Skip Navigation

Posts
125
Comments
674
Joined
3 yr. ago

  • The useful ones are still provided by big companies because the rest of us can’t afford the hardware to train them.

    We have computing power in our pockets a million times more powerful than we used to send man to the moon, why do you think we'll never have enough power?

    I have already pointed out https://eurollm.io/

    The EuroLLM project includes Instituto Superior Técnico, the University of Edinburgh, Instituto de Telecomunicações, Université Paris-Saclay, Unbabel, Sorbonne University, Naver Labs, and the University of Amsterdam. Together they created EuroLLM-22B, a multilingual AI model supporting all 24 official EU languages. Developed with support from Horizon Europe, the European Research Council, and EuroHPC, this open-source LLM aims to enhance Europe’s digital sovereignty and foster AI innovation. Trained on the MareNostrum 5 supercomputer, EuroLLM outperforms similar-sized models. It is fully open source and available via Hugging Face.

    So long as someone doesn't want to rely on big tech there will be people pushing for independence just like Linux users such as myself

  • Do you build your own Linux from scratch? If so why would you assume you can build an LLM from scratch?

  • Every corporation and government are blindly sticking with Microsoft

    Are you sure?

    France to remove Windows from government computers in sovereignty push

    https://www.rfi.fr/en/france/20260417-france-to-remove-windows-from-government-computers-in-sovereignty-push

    https://tuta.com/blog/countries-ditching-microsoft-choosing-linux-digital-sovereignty

    I doubt it will happen with governments

    It does not take much for things to change, you might like this:

    We've Hit A Wall With Transport. Here's Why | Black Swans 3 | If You're Listening

    https://youtu.be/o1R6Aq19A6Y?t=1281

  • For which you still need massive amounts of memory and compute to run reliably

    2026's average gaming PC is massive amounts of memory and compute apparently

    The gap will take decades to close, if it ever does.

    lol there are plenty of open source models in the top 100 with multiple SOTA models released in the last few months alone

    There's also smaller LLM's being made like https://eurollm.io/ which excel in their own ways

    That, and the fact that chatbots and agents nowadays rely on all sorts of proprietary customizations

    Funny that just came up: https://discourse.ubuntu.com/t/the-future-of-ai-in-ubuntu/81130?=0

    Previously, to benefit from the full power of LLMs, you had to skew to higher parameter models. Recent developments in models like Gemma 4 and Qwen-3.6-35B-A3B demonstrate advanced capabilities such as tool-calling which enable LLMs to search the web, interact with external APIs and file systems, troubleshoot live systems and fundamentally reason about topics that lie outside of their initial training data.

    The gap will take decades to close, if it ever does.

    😁

  • what a neat little site

  • so the irony is that you guys are hallucinating an answer confidently, AI is really good at translation and language

    https://www.deepl.com/en for example is far better than google translate was because they leapt onto ML/AI earlier

    If you don't want to straight up use Mistral/EuroLLM which are trained on European languages:

    The company, which launched two years ago, released on Tuesday Mistral Large 3, which the startup claims maintains the same level of performance in “a large number of languages,” particularly European ones.

    “Most AI labs focus on their native language, but Mistral Large 3 was trained on a wide variety of languages, making advanced AI useful for billions who speak different native languages,” the company said in a press release.

    https://www.euronews.com/next/2025/12/02/mistral-europes-ai-champion-releases-new-smaller-frontier-models-heres-what-to-know

    If you don't want to directly use a chatbot there's also https://morpheem.org/ and https://languatalk.com/try-langua which uses AI but builds it around the games specifically, vs just straight up telling the AI bot what to do

  • are you able to point some out on a1/b2 content on mistral for me? im interested to see if you’ve found some

