Adrian Gschwend

Finally on fediverse! Join me on @ktk@social.netlabs.org

📊 Michael Rauch delivered an insightful session on Creating RDF Knowledge Graphs from Structured Data, demonstrating how mapping languages transform structured data into powerful knowledge graphs. #CDL25 Joining us tomorrow? — secure your place here 👉 bit.ly/47TN7ks #CDL25 #SemanticWeb #RDF #KnowledgeGraphs #LinkedData

Super proud to share a new QLever milestone. We loaded one trillion triples on a single AMD Ryzen 9 machine with 16 cores and 128 GB RAM 🚀 Index size is 6.5TB and the dataset runs even with less memory. Dataset: qlever.dev/one-trillion/ #RDF #SPARQL #KnowledgeGraphs

@danbri that was funny :)

Huge if true 📈

@kidehen another way of looking at it is that they realizie their data lake layers are pretty much useless for LLMs and KGs get much better results back to the LLM so they try to re-claim/gain some of the discussions into the datalake approach.

A very important announcement was made yesterday by key Business Intelligence (BI) and Data Lakehouse/Warehouse-oriented DBMS vendors regarding an Open Semantic layer Initiative (OSI) designed to provide data harmonization in a way that enhances AI Agent context while breaking down the data silos created by applications developed, deployed, and sold by the new consortiums membership. First off, that’s absolutely fantastic news! Naturally, there will be an inevitable “Semantic Layer” vs. “Semantic Web” showdown (primarily in marketing communications)—for better or worse. In anticipation of this, I decided to try out @Grok to gain insights (gleaned from its training content) about these initiatives, as showcased in the session linked in the comments section. Enjoy!

@stanleywidianto I still laugh every time I walk by ;)

@Ugugu @FelixSchneuwly Lesenswert: inside-it.ch/e-id-die-stand…

@linkedktk @FelixSchneuwly Wäre der Moment das gesetzlich festzulegen. Grundsätzlich keine Ausweispflicht für Onlinedienste, Ausnahmen sind begründungspflichtig.

Tendiere aktuell zu e-ID zurück an den Absender...

My work here is truly done. Nobody with intellectual integrity can still believe that pure scaling will get us to AGI. GPT-5 may be a moderate quantitative improvement (and it may be cheaper) but it still fails in all the same qualitative ways as its predecessors, on chess, on reasoning, in vision; even sometimes on counting and basic math. Hallucinations linger. Dozens of shots on goal (Grok, Claude, Gemini) etc have invariably faced the same problems. That’s exactly what it means to hit a wall, and exactly the set of obstacles I described in 2022. No hypothesis has ever been given more benefit of the doubt, nor more funding. Pure scaling simply isn’t the path to AGI. Attention was not all we needed. All I am saying, is give neurosymbolic AI with explicit world models a chance. 🫳 🎤

@Ugugu @FelixSchneuwly Was meinst Du mit unklar definiert? Wenn wir das nicht vom Staat kriegen wird die gleiche Funktionalität über private kommen. Weil ohne gehen gewisse Services einfach nicht. Und genau die privaten wollten wir ja alle nicht.

@FelixSchneuwly Unklar definierte Anwendungskriterien. Mit jeder Ausweitung auf Private steigt Missbrauchspotenzial. Bin nicht grundsätzlich gegen eine eID, möchte aber vorher genau wissen worauf ich mich einlasse.

@sandhawke that is a good idea

How close are we to "tweet" being a generic word for small post, free of legal encumbrance? If X really wants to move away from that branding, they should make it totally safe for others to use the term.

@Leonidas666_ @GerryS Uhm I did not say anything even related to that. Not sure why you bring that up. All I did was posting a video I find interesting, sorry if that offends you.

@linkedktk @GerryS Cannot compare the rescue methods in the snow and volcanic areas. In Indonesia, there is an iceberg, Puncak Jayawijaya, Papua. And also, the geographical condition between the In Rinjani whose plains are not solid rocks. but the sand is easy to landslide.

