bogdiyan

I gave it a second thought and I think terms should be "self-managed" - you manage it either locally or in cloud and "self-hosted" you host it locally. Idk "self-contained" doesn't click with me, to be fair.

It is confusing, I agree. IMO we should have "self-running" and "self-hosting". Self-running if I provide the hardware, self-hosting if I manage it. Or swap the term, still not sure but there is a difference between you are the one maintaining it and where does it run.

I read the comments here and I am really surprised by many people. One picture is worth a thousand words so with better UI you can make much better user experiences. This unlocks personal assistants to be better adopted by elderly or disabled people. I see so many benefits of it and having models which can do this (if they can do it constantly with good quality) is amazing. Much better products - I am really tired of dumb chatbot - if I can do something with one button or view the whole information in one diagram/image, this is amazing.

So you never have to assemble a bike or paint a room. How those are not real world examples? I find their video pretty cool. For people with ADHD or people learning by watching or children this is spot on.

I am proud European but this whole thing of "EU only" and "sovereign X" is just nonsense. There will be no need of all this masquerade if there were more strong companies in EU building world class software. EU has SAP...mainly. EU was fast asleep in 2000 and 2010 when it comes to software, I had big hopes for AI with so many researches in EU focused on AI, Deepmind etc. but the results are the same - US is far ahead. Problem is not people, companies or software itself - problem is structural in EU as a whole - taxes, 27+ jurisdictions, languages, etc. If people and governments in EU do not start to act and do as one there is no hope.

How this is HN related?

Interesting solution for sure, just I do not find it scalable for most of the companies. Why instead of installing new software just do not put all LLM required information in llms.txt. If more information is needed put it in llms-full.txt (example: https://modelcontextprotocol.io/llms-full.txt)

I hope this project works, I see the point in it.

I am not sure how much of the work is done by OpenAI, or whether it is basically a Broadcom chip specifically built for OpenAI models. It is a necessary step, but building a high-performance chip is not easy. Look at companies like Groq, Amazon, and Google.

As a football fan (Sorry Americans) I love this. I think a nice additions will be * Metric system for distance * I do not understand what this X and Y next to distance on each goal means and what it is for. * A short video of the goal itself (this will be really cool).

I wonder what portion of the migration was contributed by Mythos. Surely the Bun team now has access to more powerful models, but could such a migration be done with just Opus 4.7? Nonetheless, nearly 7k commits is impressive.