  • tbh if you just use the free version of duolingo i don’t think you can do many lessons now, the energy runs out after like 2-3 lessons max

    im learning french so i use it just as a daily reminder to do something in french but outside of it, there should be loads of music and podcasts and audiobooks on spotify that you can use, i can’t be specific but there is things like news in slow french that surely must exist for spanish

    if you have a decent video card with 8gb vram can look into lm studio with eurollm and run your own chatbot locally

    https://eurollm.io/

    can say things like, i am an a1 spanish learner, give me a word in English to translate to spanish and tell me if i get it right or wrong

    give me a sentence in spanish to translate into english etc

    mistral.ai is your best bet if you want something online, they are a french company but do a lot with european languages

    hellotalk is good if you want to listen in on people learning to speak but the quality of audio depends on the persons mic quality so might be better for later in learning journey

    there’s tons of stuff, best is usually just to seek out books/music/sites and do your best to spend as much of your day learning, the more time you do the faster you learn but it is tough! good luck!

  • so all the reddit threads had nothing of value then i started reading the kiwifarms forum post then i realised i don’t give a shit

  • i only found recipe results for pizza cake and i have no interest in reaching out to her on whatever other social media platform im not on

  • it’s well documented?

  • i thought it was non-addictive

  • At the moment they have no reason to doubt that

    Mistral’s revenues soar over $400mn as Europe seeks AI independence

    https://www.ft.com/content/664249e7-e8d5-4425-b397-ad3ed590b305?syn-25a6b1a6=1

    And Mistral appears to be 6 months to a year behind Claude which in the AI race is a fucking eternity

    Anthropic's annualized revenue run rate has reached $14 billion, according to the company's own announcement. The company says this figure has grown more than 10x annually for the past three years. That kind of growth from a company earning the bulk of its revenue from business customers, not consumer subscriptions, is unusual.

    The enterprise numbers tell a more specific story. Over 500 customers now spend at least $1 million per year on Claude, up from a dozen just two years ago. Customers spending more than $100,000 annually have grown sevenfold in the past 12 months. And eight of the Fortune 10 are now Claude customers. Axios noted that 1 in 5 businesses using Ramp now pay for Anthropic, up from 1 in 25 a year ago, and about 79% of OpenAI's paying users also pay for Anthropic, suggesting the market isn't zero-sum.

    https://www.forbes.com/sites/jonmarkman/2026/02/13/anthropic-the-380-billion-powerhouse-hiding-in-plain-sight/

    The numbers are staggering

    At my workplace which is extremely tepid and slow moving, they initially rolled out Microsoft Copilot much to the hate of everyone who knows about AI

    The developers have been demanding Claude and they're doing a very slow roll out of it which is amazing in itself, who would have thought they could roll out anything that isn't Microsuck Dogshit Premium Pro 365

  • what do you mean? at the moment Anthropic is profitable and desperate for more compute, they are crashing all the time from too much demand and have had to take away so many features from their plans due to crashing their servers from heavy demand:

    After speculation and frustration spread, Anthropic’s head of growth, Amol Avasare, took to social media to clarify that this was a “small test on ~2% of new prosumer signups.” As for the reasoning, he explained:

    When we launched Max a year ago, it didn’t include Claude Code, Cowork didn’t exist, and agents that run for hours weren’t a thing. Max was designed for heavy chat usage, that’s it.

    Since then, we bundled Claude Code into Max and it took off after Opus 4. Cowork landed.

    Long-running async agents are now everyday workflows. The way people actually use a Claude subscription has changed fundamentally. Engagement per subscriber is way up.

    We’ve made small adjustments along the way (weekly caps, tighter limits at peak), but usage has changed a lot and our current plans weren’t built for this.

    So we’re looking at different options to keep delivering a great experience for users.

  • This is like saying, you use google? don't you know it's bad for you

    it convinces no one

  • Oh yeah I'm not sure how niche what you were looking for was, for me my stuff is mostly basic, eg I have a massive text file full of subnets from a routing table, I want a command to remove everything but the subnets, i want to convert png to jpg, i want to convert webp to jpg, how do i search a specific value within 10000 log lines of json etc

    just stuff that I could search ecosia -> read up on how X works -> guestimate what i'm looking for with trial and error -> get result or ask claude, run command, it works

  • I had to bring my dog to a chiropractor

    ?