Tidak semua helicopter rescue itu se-ideal di film-film. Ini lagi ramai kasus pendaki wanita asal Brasil 🇧🇷 yang terjatuh di gunung Rinjani. Banyak yang nanya "kenapa kok gak pake helicopter untuk rescue?" Ya, memang, BASARNAS punya helicopter untuk rescue operations... cuman... 1. Kejadian lokasi ada di ketinggian 10,000 kaki, dimana korban jatuh di lereng ke ketinggian sekitar 9400ft. 2. Mau terbang mountain rescue benar-benar tergantung dengan cuaca dan KEMAMPUAN HELICOPTERnya. 3. Helicopter yang dioperasikan BASARNAS ada: AW139 dan AS365. Yang AW139 ada 1, sisanya AS365. 4. Mau hoisting, helicopter harus hover. Untuk hover ada definisi kategori hover. Hover In Ground Effect (IGE) dan Hover Out of Ground Effect (OGE). Hover IGE itu kalau helicopter dibantu oleh bantalan tekanan udara tinggi dari baling2nya. Ini IGE cuman sampai sekitar 10 meter - 15 meter diatas permukaan tanah datar. (Jelas lokasi korban bukan di tanah datar). Nah, kalau tidak hover IGE, ya harus hover OGE. Disinilah kita ketemu masalahnya... 📸AS365 US Coast Guard 5. Untuk helicopter AW139, ketinggian maksimum untuk hover OGE adalah 8130 kaki diatas permukaan laut. Untuk AS365, hover OGE maksimum bisa dilakukan di 3740ft. Jadi disini bisa kelihatan, heli BASARNAS tidak akan bisa melakukan hoisting rescue korban, mau cuacanya bagus sekalipun. 📸AW139 Untuk perbandingan, Blackhawk aja hover OGE max di 6200ft. 📸B206-L4 7. Kok, ujung2nya di evakuasi pake helicopter dari Sembalun? Iya lah, Sembalun ketinggiannya cuman 3000ft... apalagi pake Bell 206L4 sipil yang hover OGEnya bisa sampai 6500ft... Cuman tetep, gak bisa rescue di lerengnya pakai heli ini juga. Semoga jelas dan bermanfaat... 🙏

@Rockoonde @GerryS I never said it's Indonesia.

@linkedktk @GerryS But this indonesia not nepal

@GerryS this is the story about how they created the rescue operation in Himalaya (around 2010) youtube.com/watch?v=Wa06CB…

@GerryS You might enjoy the documentary about Air Zermatt in Switzerland, which does alpine rescue in Switzerland. They also trained the team in Nepal. See for example youtube.com/watch?v=xjG9aP… (enable auto-generated subtitles)

@cyberandy Interesting slides thank you. What did it say why on the second slide? "Mostly due to ..."?

Why it is still so hard to build innovative businesses in Europe 🇪🇺

@muwlgr @Gammitin We asked for that formally some years back, got declined.

@Gammitin IBM should release OS/2 as open source. They don't get any profit from it since long ago, anyway.

This is my vibe.....

Our paper is finally published. It's an honor to work together with all, both from @univ_indonesia and WU Vienna.

@nuicemedia This. An absolute failure of governance that the first requirement for developing North Jakarta wasn't to build a metro and dedicated bus lanes before anything else.

Plenty enough responses say that Ancol has basically failed to keep up with the times, failed to provide adequate public transport access, facilities and accessibility are not up to acceptable standards, and of course, people don’t have money like they used to.

Ancol is struggling to meet its goal of 660,000 visitors during this holiday period, having only admitted around 315,000 or not even half the targeted number. Management cited rain and longer holiday period as reasons for the weak demand to enter the beach park.

@chtrachsel Jep, und Deine Daten :-D

@linkedktk Sehr schön. Sehr beeindruckend

How close is every building 🏠 in Switzerland to the nearest public transport stop? 🚆 With QLever and some clean RDF data, this is easy. Magdalena and I show it in our new demo, crazy fast, super visual. 🎥 youtu.be/gnocVIoHVKk?si… #SpatialAnalytics #KnowledgeGraphs #OpenData

@Ugugu Er dreht immer mehr im roten leider...

Kann man machen, einen Rechtsstaat mit einer Quasidiktatur vergleichen, sollte daraus aber nicht die falschen Schlüsse ziehen. (Für mich als Linksliberalen eine der grösseren Desillusionen der letzten Jahre...)