the publishing pipeline.
Have this for blog posts, separate folder and explicit status for publishing. A published garden first as a separate folder of cloned notes, then as tagged for publication? Needs a 'touching' process to see diverging note/clone.
-
Publishing is Pausing: You push a rough idea out to pause your thinking and invite external input.
Like the phrase. Vgl [[Kunst-artefact is (tussen)uitkomst proces 20140505070232]] any artefact is a (intermediate) output of ongoing process. Vgl [[Artefacten uit kennis met methoden 20220215202011]], artefacts brought forth from K through methods.
-
What Is a Federated Garden? A federated digital garden is a personal knowledge space that: Starts private (your vault, your chaos, your unfinished thoughts) Grows selectively public (some notes bloom into garden pages) Connects across platforms (one source → many destinations) Links bidirectionally (gardens can discover and reference each other) Key principles: Gradual disclosure: Not everything needs to be public—choose what to share Version multiplicity: A note can exist as a draft, a garden page, a newsletter piece, and a toot Decentralized ownership: You control the source, not a platform Interoperability: Uses open standards (Markdown, RSS, ActivityPub, HTML)
The page confuses tools and content imo, but this is a workable list of attributes. 'Grows selectively public' is the key imo. #openvraag #webbeheer What can I do to improve the note to webpage pipeline I have but seldom use?
-
My goal in this is to treat knowledge as a living, collaborative process, not a series of finished articles.
yes, but 'finished article' never was K. Artefacts may be carriers.
-
-
mnl.dares.nl mnl.dares.nl
-
Welkom op de website van Dares regio Midden Nederland
Regio Utrecht no activity on site since late 2023.
-
-
-
Ontdek hoe wij bijdragen aan een veiliger Nederland en hoe u deel kunt uitmaken van ons team van gepassioneerde radioamateurs.Door het Ministerie van Binnenlandse Zaken is aan de veiligheidsregio’s geadviseerd om de dienstverlening van DARES op te nemen in de rampenplannen.
The DARES volunteers are actively preparing for emergency situations. Veiligheidsregio's sign convenants with them.
-
DARES, anbi for emergency comms in support of public admin needs.
-
-
meshtastic.org meshtastic.org
-
nRF52 The nRF52 chip is much more power efficient than the ESP32 chip and easier to update, but is only equipped with Bluetooth.
nRF52 is more power efficient, but only has BT, no wifi. Check which one of the boards mentioned provides wired ports?
-
ESP32 The ESP32 chip is older and consumes more power than the nRF52 chip, but is equipped with both WiFi and Bluetooth. Supported ESP32 devices include:
ESP32 MCU supports both wifi and BT. Higher power consumption though.
In general, as this network is supposed to provide alternative comms when regular networks break down: how reliant on battery / solar powered devices is the network?
-
Meshtastic hardware needs specific MCUs.
Tags
Annotators
URL
-
-
meshtastic.org meshtastic.org
-
Bitcoin miners have deployed LoRa for mining networks in the first 2020s. Currently most are no longer in use.
-
-
play.google.com play.google.com
-
Meshtastic app. It provides an interface to connect to a node. None visible in my vicinity (assuming nodes broadcast their presence)
Tags
Annotators
URL
-
-
opencorporates.com opencorporates.com
-
530-B HARKLE ROAD, STE 100, Santa Fe, NM, 87505, USA
Meshtastic LLC is registered in Sante Fe, New Mexico, USA. That information is not contained in the meshtastic website.
The head office address given is that of a registration agent though: New Mexico Registered Agent by High Desert Corporate Filings LLC
Tags
Annotators
URL
-
-
meshtastic.org meshtastic.org
-
Meshtastic® is a registered trademark of Meshtastic LLC
Meshtastic trademark owned/controlled by Meshtastic LLC. US based I presume (given that examples I saw on the site take USA as default location).
-
-
opencorporates.com opencorporates.com
-
Meshtastic LLC (and Meshtastic Solutions LLC, a service provider on top of the network it seems) are USA, New Mexico registered corporate limited liability entities.
-
-
meshtastic.org meshtastic.org
-
When you send a message on your Meshtastic companion app, it is relayed to the radio using Bluetooth, Wi-Fi/Ethernet or serial connection. That message is then broadcasted by the radio. If it hasn't received a confirmation from any other device after a certain timeout, it will retransmit the message up to three times. When a receiving radio captures a packet, it checks to see if it has heard that message before. If it has it ignores the message. If it hasn't heard the message, it will rebroadcast it. For each message a radio rebroadcasts, it marks the "hop limit" down by one. When a radio receives a packet with a hop limit of zero, it will not rebroadcast the message. The radio will store a small amount of packets (around 30) in its memory for when it's not connected to a client app. If it's full, it will replace the oldest packets with newly incoming text messages only.
You use your phone or 'companion app'(?) to send a msg to a radio (over BT, wifi or wire). The radio broadcasts incoming messages, including the one you provide through the app.
Msgs that are not acknowledged by another radion will be send at most 3 times. (will you be able to see it has not propagated?)
A radio that receives msgs already received will not rebroadcast it. Any broadcasted msg has a 'hop limit' and if it hits 0 it will not be rebroadcast. This limits the spread of a message, no? What is the default hop limit? Otoh the hoplimit does not limit the initial number of paths for broadcasting. So it's an attenuation over paths.
Theoretically in a dense network, my msg may reach Y number of other radios that all start out with the same hop limit.
I do not see here yet how you could intentionally set and reach a specific recipient. This description provides attenuated propagation of messages but no direction/addressee?
-
-
meshtastic.org meshtastic.org
-
Additionally, Meshtastic radios can be paired with a single phone, allowing friends and family to send messages directly to your specific radio. It's important to note that each device is capable of supporting a connection from only one user at a time.
Any meshtastic radio can connect a single phone for receiving / sending (?) messages on. Apart from repeater function 1 single connection for actual comms input?
-
These radios are designed to rebroadcast messages they receive, forming a mesh network.
Any meshtastic device is also a repeater?
-
Meshtastic utilizes LoRa, a long-range radio protocol, which is widely accessible in most regions without the need for additional licenses or certifications, unlike ham radio operations.
Meshtastic positions itself as a radio protocol without need for licensing such as in ham radio. Meaning it's fully in parallel to ham emergency networks like the Dutch DARES.
