@tappylappy@T00ManyCommies do you think having the highest retirement age and 15% of its GDP going to pensions has something to do with the lack of funding in education, which is precisely what they're protesting for? it's almost like it's a systemic issue that isn't going to be fixed until you address it
anyone ever use hermes kanban cards to make a plugin that instead of calling up a model to deploy to a task, you could deploy a model + harness underneath harness to a task, and for that harness to report back to hermes? kinda interesting idea i was playing with
@landontgreen 45 is the 1017 vs 42 for glm 5.3 flash on the ai analysis index. haven't tried 1017 yet, downloading now. i've liked 5.3 a lot though, same with qwen flash, we'll see!
@joaosump@lxe holy crap! i am downloading 1017 right now, was trying tout the Maya S quant, 96gb of ram and 5090, can you point me at the project you're running? 50-60 feels very usable if we can hit a 45 on the ai analysis index
@BrandonLBradfor and yet it pushes you towards a president that is explicitly campaigning on anti immigration. HMMM I WONDER WHAT COULD BE THE CONNECTION lmao.
@Teknium@iamlukethedev i wonder if you could make it so there was like a 1 click install "dev studio" that had a suite of plugins for asset generation. would love a bundle that linked all the ones you listed
@vLeocadiio@seanhighness i pointed the ai to the link to the quant on hf, and then i linked this https://t.co/2HzHXUg9G1 and told it to get strata to run the quant. it figured out the rest. i used glm 5.3 flash to do it, which is free to use on opencode
@pleometric this reminds me of old videogames. world of warcraft, for example holds up very well over time since it eschewed realism for a distinctive style, realistic games at the time are long gone but wow stays. new models need a distinctive hook, a deliberate imperfection as signature