Quin tipus de tasques mecàniques?
Jo encara estic peix amb la IA però faig servir Public AI, que és de programari lliure i sense ànim de lucre. La faig servir per a consultes que em duria més temps a través del cercador.
#PublicAI
RE: https://social.coop/@Wtebbens/115860069680612667
#publicai doet prachtige dingen, onder andere:
⁃ alles #opensource
⁃ transparante system prompts
⁃ privacy-first
⁃ publiek beheerd, zoals een nutsbedrijf
zie: https://publicai.co/stories/utility
En ze bieden een chat-interface: https://publicai.co/chat
Bedankt voor de tip, @Wtebbens!
#WorkAdventure 👾 #Apertus 🗻 a quiet chalet on the top of a virtual mountain ❄️ Join me here to explore #PublicAI or just say hello and chat about life, the universe, and whatever's on your mind at https://ciao.spiu.ch
Public AI, Built On Open Source, Is The Way Forward In The EU
Maybe this is a symptom of a broader issue in AI: imagining it as a tool to substitute existing processes rather than something that can extend our capabilities to do things we never were able to before…
As we think about responsible #CivicAI and even #PublicAI, I think it would be really valuable to ground ourselves in concrete examples of low risk applications, questions of missing capabilities over substitutes, and priorities the public actually WANTS to implement AI to address.
Why “public AI”, built on open source software, is the way forward for the EU
A quarter of a century ago, I wrote a book called “Rebel Code”. It was the first – and is still the only – detailed history of the origins and rise of free software and open source, based on interviews with the gifted and generous hackers who took part. Back then, it was clear that open source represented a powerful alternative to the traditional proprietary approach to software […]
#ai #aiAct #cdsm #china #cloudComputing #copyrightDirective #dataMining #eu #freeSoftware #openSource #openSourceInitiative #paulKeller #publicAi #rebelCode #research #startups #supercomputers #tdm #textMining #us #ventureCapital
https://walledculture.org/why-public-ai-built-on-open-source-software-is-the-way-forward-for-the-eu/
My input into the AI track at #DINAconCH covered some history and shared an outlook for the civil sector, in particular the #PublicAI movement as a way to get behind sovereign ML/AI projects, and build on the foundations of #PublicCode
My slides - and ideas for how to get started - are here: https://hacknight.dinacon.ch/project/103
Closing remarks, with allusions to trade policy and #EuroStack:
@pluralistic: “I think we’re in the in the first days of the post-American Internet. It may not be a good Internet. But we could make it a good Internet.”
Lina Khan: “We’re going to have to reckon with the question of whether cloud computing should be treated as core infrastructure.”
Cory and Lina for #PublicAI? 👀
He programat un bot que connecta amb Apertus: @aibot
Li podem preguntar o consultar el que vulguem.
El codi font del bot l'he pujat al meu repositori https://codeberg.org/spla/aibot.
#IA #Apertus #PublicAI https://mastodont.cat/@spla/115405903970475086
Heading back to #NYC today for #UNGA! Excited to connect with folks working on #FoodSecurity to talk about defining a new research agenda for #AI and what it means to build truly accountable #PublicAI infrastructure ✨
https://www.aspendigital.org/feeding-the-future/
@loleg Other tools/services like ChatGPT and Copilot seem to give a fair summary for the exact same question. I know they crawl/sponge what they can get a hold of. I don't know what data source #PublicAI #Apertus use. I'm also on Wikidata (although some outdated) if that matters. But as you can see from the input question, the output is really a hallucination of some sort.
The evaluation criteria of the Swiss {ai} Weeks hackathons was shared with our participants on the platform, where they have documented their results. After the presentations, the #Apertus model powered by #PublicAI was asked to apply the evaluation criteria to each individual project, from which the AI scoring was derived. There is an uncanny overlap in the distribution with the human votes, which were gathered completely independently of the AI evaluation. "Local produce transportation" and "Tibetan chatbot" tied for first place.