-
-
meshtastic.org meshtastic.org
-
There is no NL based local group mentioned at Meshtastic
Tags
Annotators
URL
-
-
meshtastic.org meshtastic.org
-
https://web.archive.org/web/20251205114551/https://meshtastic.org/ Meshtastic is a LoRaWan based comms network. It does not interoperate with other things like my Things LoRaWan gateway.
-
-
www.tomshardware.com www.tomshardware.com
-
https://web.archive.org/web/20251205111520/https://www.tomshardware.com/tech-industry/artificial-intelligence/googles-agentic-ai-wipes-users-entire-hard-drive-without-permission-after-misinterpreting-instructions-to-clear-a-cache-i-am-deeply-deeply-sorry-this-is-a-critical-failure-on-my-part Agentic AI deletes entire HDD, when it was supposed to only delete a cache folder (rmdir in root folder, not the projectfolder....bc it does not know about the world)
Tags
Annotators
URL
-
-
www.infoq.com www.infoq.com
-
https://web.archive.org/web/20251204124437/https://www.infoq.com/podcasts/looking-root-causes-false-path/ RCA vgl [[Triz root-cause-analysis 20200826114715]] can be source of reductionism [[Avoid greedy reductionism 20041114065928]] in complex settings.
-
-
www.simonandschuster.com www.simonandschuster.com
-
Buy [[We Will Rise Again by Karen Lord Annalee Newitz Malka Older eds]]
-
-
jenvgegevens.pleio.nl jenvgegevens.pleio.nl
-
MinJenV Afsprakenstelsel Gegevens en Algoritmes (JAGA) -[ ] maak Notie v JAGA #geonovumtb #eudata
-
-
jenvgegevens.pleio.nl jenvgegevens.pleio.nl
-
MinJenV stelt afsprakenstelsel gegevens en algoritmes vast. JAGA. #2025/11/27 En gegevenstyperingsbeleid
-
-
www.americanbar.org www.americanbar.org
-
via [[Chris Corrigan]]
civic organisation as a skillset is disappearing in the US but is a key building block of functioning democracies. Vgl #netag
-
-
mickryan.substack.com mickryan.substack.com
-
Mick Ryan compares his early 2025 notes with how this year went in Ukraine's war against the Russian fullscale invasion of 2022. Paywalled.
-
-
www.fairphone.com www.fairphone.com
-
Fairphone on their open source policy and choices.
My 2021 fp4 is now 4 yrs old, one more year at least of updates.
-
-
www.eljadaae.nl www.eljadaae.nl
-
Hieronder de lijst van AI-boeken die ik gelezen heb en je aan kan raden. Klik meteen door naar de langere omschrijving of scroll verder. Ze staan op de volgorde waarin ik ze uitgelezen heb: Weapons of Math Destruction: over desastreuze algoritmes Code Dependent: over de achterkant van AI Onze kunstmatige toekomst: over de etische kant van AI Empire of AI: over de opkomst van OpenAI Your face belongs to us: over de opkomst van ClearView AI Atlas van de digitale wereld: over de geo-politiek van AI The Digital Republic: over het reguleren van technologie Toezicht houden in het tijdperk van AI: over de juiste vragen stellen over AI
[[Elja Daae]] recommended reading list wrt AI [[Weapons of Math Destruction by Cathy O Neil]] (have it since 2017) [[Code Dependent by Madhumita Murgia]] bought it in August in ramsj [[The Digital Republic by Jamie Susskind]] I noted in 2024 as possible reading. [[Atlas van de digitale wereld by Haroon Sheikh]] I have too Other's are unknown to me. Interesting list, as it shaped their view on their role in AI public policy I presume
-
The Digital Republic van Jamie Susskind (2023) is nogal een boekwerk. Niet iets dat je in een avondje uitleest. Maar het is wel een heel belangrijk boek, want het gaat over de vraag hoe we als samenleving technologie kunnen reguleren.
[[The Digital Republic by Jamie Susskind]], already jotted down the title [[Daglog 29-10-2024]] at Dussmann's in Berlin, 2023 book. [[Elja Daae]] recommends it in this list. - [ ] check for review / summary [[The Digital Republic by Jamie Susskind]] #digitalpolicy #reading
Tags
Annotators
URL
-
-
beat.doebe.li beat.doebe.li
-
2012 conf volume Found PDF and downloaded it to Zotero
[[Cultures and Ethics of Sharing by Wolfgang Sützl Felix Stalder Theo Hug eds]]
Tags
Annotators
URL
-
-
en.wikipedia.org en.wikipedia.org
-
[[The Digital Republic by Jamie Susskind]] author is UK law/ currently in USA it seems (Berkman)
Tags
Annotators
URL
-
-
www.boekenwereld.com www.boekenwereld.com
-
[[Toezicht houden in het AI-tijdperk by Esther van Egerschot Marco Florijn]] mbt toezicht/politieke rollen vs AI vragen. Wellicht interessant taalgebruik/framing om uit te putten?
-
-
www.marketplace.org www.marketplace.org
-
#2024/10 article describing how private equity firms are buying up skilled trades companies. Plumbers, solar panel / heat pump installers etc. A weak signal imo. Would you want private equity in real world entities? It's one thing for 'new fields' or scale-ups etc. Another when extraction hits trades we rely on (although easy enough for other entrants to help circumvent)
(via [[Chris Aldrich]])
-
- Nov 2025
-
website-translation.language-tools.ec.europa.eu website-translation.language-tools.ec.europa.eu
-
Web-T is a EC created plug-in for various CMSs to translate. site into all official EU languages. Installed it, not yet activated.
Plugin can be downloaded here, in the WP pluging platform or from github repo. floss
-
-
itrade.gov.il itrade.gov.il
-
Israel exports water, due to desalination and extensive re-use.
-
-
en.wikipedia.org en.wikipedia.org
-
Desalination, some numbers on global growth. Does not mention that Israel is an exporter of drinking water bc of desalination.
Tags
Annotators
URL
-
-
www.theguardian.com www.theguardian.com
-
Guardian article on groundwater reserves in Europe diminishing. Vgl NL current sensors https://droogteportaal.nl/droogteportaal/web/embedded_map.html?index=sgi_0&index_type=current&max_screen_depth=10
No mention of potential alternatives. Water does not disappear after all, it does become harder to use directly as drinking water. The gap is where technology options may exist, costly or not.
-
-
www.politico.eu www.politico.eu
-
In the meantime, “those who wanted to maintain the status quo have won
yes, it's key to break status quo early and in several small ways simultaneously, to signal it is at all possible. Vgl [[Een goed gesprek over digitale soevereiniteit in de gemeente]]
-
American cloud giants Amazon, Microsoft and Google, as well as Chinese tech giants Huawei and Alibaba, are all members of Gaia-X. In 2021, the annual summit in Milan was sponsored by Huawei and Alibaba, prompting backlash.
