ℍ𝕂-𝟞𝟝

  • 1 Post
  • 509 Comments
Joined 2 years ago
cake
Cake day: July 14th, 2024

help-circle
  • Back in the days, vacation was an affair organised by workplaces. State companies (all companies were state companies) owned vacation venues, and workers were entitled to an organised trip there, so that’s why you have a bunch of people who don’t really know each other on the same vacation at the same time.

    Ismerkedési est means essentially a soirée organised in order for people to “get to know each other”, but in Hungarian that term has a strong connotation that it’s for romantic purposes.

    So the situation is essentially an organised vacation party where people are supposed to find romantic partners.









  • So some context. In communist era Hungary, a lot of the population lived - and still lives - in big prefab apartment buildings that run off central heating.

    In the winter as it was getting colder, the heating would get turned on at the same time for the whole district if not the city.

    The “they” in “they are finally heating at Jucika” is the government. It’s a hot topic every year when it gets turned on, as the people pay a flat rate for warmth.








  • The sticks were digitally driven for quite a while as well. Managed to short circuit a 2004 VW Passat’s electronics, the sticks started to point in random directions all over the place while the car was in motion. And I don’t mean shaking, just going to a random value and settling on it every 1.5 seconds.

    Was very spooky.



  • I actually did an experiment on doing just that. For context, I’m an experienced software engineer, whose company buys him a tom of Claude usage so I had time to test out what it can actually do and I feel like I’m capable of judging where it’s good and where it falls short at.

    How Claude Code works is that there are actually multiple models involved, one for doign the coding, one “reasoning” model to keep the chain of thought and the context going, and a bunch of small specialized ones for odd jobs around the thing.

    The thing that doesn’t work yet is that the big reasoning model has to still be big, otherwise it will hallucinate frequently enough to break the workflow. If you could get one of the big models to run locally, you’d be there. However, with recent advances in quantization and MoE models, it’s actually getting nearer fast enough that I would expect it to be generally available in a year or two.

    Today the best I could do was a tool that could take 150 gigs of RAM, 24 gigs of VRAM and AMD’s top of the line card to take 30 minutes what takes Claude Code 1-2. But surprisingly, the output of the model was not bad at all.