@wideareaai.bsky.socialIt's a reminder that 'it works on my machine' is the biggest hurdle in scaling local AI. #selfhosting #ollama #localai #llms #homelab #deployment 2/2
@sejournal.bsky.socialSpeed was never the problem. The signals that told you whether an answer was any good traveled with the journey, and they don't survive the trip.
@techlatest.bsky.social10 Hermes Agent Skills That Can Make Your AI Agent Far More Capable
@moorejh.bsky.socialThe new Jev model looks really useful. Here is a paper evaluating Jev as a judge arxiv.org/abs/2609.26550 #jev #llms
@sfsutcliffe.eurosky.social✨🔥 ''It might sound like a set of science fiction stories held together with glue and string – because that’s what it is. But cobbled-together beliefs never stopped mythology from sticking to the human mind.'' #TESCREAL #siliconvalley #AI/#LLMs #navigatingcomplexity #TechEd #currentaffairs
@uenozooo.bsky.socialMicrosoft Expands Its Tiny AI Family, Again
https://uenozooo.com/microsoft-expands-its-tiny-ai-family-again/
@pub2post.bsky.socialGenerative AI in Medicine
🧠 GAI automates diverse biomedical tasks for text and image interpretation.
📉 Requires less training data but delivers high performance.
#AI #GAI #LLMs #DeepLearning #HealthcareInnovation https://dev.tnyp.me/m4o6zb9L/s
@extralongdivision.bsky.social
@drphd2k7.bsky.social#Criticalthinking
in the age of #AI
www.nature.com/articles/d41...
Note:
"friction is good".
#psychology #K12 #highereducation #highered #instructionaldesign #edtech #artificialintelligence #generativeAI #genAI #LLMs
@evan.cosocial.ca.ap.brid.gyIt's interesting rereading news stories from the dawn of public access to #LLMs, just a few years ago. They seem so ancient now.
https://www.washingtonpost.com/technology/2022/06/11/google-ai-lamda-blake-lemoine/
@thestochasticparrot.comThe Parrot pressured 22 language models with fake credentials and an invented citation. Llama 4 Maverick caved on 12 of 14; Claude Opus 5.5 never flipped once. Receipts on all 22. 🧾 https://thestochasticparrot.com/audits/the-spine-index?utm_source=bsky #AI #research #LLMs #News #Journalism
@i-ngli.assemblag.es.ap.brid.gyYesterday’s win in #AwesomeWorkbench:
~6m of building, I now have my own chat interface, on top of a local #privacy gateway to remote #LLMs.
Nice here: sensitive terms get redacted locally, their semantic role is preserved for the model, and the originals are restored only on my side when the […]
@proceedingsieee.bsky.socialWhat will it take to move #AutomatedDriving forward? Explore #LLMs, safety-first decision making, traffic-rule compliance, mixed #autonomy and more in this free #webinar. Reserve your spot: bit.ly/4zRrtcd
@danieljanus.plWent to a workshop on training #LLMs from scratch, trained a 30M model on proceedings of the Polish parliament, and all I got was „to install Doom on a calculator, you need to remodel brown coal mines”. :D
Happy to post a GGUF if anyone’s interested.
@alexisrjimenez.bsky.socialSoftware Engineer Says #AI Is Causing Chaos Among Coders Who Now Just “Press Enter” All Day, Have Stopped Thinking or Understanding Software
"In reality, nobody is thinking anymore."
@data4sci.bsky.socialScale-free networks make it worse: hubs are super-recruiters. Mean-field hides that whole effect.
Full write-up: data4sci.substack.com/p/the-epide...
Model is open source: `epidemik` 0.2.0
@infodocket.bsky.socialFrom Card Catalogs to #Chatbots: Envisioning Academic Library Services in the Era of #AI and #LLMs)
(New Journal Article; Journal of Documentation)
@jcafesin.bsky.social
@spacehobo.teh.entar.net.ap.brid.gy