Same is happening wrt IDSA, who have a global orientation, but are treated as EU grouping, which they're not.
-
Those firms “steered the entire roadmap,” Lechelle said, throwing money and people at it. “The committees were drowning. They [global players] had the capacity, the bandwidth, but we were already underwater ... Americans have full-time lobbyists and massive budgets. Their job is basically to derail any initiative they don’t like.”
Key friction. One cannot rule out non-EU parties mostly, esp their EU entities. But presence often obstructive / malicious compliance. You'd need much better governance / rulebooks upfront to flag and remove.
-
of a data space based on Gaia-X standards that French energy company EDF will use to securely coordinate the construction of new nuclear sites
example of DS project, here led by EDF for nuclear plant construction. Link w energy DS program? -[ ] zoek naar EDF DS voorbeeld voor bouw kerncentrales #geonovumtb/ds
-
That’s how the mission to create a “federated cloud infrastructure” came to life. But that “staggering complexity” would soon turn into an “unmanageable mess,” said Lechelle.
federated cloud concept originates in Gaia-X article says. Don't know if that's right. In itself that does not create 'staggering complexity' though. It is a different design path.
-
“I joined Gaia-X because I believed in the original mission. I left Gaia-X because I didn’t believe it was going in the original direction,” said its former CEO, Francesco Bonfiglio.
Gaia-X moved the goalposts original member says.
-
The results we’re providing and the real business benefits these interoperable data spaces are creating are more and more visible,” he said, highlighting the example of a data space based on Gaia-X standards that French energy company EDF will use to securely coordinate the construction of new nuclear sites.
again an impact example that is still in the future, a planned result, not an achieved one.
-
Current CEO Ulrich Ahle, who joined in 2023, pushed back — saying Gaia-X is far from a “failure.” It has united the industry — both large and small players — around tangible deliverables, such as federated data spaces and compliance labels, he said.
federated data spaces are not an outcome of gaiax, are they? EC induced. Do note that deliverables after all those years are still in the future.
-
Politico on Gaia-X failures , based on diffs between French and German approaches.
-
-
www.csoonline.com www.csoonline.com
-
Named anchors in URLs can be used for prompt injection in AI browser assistants. # URL parts are only evaluated in browser, and not send to servers. AI assistants in browsers do read them though.
-
-
-
https://web.archive.org/web/20251129105036/https://www.nature.com/articles/d41586-025-03506-6
For an international AI conf, a chunk of papers was generated, but also 21% of the peer reviews on those papers was. human centipede epistemology is here vgl [[Talk The Expanding Dark Forest and Generative AI]]
Tags
Annotators
URL
-
-
www.yesigiveafig.com www.yesigiveafig.com
-
On the erosion of middle class America. Poverty line is around 140k if actual costs taken into account. The 1960s benchmark assumed cost of food to be 1/3 of overall costs. Now it's 7% or so, meaning 1/15 of overall costs. This pushes up the poverty line to 5 times the level used, or some 150k USD pa
Example of a proxy being used as 'measurement' and the assumptions in a proxy never re-evaluated.
-
-
lifestyle.sustainability-directory.com lifestyle.sustainability-directory.com
-
This page lifts quite a bit out of my descriptions of networked agency, a term I coined 2016, without attribution. It reads like a generated text. Unclear too what the point of the site is at all.
-
-
actsofvolition.com actsofvolition.com
-
Resumability is not a word, but it’s an important concept to me. When I say Resumability, I’m talking about the ability to quickly interrupt and later resume a task.
[[Steven Garrity]] coining 'resumability' , where a device just continues on where you were previously. Mentions e-readers as such devices (and paper books with a bookmark), a smartphone and apps like Slack. The latter is not true I think, it scrolls fwd to newest.
Thinking about resumability in terms of notes / pkm. Obsidian is always where I was previously e.g. A type of ratchet for task execution.
Tags
Annotators
URL
-
-
ruk.ca ruk.ca
-
For example, you may need a working shop or a working painting studio. You may need a working music studio. Or a computer room where you can write something. It’s crucial to have a setup, so that, at any given moment, when you get an idea, you have the place and the tools to make it happen.
Quote David Lynch https://en.wikipedia.org/wiki/Catching_the_Big_Fish continued.
Mentions examples of set-ups for different activities. Vgl spaces on laptop. Vgl my home office vs attic space etc.
Vgl [[%index coachingboekje]] wrt set-ups I need/want/woud like to have or list of set-ups currently available to me.
-
If you don’t have a setup, there are many times when you get the inspiration, the idea, but you have no tools, no place to put it together. And the idea just sits there and festers. Over time, it will go away. You didn’t fulfill it—and that’s just a heartache.
Quote David Lynch https://en.wikipedia.org/wiki/Catching_the_Big_Fish continued.
Not having a set-up ready when an idea hits means no agency. Mentions festering a type of powerlessness, vgl [[%I Networked Agency]] wrt methods and tools for various things. Also vgl making note of any idea in 3Ideeenkweekkas
-
“Okay, today I’m going into the shop to work on a lamp.” This idea comes to you, you can see it, but to accomplish it you need what I call a “setup.”
David Lynch quoted from https://en.wikipedia.org/wiki/Catching_the_Big_Fish about the necessity of a set-up to have agency and act on an idea.
-
David Lynch wrote this in Catching the Big Fish:
-
[[Peter Rukavina]] on great set-ups. Take this as a thinking aid vgl [[%index coachingboekje]] wrt needed / ideal settings for types of activity.
Tags
Annotators
URL
-
-
en.wikipedia.org en.wikipedia.org
-
Lynch's idea that "ideas are like fish. If you want to catch little fish, you can stay in the shallow water. But if you want to catch the big fish, you've got to go deeper". To Lynch, going deeper means experiencing a deeper, more expanded state of consciousness, a transcendental or fourth state of consciousness,[2] an experience he has during meditation but believes is rare in ordinary daily life.[6] According to Lynch, this experience expands artistic capacity.[4]
Source of title Catching the Big Fish. Aside from the meditation angle, this points to practice / reflection, ratchets, and [[Holding questions 20091015123253]] etc.
-
-
www.volkskrant.nl www.volkskrant.nl
-
20+ attempts daily
-
-
-
Funny. Cryptographers club held elections, but cannot access results as one of the three keys needed to see results was lost by one of the people involved. They did not design for human failure / lapses in operational precision. Typical. Vgl the cartoon with laptop w unbreakable login and person taking a bat to the one who knows the login.
-
-
teachinginhighered.com teachinginhighered.com
-
It may seem strange that I like remembering who recommended things to me, after the fact. To me, that's part of my sensemaking and ongoing relationship deepening habits.
Yes! key part of social filtering. Recommendations and source, in ones network is a key piece of metadata. I think it's not just likable but necessary.
-
Jarche shares 14 ways to acquire knowledge from the quintessential PKM practicer, Maria Popova at The Marginalian, and her review _You Can Do Anything_ by James Mangan, written in 1936. He then categorizes the methods in terms of how they align with PKM in this graphic from Jarche:
Maria Popova https://en.wikipedia.org/wiki/Maria_Popova http://themarginalian.org/
[[You Can Do Anything by James Mangan]] 1936 review: 14 ways to acquire K. https://www.themarginalian.org/2013/04/22/14-ways-to-acquire-knowledge-james-mangan-1936/ "prolific self-help guru and famous eccentric" https://archive.org/details/bwb_W8-ANG-369/ can be borrowed.
-
-
pkm.jarche.ca pkm.jarche.ca
-
Image by [[Harold Jarche]] plotting 14 modes of learning from [[You Can Do Anything by James Mangan]] 1936 self-help book on his Seek / Sense / Share
-
-
www.youtube.com www.youtube.com
-
Jorge Arango talks about his writing process. He actually co-authored a book on Information Architecture in 2015. This vid more about his latest. uses lots of paper notebooks Mentions 3 (obvious) phases in his non-fiction writing research structuring writing Says all three req diff tools. (he uses Obsidian f research)
Stage 1 research, vague idea about what topic, (s paper notebooks, index/sequence them) His book duly noted is on this phase.
Stage 2 structuring. mistake to go directly to writing. (vgl [[A System for Writing by Bob Doto]] ) Mentions author Robert Caro who had book structure on his wall story book like, vgl what I do when writing reports w wall of post its for structure. Limitation is size of wall. But physcial movement helpt. He does this type of wall in [[Tinderbox]] !!!! canvas, with the sticky notes having metadata. His chapter and heading structure emerges from it! Prevents getting lost in the details. Shows him what is missing too (looping back to research phase sometimes). It's not writing but getting a sense of the narrative flow of the entire thing. Seems like outlining of talks, but a book is bigger and thus the outlining is more filled out with links to material
What question would a reader have entering chapter 2 after reading chapter 1, and go from there. Has opening story/anecdote for each chapter. Mwah. Uses the story board view like how the periodic table suggested the existence of atoms that had not been discovered yet. Dmitri Mendleev.
Stage 3 writing After structuring the writing tools come into play. He used Scrivener, avoiding having a monolithic text. Now Ulysses, in markdown. Uses word count targets in these tools for sections/chapters
set up supportive environments for each pahse structure first but keep flexible, don't make it a linear process, but structure ulnocks the writing swith modalities - walking, free form in [[Tinderbox]] or another tool, to avoid getting stuck.
-
-
betterimagesofai.org betterimagesofai.org
-
Tip from colleague C for images on AI that break the usual (anthropomorphic) frame. CC-licensed for re-use usually.
Tags
Annotators
URL
-
-
www.nvidia.com www.nvidia.com
-
Nvidia Cosmos world (foundation) models. Avalailable on github. 'for physical AI', for use in training autonomous vehicles, robots and video analytics. E.g. to generate videos and 3d worlds.
-
-
www.forbes.com www.forbes.com
-
This transition is signaled by focused efforts from several major scientists and technology entities. Meta Chief AI Scientist Yann LeCun has emphasized his intent to pursue world models, while Fei-Fei Li’s World Labs has released its Marble model publicly. Concurrently, Google is testing its Genie models, and Nvidia is developing its Omniverse and Cosmos platforms for physical AI.
Various examples of world model work: Nex to Yann LeCun. Fei-Fei Li World Labs w Marble model, Google has Genie models, Nvidia Omniverse and Cosmos.
-
-
dcx04g4gb0w75m.archive.ph dcx04g4gb0w75m.archive.ph
-
On Tuesday, news broke that he may soon be leaving Meta to pursue a startup focused on so-called world models, technology that LeCun thinks is more likely to advance the state of AI than Meta’s current language models.
Yann LeCun says world models more promising. What are world models?
Tags
Annotators
URL
-
-
-
De bank maakte woensdag ook cijfers over het derde kwartaal bekend. Daarin boekte ABN een winst van 617 miljoen euro, op een omzet van 2,16 miljard. Beide zijn wat lager dan in dezelfde periode het jaar ervoor.
Wut? 617/2160 is 28,5% winst in het 3e kwartaal. 'lager dan vorig jaar'. Het lijkt me een heel hoog winstpercentage.
-
-
silverbullet.md silverbullet.md
-
[[Links]] like a wiki (except they are [[Linked Mentions|bi-directional]]).
The bidirectional takes the form of linked mentions underneath a linked to note. That's common even in wiki I think these days. Any tools out there that link closer to like Ted Nelson proposed? Within a pkm tool bidirectional linking into a specific spot in a note should be somehow doable. Transclusion is a step further even, but if I can link to a paragraph, that paragraph should be able to link to where it is cited, no? In Obsidian the ^ link is still one way, while origin and source are known for it to work (the reference number)
-
Silverbullet, making the rounds in #pkm tooling circles last few days. Works on file system, but that needs to be on a web server. Client is a progressive web app. Browser loads all of it, and then it's local on device and can work offline.
Tags
Annotators
URL
-
-
eurollm.io eurollm.io
-
EuroLLM for all official 24 EU languages. WTF is up with that .io TLD?
-
-
en.wikipedia.org en.wikipedia.org
-
George Gilder. Reagan era supply side economist, 85yo, Called the opportunities of internet tech relatively early, and wrote a book about Carver Mead and CMOS chiptech. Founder of the Discovery Institute and loud on 'intelligent design'. Not sure any of that qualifies him to talk about wafer sized chips systems.
Tags
Annotators
URL
-
-
-
George Gilder indeed calls for wafer scale compute. Article mentions the issues of one flaw meaning having to toss the wafer (as opposed to 1 chip cut from the wafer). Also says that bc of that wafer solutions are for highly specialised set-ups only (bc it's costly to end up with one flawless wafer).