#PublicAI now hosts #Apertus across multiple data centres: with a little luck, your prompts are processed at CSCS in Lugano #sovereigncompute
https://publicai.co/about
Public AI inference utility, nonprofit, open-source project.
https://raw.githubusercontent.com/swiss-ai/apertus-tech-report/main/Apertus_Tech_Report.pdf
Big News! The completely #opensource #LLM #Apertus 🇨🇭 has been released today:
📰 https://www.swisscom.ch/en/about/news/2025/09/02-apertus.html
🤝 The model supports over 1000 languages [EDIT: an earlier version claimed over 1800] and respects opt-out consent of data owners.
▶ This is great for #publicAI and #transparentAI. If you want to test it for yourself, head over to: https://publicai.co/
🤗 And if you want to download weights, datasets & FULL TRAINING DETAILS, you can find them here:
https://huggingface.co/collections/swiss-ai/apertus-llm-68b699e65415c231ace3b059
🔧 Tech report: https://huggingface.co/swiss-ai/Apertus-70B-2509/blob/main/Apertus_Tech_Report.pdf
After #Teuken7b and #Olmo2, Apertus is the next big jump in capabilities and performance of #FOSS #LLMs, while also improving #epistemicresilience and #epistemicautonomy with its multilingual approach.
I believe that especially for sensitive areas like #education, #healthcare, or #academia, there is no alternative to fully open #AI models. Everybody should start building upon them and improving them.
#KIMündigkeit #SovereignAI #FOSS #ethicalAI #swissai #LernenmitKI
@stefanwagner @tu_muenchen Richtig. Gleichzeitig wartet die KI-Szene da noch auf die ersten Resultate mit mehr Impact. International wird - wenn überhaupt - Black Forest Labs mit Flux gesehen. Aleph-Alpha haben sich in die 3. Liga verabschiedet. Vielleicht dreht sich ja der Diskurs etwas, wenn die ETH bald #Apertus veröffentlicht. Dann könnten bspw. Deutschland und die EU stärker auf KI als öffentliches Gut setzen und Transparenz und Provenienz von Daten gegenüber Systemen aus USA und China zum Alleinstellungsmerkmal machen. #PublicAI
nicht mit der ausreichenden Fachexpertise.
Das grundlegendere Problem ist aber die illusorische Perspektive der Politik, Deutschland könnte - ohne strategische staatliche Steuerung wie in China, ohne eine Venture-Capital-Kultur wie in USA, und sogar ohne einen entsprechenden Pool an Expert:innen "in der ersten Liga mitspielen". Statt solcher Fantasien sollte man sich hierzulande auf vielversprechende Nischen oder auf unausgelotete neue Wege konzentrieren.
Nimmt man dagegen die geopolitsche Bedeutung von #KI ernst, muss man *jetzt* auf EU-Ebene konzertiert agieren und Mittel, Forschungsansätze, Human Resources, Know-How so teilen, dass man ein Maximum herausholt.
Die Schweiz ist da gewohnt furchtlos und mit #Mistral in Paris gäbe es zumindest eine gute Ausgangsbasis. Denn es geht bei den politischen Implikationen von #LLM s zunehmend um #Werte - gerade im Bereich Bildung. Hier zwischen China und USA ganz auf seine eigene Stimme zu verzichten, könnte Europa noch bereuen.
When thinking about “sovereign” AI consider the difference between digital sovereignty and digital nationalism. Take the Swiss model (https://www.forbes.com/sites/corneliawalther/2025/07/14/swiss-ai-for-public-good-a-prosocial-ai-blueprint-for-the-world/) vs the Korean (https://www.cnbc.com/2025/08/08/south-korea-to-launch-national-ai-model-in-race-with-us-and-china.html). Listen to @ceciliarikap.bsky.social and think about our approach in Canada as we pursue #publicAI.
When thinking about “sovereign” AI consider the difference between digital sovereignty and digital nationalism. Take the Swiss model vs the Korean. Listen to @ceciliarikap.bsky.social and think about our approach in Canada as we pursue #publicAI.
Client Info
Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst