bloomingkales

43 ms·

1. 571.▲
   by bloomingkales 2y ago

   Mobile scrolling breaks for me too. Off topic for OP, I kinda want to use this UI to browse HN or a lot of sites.
2. 572.▲
   by bloomingkales 2y ago

   I would say ELO doesn’t let you walk out of a team based game with anything other than a win or loss. So, in sports, even if your team losses there will be players who will leave the season as All stars, or even MVPs (or vice versa, you are
3. 573.▲
   by bloomingkales 2y ago

   I doubt websites as we know it will be what we’ll be dealing with going forward anyways. What is a browser if we just digest all the HTML and spit out clean text in the long run? We handed over something of some value I guess, once upon a t
4. 574.▲
   by bloomingkales 2y ago

   Can a browser run a web server with this?
5. 575.▲
   by bloomingkales 2y ago

   I don’t tell the model that it is 100% wrong, in which case it would contradict the first half of its response with the second half of its generation. We basically want it to enter double-checking mode on it’s own from its initial rough dra
6. 576.▲
   by bloomingkales 2y ago

   making multiple calls and feeding the responses back to the model. By asking it to reconsider half its generated response, aren’t I essentially asking it to formulate the second half of its response from the first half internally? I’m byp
7. 577.▲
   by bloomingkales 2y ago

   I don’t think this is for humans to ever write. It’s mostly readable code for AI to generate, Lego blocks to achieve a task. That’s my best guess.
8. 578.▲
   by bloomingkales 2y ago

   The gist of it is I think these large hosted models have system prompts that are not as skeptical of its own outputs. You are an helpful AI Assistant seems to lead to more lax responses. Adjusting the system prompt to be more incredulous
9. 579.▲
   by bloomingkales 2y ago

   There is no rigor to this, this is just from throwing stuff against the wall. See my response to the other poster above.
10. 580.▲
    by bloomingkales 2y ago

    Your prompt paints a context that might keep it more on the rails, but it won't do multiple passes. This is probably the truth behind the black magic I’m imagining. You could have it explicitly spit out this process, in which case yo
11. 581.▲
    by bloomingkales 2y ago

    Honestly, try prompting it with “you are wrong 80% of the time, therefore you will need to double check your answers, first factually, then numerically, then double check the time/date. You are still probably wrong so do a third accura
12. 582.▲
    by bloomingkales 2y ago

    Qwen2.5 3b is very very good.
13. 583.▲
    by bloomingkales 2y ago

    AAA games need to do honest pricing. They can release at $70 dollars but if it’s not a hit, drop the price and release modding tools so the community can have a go at reinventing it. What they seem to do is drop the price and give it anothe
14. 584.▲
    by bloomingkales 2y ago

    Anyone have spare machines and want to one v. one my computer-use AI? We just tell it to hack each other’s computers and see how it goes.
15. 585.▲
    by bloomingkales 2y ago

    So long as we make the launch nuke methods private, we should be okay I think. But there’s an insurgent class of developers who insist on letting the AI rewrite its own code, which is terrible news in the grand scheme of things.
16. 586.▲
    by bloomingkales 2y ago

    It is inevitable. Someone please just make the Matrix repo so we can all begin contributing, enough the with the charades.
17. 587.▲
    by bloomingkales 2y ago

    I say it because at any point either of those two things could have shown up at your front door (as it was happening). In this case, yeah, CNN may give you AI synthetic news sooner rather than later to preempt the inevitable. I’m not really
18. 588.▲
    by bloomingkales 2y ago

    What in god’s name are you talking about? https://www.congress.gov/bill/118th-congress/house-bill/1157... Oh. Had to look it up. (6) to expose misinformation and disinformation of the Chinese Communist Party
19. 589.▲
    by bloomingkales 2y ago

    Depends. There’s no such thing as too much news during ww2 or Industrial Revolution. This is news about how news as we know it is going away, and there’s more news on that coming 24/7. And that’s just the news on the news. Idea for OP:
20. 590.▲
    by bloomingkales 2y ago

    Something like this: https://xmpp.org/extensions/xep-0030.html That way AI can handle any arbitrary app so long as it has a public interface of discoverable commands. It will make sense of your apps and handle complex
21. 591.▲
    by bloomingkales 2y ago

    I was thinking about something like this. It would be cool if each app has a discovery service of commands so AI can manage the ecosystem automatically.
22. 592.▲
    by bloomingkales 2y ago

    Guess we need an Agent that logs and re-contributes to Stackoverflow (for example) automatically. Then also have agents that automatically give upvotes for used solutions. Weird world. I’m just imagining the precogs talking to each other in