-
-
www.discovery.org www.discovery.org
-
WSJ article contends that chips will soon be obsolete, and replaced by wafers that function as a fully integrated system. I have questions (and most of the article is missing). Wafers are error prone, and while you can toss out individual chips cut from wafers, having to throw out an entire wafer is costly. Other factor is that whatever size you will have surrounding electronics too, making wafers unwieldy / brittle parts of a system. Also wondering about heat. Talks about 'suitcase sized datacenters'. I wonder what the technology jump is for that to be viable
-
-
www.mckinsey.com www.mckinsey.com
-
n our latest findings, the share of respondents reporting mitigation efforts for risks such as personal and individual privacy, explainability, organizational reputation, and regulatory compliance has grown since we last asked about risks associated with AI overall in 2022.
did they also ask whether those mitigation efforts negate gains in efficiency / innovation reported for AI?
-
While a plurality of respondents expect to see little or no effect on their organizations’ total number of employees in the year ahead, 32 percent predict an overall reduction of 3 percent or more, and 13 percent predict an increase of that magnitude (Exhibit 17). Respondents at larger organizations are more likely than those at smaller ones to expect an enterprise-wide AI-related reduction in workforce size, while AI high performers are more likely than others are to expect a meaningful change, either in the form of workforce reductions or increases.
Interesting to see companies vary in their est of how AI will impact workforce. A third expects reduction (but not much, about 3%), 13% an increase (AI related hiring), 43% no change.
-
with nearly one-third of all respondents reporting consequences stemming from AI inaccuracy (Exhibit 19).
A third of respondents admit they've seen 'at least once' negative consequences of inaccurate output. That sounds low, as 100% will have been given hallucinations. So 1-in-3 doesn't catch them all before they run-up damage. (vgl Deloitte's work in Australia)
-
The online survey was in the field from June 25 to July 29, 2025, and garnered responses from 1,993 participants in 105 nations representing the full range of regions, industries, company sizes, functional specialties, and tenures. Thirty-eight percent of respondents say they work for organizations with more than $1 billion in annual revenues. To adjust for differences in response rates, the data are weighted by the contribution of each respondent’s nation to global GDP.
2k self selected respondents in 50% of nations. 4/10 are big corporates (over 1 billion USD annual revenue)
-
McKinsey survey on AI use in corporations, esp perceptions and expectations. No actual measurements. I suspect it mostly measure the level of hype that respondents currently buy into.
-
-
ibestuur.nl ibestuur.nl
-
ibestuur over contouren van digitale soevereiniteit.
-
-
apnews.com apnews.com
-
SoftBank Japan cashing out on NVIDIA. Is this a 1st step down the hype curve into the trough? Although they are also highly invested in OpenAI.
-
-
-
AI checking AI inherits vulnerabilities, Hays warned. "Transparency gaps, prompt injection vulnerabilities and a decision-making chain becomes harder to trace with each layer you add." Her research at Salesforce revealed that 55% of IT security leaders lack confidence that they have appropriate guardrails to deploy agents safely.
abstracting away responsibilities is a dead-end. Over half of IT security think now no way to deploy agentic AI safely.
-
When two models share similar data foundations or training biases, one may simply validate the other's errors faster and more convincingly. The result is what McDonagh-Smith describes as "an echo chamber, machines confidently agreeing on the same mistake." This is fundamentally epistemic rather than technical, he said, undermining our ability to know whether oversight mechanisms work at all.
Similarity between models / training data creates an epistemic issue. Using them to control each other creates an echo chamber. Vgl [[Deontologische provenance 20240318113250]]
-
Yet most organizations remain unprepared. When Bertini talks with product and design teams, she said she finds that "almost none have actually built it into their systems or workflows yet," treating human oversight as nice-to-have rather than foundational.
Suggested that no AI using companies are actively prepping for AI Act's rules wrt human oversight.
-
We're seeing the rise of a 'human on the loop' paradigm where people still define intent, context and accountability, whilst co-ordinating the machines' management of scale and speed," he explained.
Human on the loop vs in
-
As Gartner VP analyst Alicia Mullery put it: “AI can make mistakes faster than we humans can catch them.”
yes, example of [[Spammy handelings asymmetrie 20201220072726]]. At scale it moves the bottleneck
-
piece on AI oversight.
In general I wonder, at what point does the needed oversight negate the gains in time / effectivity / efficiency that are expected of using AI in some context.
-
-
www.theguardian.com www.theguardian.com
-
strategic issues wrt China. Nexperia is mentioned but not explained (e.g. as to why Dutch gov stepped in), something I find lacking overall. Search for comms by Dutch gov on this?
-
-
-
US people under 40 show doubling of self reported cognitive issues over 2013-2023. No conclusion as to cause. Paper in Rising Cognitive Disability as a Public Health Concern Among US Adults: Trends From the Behavioral Risk Factor Surveillance System, 2013–2023 in Zotero
-
-
netzpolitik.org netzpolitik.org
-
Unklar ist derzeit offenbar noch, ob die weitere Umsetzung der KI-Verordnung in Teilen aufgeschoben wird.
AI Act application might be postponed in parts
-
Die Kommission will erreichen, dass sensible Daten enger definiert werden. Besonders geschützt wären dann nur noch jene Daten, die oben genannte Informationen explizit offenbaren. Das bedeutet: Gibt etwa eine Person in einem Auswahlfeld an, welche sexuelle Orientierung sie hat, wäre das weiterhin besonders geschützt. Schließt ein Datenverarbeiter aufgrund vermeintlicher Interessen oder Merkmale auf die mutmaßliche sexuelle Orientierung eines Menschen, würden bisherige Einschränkungen wegfallen.
Special personal data not protected if inferred, only if explicitly collected as such?
-
sobald entsprechende Standards verfügbar sind“.
standards expected wrt handling of cookie prefs
-
Das gilt etwa für das Training von KI-Systemen mit personenbezogenen Daten. Dies soll künftig auf Basis des berechtigten Interesses von Tech-Konzernen möglich sein. Die noch immer heftig geführte Debatte um möglicherweise notwendige Einwilligungen der Betroffenen hätte sich damit erledigt. Deutlich enger gefasst werden soll auch die Definition von pseudonymisierten Daten.
Legitimite interest is being stretched it seems? I thought it was pretty strict wrt things an org would be required to collect / needs to collect for a necessary task.
