Только недавно мысленно грустил о том, что давно не выходило значительных открытых LLM, которые можно селфхостить как подвезли Ministral 3.

Надо будет попробовать её в Home Assistant. На первый взгляд выглядит неплохо: даже 3B модель умеет в tool calling и даже такая маленькая версия прилично […]

All LLM/AI code tools should be banned for junior/mid level developers in a company.

You're regressing developers in your team by being junior prompting assistants rather than senior developers

#llm #AI #coding

⁨📷⁩ ⁨Crime-based CAPTCHAs 😏

https://peterwhiting.me/alignment_captchas/⁩ #ai #aialignment #llm #captcha

How is AI really impacting jobs? @brianmerchant writes for Blood in the Machine about what the numbers are saying: "Generative AI is not reliable enough when it comes executing complex tasks to enable most organizations to displace jobs at scale, and it certainly can’t do jobs that require […]

Math eint mathing with #LLM $IBM CEO says that at today’s costs it takes about $80B to build a 1 GW AI data center, so the ~100 GW of announced capacity implies roughly $8T of capex & “no way you’re going to get a return on that,” since you’d need “about $800B of profit just to pay for the interest”

Pourquoi #OpenAI décrète l’« #alerte #rouge » face à #Google #Gemini

www.silicon.fr/business-136...

AI macht uns unglücklich. #ai #openai #gemini #blog https://trancefish.de/2025/12/02/ai-macht-uns-unglucklich

#rwe hat am Rand des Tagebaus #hambach eine neue #PV-Anlage mit 13,9 MW Spitzenleistung in Betrieb genommen.

Auf zwölf Hektar erzeugen über 22.000 Module Strom für rund 4.500 Haushalte. Zum Einsatz kam erstmals ein KI-gestütztes #Bauüberwachungssystem. Es ist der dritte #Solarpark an diesem […]

“Even when the generated video isn’t recognizably lifting from someone else’s work, that’s what it’s doing. There’s literally nothing else there.”

https://www.404media.co/openai-cant-fix-soras-copyright-infringement-problem-because-it-was-built-with-stolen-content/

College Students Choosing A.I. Majors Over Computer Science. Via @nytimes #AI #ArtificialIntelligence 💻 🤖 🧠 #Education

@theinformation.com

#openai CEO Declares ‘Code Red’ to Combat Threats to #chatgpt, Delays Ads Effort
By Stephanie Palazzolo and Erin Woo

#AI 12/2/2025

https://www.theinformation.com/articles/openai-ceo-declares-code-red-combat-threats-chatgpt-delays-ads-effort

[OT] Sam Altman is suddenly terrified futurism.com/artificial-i...

WSJ: #OpenAI declares 'Code Red' as #Google threatens #AI lead archive.ph/unKRQ

OT] WSJ: <a href="https://mas.to/tags/Google" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">#OpenAI declares 'Code Red' as <a href="https://mas.to/tags/AI" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">#Google threatens <a href="https://archive.ph/unKRQ" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">#AI lead [https://archive.ph/unKRQ

Language translation is an actually useful application of AI

#ai #multilingual #translation #languages #artificialintelligence

https://newslttrs.com/language-translation-an-actually-useful-ai/

I would like to invite you to read my article ‘Self-Awareness in Artificial Intelligence’, as it is important to reflect on this subject…

#SelfAwareness #ArtificialIntelligence #AI #consciousness #human #machine #mind #Superintelligence #AGI #Singularity

I've been playing around with LLMs and all this MCP stuff. I created an agentic workflow that connects @planifyapp, evolution and so on. It's pretty rad. It improved my productivity.

But, I find LLMs to be very fragile. If you change the LLM or update the LLM the flows break. I don't know how […]

Just read about DeepSeek Sparse Attention.
Essentially, they gave ADHD to an AI to make it work better, and market it as a big innovation.

This would be a chance for Neurotypicals to question their attitude towards Neurodiversity, but I doubt they take the hint.

#adhd #neurodiversity […]

An independent effort says AI is the secret to topple 2-party power in the U.S. Congress.

From @npr: "The goal is to elect a handful of independent candidates to the House of Representatives in 2026, using AI to identify districts where independents could succeed and uncover diamond in the […]

So you're saying...I must never restart Teams, ever again. Tricky.

(I have to use Teams for a client.)

#ai #llm

i am sitting in a coworking next to some AI startup bros who apparently try do develop an #llm based application. their ramblings go sth like this
- "why does it do this"
- "idk"
- "can we fix it?"
- "no idea"
- "could we add (some natural language about how to fix that specific case)"

lost […]