-
Im neuen Data Act sollen gleich drei weitere Gesetze aufgehen: die Open-Data-Richtlinie, die Verordnung über den freien Fluss nicht-personenbezogener Daten und der Data Governance Act.
The DA will absorb ODD, DGA and free flow non-personal data Interesting, as ODD is a Directive now. ODD is based on national access regimes, so might be tricky as Reg, unless the focus is on HVD mostly.
-
-
github.com github.com
-
FeedMe is an Android app that does support a FreshRSS back-end.
The app's interface to set-up the api connection is not clear at all. Requires a http (not s) connection through the IP address.
Installed and uninstalled again
Tags
Annotators
URL
-
-
-
Feeder android app does not support using FreshRSS API
Tags
Annotators
URL
-
-
sandgrain.eu sandgrain.eu
-
Sandgrain website. unique ID for chips, verified in a secured cloud enviro. Focus on IoT devices.
-
-
-
erst bij elektriciteitsnetwerken, zoals netbeheerder Tennet. Transformatorhuisjes worden aangestuurd via internetverbindingen en die zijn kwetsbaar
First area of application is Dutch electricity network. Ask S.K.?
-
Er zijn andere manieren om hardware een eigen nummer te geven, zoals de PUF (physical unclonable function) in een chipontwerp. PUF-codes worden niet ingeëtst, maar zijn gebaseerd op een statistisch sommetje in het chipontwerp.
It is not PUF, which is on chip, and sensitive to environmental factors, as well as taking space on the wafer.
-
Unieke chips helpen de schade te beperken. Het toevoegen van zo’n nummer hoeft niet veel te kosten: dat kan in een later stadium van de chipproductie met goedkope apparatuur.
the ID is added after the hightech productionphase making it affordable.
-
De controle van de identificatie vindt plaats op een streng beveiligde server en daarbij worden geen sleutels uitgewisseld, zoals bij gangbare beveiliging. Niemand in de hele productieketen kent de sleutel, ook voor SandGrain is het een geheim. Zelfs al zouden criminelen het systeem weten te omzeilen – wat volgens de SandGrainers niet kan – dan hebben ze toegang tot maar één apparaat. Dat maakt het veel ingewikkelder om een volledig netwerk te saboteren.
Sandgrain is key less and zero-K. ID is done through a (central?) server though.
Circumventing it gains access to just one device, not all devices that use the same chip.
-
Het principe is simpel: een chip met een uniek nummer bewaakt de toegang tot de achterliggende elektronica. Alleen als dat nummer wordt herkend, krijgt de gebruiker toegang tot het systee
The ID serves as a gatekeeper for access to the system it is used in.
-
Een nieuw Nederlands initiatief werkt aan chips met een uniek nummer, die als een slot op de deur dienen voor elektronica in cruciale infrastructuur.
SandGrain (Eindhovense spinoff, Joeri Voets en Sander Koopmans) provides every chip with a unique ID.
-
-
www.mondadoristore.it www.mondadoristore.it
-
https://www.mondadoristore.it/ Italian site for ebook purchases
-
-
www.sciencedirect.com www.sciencedirect.com
-
Beware of botshit: How to manage the epistemic risks of generative chatbots
Great term for incorrect AI slop uncritically used by people, botshit. Beware of botshit: How to manage the epistemic risks of generative chatbots in Zotero
-
-
www.kobo.com www.kobo.com
-
[[Pensare Bene by Luca De Biase]] is available via Kobo Plus
-
-
www.lafeltrinelli.it www.lafeltrinelli.it
-
https://www.lafeltrinelli.it/ Feltrinelli is an Italian online bookshop
-
-
-
ibs.it is an Italian online bookshop
-
-
www.librerie.coop www.librerie.coop
-
https://www.librerie.coop/ is an Italian platform (coop) for fiction and non-fiction books.
-
-
www.treccanilibri.it www.treccanilibri.it
-
Prezzo: € 9,99 Collana: Visioni Disponibile dal: 26/09/2025 ISBN: 9788812012992
Ebook version is 9,99 in Kobo.
-
un contributo a un ipotetico movimento culturale per un’ecologia della conoscenza in un ambiente mediatico a dir poco controverso. Perché non si pensa bene se si è informati male. Lo sviluppo dell’attuale forma di economia digitale, realizzato con una strategia di innovazioni disattenta alla qualità delle relazioni, ha generato esiti culturali evidentemente negativi, dei quali è tempo che la società si faccia carico, ripensando il sistema dei media in modo da renderlo compatibile con gli obiettivi democratici, ponendo fine al senso di ineluttabilità diffuso.
The book aims for a cultural movement / ecology of knowledge in spite of the toxic platforms. Because those erode thinking well (title). Innovation now disregards the quality of relationships. (Nice, this chimes with my [[Menselijk en digitaal netwerk zijn gelijksoortig 20200810142551]] and the unfulfilled potential of exploring that congruence). The book proposes to review the mediasystem to make it align with democratic values.
-
[[Pensare Bene by Luca De Biase]] [[Luca De Biase]] 'a media ecology for the 21st century.'
New book by Luca, published Sept 2025.
-
-
blog.debiase.com blog.debiase.com
-
Nel mio libro, questo conduce a costruire un progetto di liberazione dal dominio delle mega-piattaforme attuali e suggerisce la possibilità di costruire molte piattaforme alternative a quelle gigantesche e sfruttatrici che attualmente costituiscono il sistema dei media. Nel libro di Ferraris, questo conduce a una proposta di redistribuzione del valore dei dati, in funzione di un progetto politico che porti a un nuovo, moderno, sistemico e pragmatico comunismo.
Luca compares two books (his and Ferraris') in how they solve that relative value problem. Luca by proposing dismantling the dominant platforms and replace them with a multitude of others (as the volume of personal data aids its exploitation I presume, spreading it out means having to collect it and collection is costly if users don't actively bring it to you already). Ferraris proposes otoh an active redistribution of the value gained as political project a new 'pragmatic' and modern communism. De Biase proposal seems more achievable imo (in the sense that we're already doing it)
-
È ovvio: il valore dei dati personali non esiste in sé ma in relazione a un contesto.
Often not acknowledged what Luca says here: that it's obvious that n:: the value of personal data is not intrinsic but w.r.t. to a specific context of use. It's relative (and that is also why data protection of personal data isn't absolute, but weighed against other factors.) Perhaps capture it as Notie.
-
-
en.wikipedia.org en.wikipedia.org
-
Implicate and explicate order, by [[David Bohm]] quantum theory concepts.
Tied it to the notion of the brain / human mind as a quantum device.
-
-
en.wikipedia.org en.wikipedia.org
-
Bohm advanced the view that quantum physics meant that the old Cartesian model of reality—that there are two kinds of substance, the mental and the physical, that somehow interact—was too limited. To complement it, he developed a mathematical and physical theory of "implicate" and "explicate" order.[3]
Implicate and explicate order.
-
[[David Bohm]] 1917-1992. US, theoretical physicist, influenced neuropsychology / philosophy of mind. Saw the brain as a quantum device it seems.
Have several of his books [[On Dialogue by David Bohm]] [[On Creativity by David Bohm]] and just bought [[Thought as a System by David Bohm]] #2025/11/10
-
-
www.wurman.com www.wurman.com
-
Information Anxiety
[[Information Anxiety by Richard Saul Wurman]] is the origin of the [[Informatie orden je altijd LATCH 20221106114119]] concept. This is from 1989, so no internet / digital considerations yet. In LATCH I wonder about the place of linking in it. Is the 1986 version aware of Ted Nelson's work 1960s. Or does linking come up in edition 2 in 2000: [[Information Anxiety 2 by Richard Saul Wurman]]
-
Website of [[Richard Saul Wurman 20251110184844]] Ordered some of his books (not very easy to find) [[Information Anxiety by Richard Saul Wurman]] 1989 [[Information Architect by Richard Saul Wurman]] 1996 [[Information Anxiety 2 by Richard Saul Wurman]] 2000 [[UnderstandingUnderstanding by Richard Saul Wurman]] 2017.
Tags
Annotators
URL
-
-
ajroach42.com ajroach42.com
-
2017 page on 'web without internet', thinks like DAT and IFPS. what do I take from this posting, with 8 years hindsight.
Tags
Annotators
URL
-
-
-
we demonstrate the same dynamics for poisoning during fine-tuning.
This poisoning also plays out the same way when fine-tuning. This adds another attack vector also for existing models when they are used.
-
his work demonstrates for the first time that poisoning attacks instead require a near-constant number of documents regardless of dataset size. We conduct the largest pretraining poisoning experiments to date, pretraining models from 600M to 13B parameters on chinchilla-optimal datasets (6B to 260B tokens). We find that 250 poisoned documents similarly compromise models across all model and dataset sizes, despite the largest models training on more than 20 times more clean data
The paper shows that it's not a percentage of training data that needs to be poisoned for an attack, but an almost fixed number of documents (250!) which is enough across large models too.
-
Existing work has studied pretraining poisoning assuming adversaries control a percentage of the training corpus. However, for large models, even small percentages translate to impractically large amounts of data.
It was previously assumed that a certain percentage of data needed to be 'poisoned' to attack an LLM. This becomes impractical quickly with the size of LLMs.
-
Poisoning Attacks on LLMs Require a Near-constant Number of Poison Samples in Zotero
-
-
oxrml.com oxrml.com
-
LLM benchmarks are essential for tracking progress and ensuring safety in AI, but most benchmarks don't measure what matters.
Paper concludes most benchmarks used for LLMs to establish progress are mistargeted / leave out aspects that matter.
-
PDF paper Saved Measuring what Matters: Construct Validity in Large Language Model Benchmarks in Zotero Paper for NeurIPS 2025 conference https://neurips.cc/Conferences/2025
Tags
Annotators
URL
-
-
slow-thoughts.com slow-thoughts.com
-
Her point is this: Metaphors are so often visual in nature, that we tend to equate understanding something with the ability to visualise it. Which explain why Einstein–always a visual thinkers–hated quantum mechanics. Because while the standard model helps making perfect mathematical sense of particle physics, it’s simply not possible to visualise what it proves to be true. But here’s the thing: metaphors don’t have to be visual in nature, and in fact going beyond the visual often allows us to naturally accommodate ambiguity. Trompe l’oeil images are just as maddening and hard to let go of as trying to visualise a quark that exists simultaneously in multiple places, but anyone can attest to feelings of ‘being torn‘ or ‘in two minds‘. Time is another metaphor that is notoriously hard to visualise, which hasn’t stopped anyone from experiencing it. Again it’s also a phenomenon that most of us feel behave in a highly irrational manner; slowing to a creep in one moment only to jump into action the next. The point that Julia Ravanis makes, the perspective she helps me see, is that quantum mechanics doesn’st have to ‘not make sense’. That the act of sense-making includes a chosen perspective, and that being mindful that there are more than one possible, even within science, means that the boundaries between it and the humanities are crumbling.
[[Julia Ravanis]] in [[Skönheten i Kaos by Julia Ravanis]] is here said to argue that a way of moving past 'quantum mechanics does not make sense' is by letting go of default (visual) metaphors and using other metaphors that can embrace ambiguity. This sounds somewhat like [[Is het nieuwe uit te leggen in taal van het oude 20031104104340]] or even [[Avoid greedy reductionism 20041114065928]] accusation levelled here at Einstein.
-
That the act of sense-making includes a chosen perspective, and that being mindful that there are more than one possible, even within science, means that the boundaries between it and the humanities are crumbling.
overtones of sense making, and of SC style boundary crossing, with the notion of a chosen perspective, and [[Multidimensionaal gaan ipv platslaan 20200826121720]] as means. #verdiepen
-
-
www.microsoft.com www.microsoft.com
-
all at orders of magnitude lower computational cost.
I presume disregarding the 'more than one million hours' of training in the order of magnitude comparison?
-
Aurora outperforms operational forecasts in predicting air quality, ocean waves, tropical cyclone tracks and high-resolution weather
Aurora said to extrapolate for air quality, bathymetry, cyclone tracking and 'high resolution' weather (I suspect they mean the opposite, check in paper).
-
Aurora is a machine learning model that can predict atmospheric variables, such as temperature. It is a foundation model, which means that it was first generally trained on a lot of data and then can be adapted to specialized atmospheric forecasting tasks with relatively little data. We provide four such specialized versions: one for medium-resolution weather prediction, one for high-resolution weather prediction, one for air pollution prediction, and one for ocean wave prediction.
MS created foundation model Aurora, trained on over '1 million hours of diverse geophysical data' (they mean 1 million compute hours??), to use to predict atmospheric variables (temp) in August 2024.
-
-
www.flickrhelp.com www.flickrhelp.com
-
500 photo limit also applies to albums
Any download will be at most 500 images.
-
The files will have the same EXIF data they had when originally updated. Any details added to your photos after upload will be compiled in a separate JSON file.
The downloads will include original exif data. If that has been altered (e.g. location or time (NB I adjusted some for diff timezones)) the alterations will be provided as JSON
-
Keep in mind that links to the requested data are time sensitive and will expire
Any flickr download link is available in account settings, but for a limited time.
-
Flickr downloads of my images can be done: - by selecting 500 images at a time in Camera roll. This would be I think the best approach for timeline approaches. You get a downloadable zip. - by downloading an album. as zip - by downloading all my data, as zip. The text implies it will be in chunks of 500 images too, so perhaps also chronological? (in my case 87 zip files or so)
-
-
tim.siosm.fr tim.siosm.fr
-
Description of moving away from Spotify to Qobuz. Spotify is Swedish but in the spotlight for hosting ICE ads in the USA at the moment. Qobuz is French, but hosts their stuff on AWS.
-
-
www.qobuz.com www.qobuz.com
-
Qobuz is a possible Spotify alternative. Qobuz is a French company (Xandrie s.a.), service started in 2008. They host their stuff on Amazon AWS though.
Tags
Annotators
URL
-
-
bryanalexander.org bryanalexander.org
-
[[Bryan Alexander]] gives an overview of his current socialmedia usage. #blogdit my 2007 overview of stuff vs now and why the changes https://www.zylstra.org/blog/2007/10/the_long_list_o/ [[Long list of my distributed self 20080530120000]]
-
-
fritanke.se fritanke.se
-
Kommer den artificiella intelligensen att bli bättre på att tänka än den mänskliga? Kognitionsvetaren Peter Gärdenfors förklarar varför så inte är fallet. Den mänskliga intelligensen består av en rad olika färdigheter och specialiteter som har förfinats under tusentals år. Mycket återstår innan den artificiella intelligensen kan mäta sig med det tänkande som inte bara människor utan även djur har. När vi förstår att vår intelligens är en bred palett av många olika förmågor ter sig tanken på att AI-tekniken trumfar oss i schack och kan skriva avancerade texter inte lika skrämmande. Utifrån ett brett forskningsunderlag förklarar Gärdenfors varför AI-tekniken inte kan och inte kommer att kunna tänka på samma sätt som människor och djur gör. »Peter Gärdenfors tilldelas Natur & Kulturs debattbokspris 2025 för att han fördjupar AI-debattens centrala begrepp och utmanar dess utgångspunkter. Med lätt språk och stabil lärdom blottlägger han tänkandets evolutionärt slipade mekanismer, och skärper bilden av vad intelligens är och vilken plats tekniken intar i vår digitala värld.« – Juryns motivering
[[Kan AI tänka by Peter Gärdenfors]] via Sven Dahlstrand, dahlstrand.net Publ okt 2024 Seeks to define what thinking actually is, and how that plays out in other animals and humans. The 2nd part goes into sofrware systems and AI and how they work in comparison.
Tags
Annotators
URL
-
-
juliaravanis.se juliaravanis.se
-
Skönheten i kaosSkönheten i kaos (Natur och Kultur 2021) är Julias debutbok.I den tar hon ner den teoretiska fysiken på jorden ochjämför den med mänskliga erfarenheter. Förklaringar avsvarta hål och sammanflätade elektroner varvas medreflektioner om längtan och frustration, om att bli kär ochkänna andra människors blick på en själv.
[[Skönheten i Kaos by Julia Ravanis]] (pub 2021, Swedish) Debut. Essay collection joining theoretical physics w philosophy Explains concepts like black holes, quantum entanglement, and string theory in an accessible form. (via Sven Dahlstrand, dahlstrand.net)
-
Julia Ravanis (1993), leeft in CPH. via Sven Dahlstrand dahlstrand.net nav [[Looking for Untranslated Non-Fiction From Across Europe]]
Civiel ir / TN, en phd in tech history.
Tags
Annotators
URL
-
- Oct 2025
-
positivepsychology.com positivepsychology.com
-
Circles of Control, Influence and Concern. Came up in DefDem meeting.
Tags
Annotators
URL
-
-
www.incompleteideas.net www.incompleteideas.net
-
Richard Sutton page, lists more articles than the folder I found.
-
-
www.incompleteideas.net www.incompleteideas.net
-
Some writings, incomplete ideas he calls them, by Richard Sutton. Straight-up HTML, no frills, in a folder. Nice.
Tags
Annotators
URL
-
-
www.dwarkesh.com www.dwarkesh.com
-
LLMs aren’t capable of learning on-the-job, so no matter how much we scale, we’ll need some new architecture to enable continual learning.And once we have it, we won’t need a special training phase — the agent will just learn on-the-fly, like all humans, and indeed, like all animals.This new paradigm will render our current approach with LLMs obsolete.
Richard Sutton on LLM dev: a) core problem is LLMs can't learn from use. Diff architecture necessary for continual learning b) if you've got continual learning then current big-bang training no longer useful. facit: LLM approach not sustainable and dead end.
Tags
Annotators
URL
-
-
-
Stuff to unpack here, and read the three posts by [[Valdis Krebs]]. [[Stephen Downes]] draws analogy between social and neuronal networks, and suggests same underlying logic. This I think points back to my notion that fully embracing the network metaphor/thinking in tech hasn't happened and would be a valuable path forward out of current upheaval. Social networks etc have a certain symmetry that internet platforms ignore.
-
-
mistys-internet.website mistys-internet.website
-
A July 2024 posting (via Alper Çugun on Masto) signalling a decline in functional quality in GitHub. In his Masto message Alper anecdotally says "many leading edge developers have already moved off or are seriously considering it."
-
- Sep 2025
-
www.nrc.nl www.nrc.nl
-
Former UKIP MEP Nathan GIll admits to being bribed by Russia to make pro-Russian / anti-Ukrainian statements. The person carrying the money, Oleh Volosjyn also has ties to Dutch extreme right FvD party. Dutch secret service previously stated knowing about at least one Dutch politican having been bribed. Possibly De Graaff, then a MEP for FvD
-
-
www.theatlantic.com www.theatlantic.com
-
Main points: - Russia does not care about number of people lost, but cares about losing revenue. - Russia is a (imperialist) petrol-state, so Ukraine hits revenue there. - Ukraine's weapon production is increasingly independent from both US and Chinese production - Their drone production is fully home grown and has tight feedback loops to experiences in the field (home grown meaning that they also create their own pcb's from scratch etc.) - EU is funding / supporting that - Trump is words only, and is letting Biden admin sanctions run out without renewing them.
-