Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)DH
Posts
0
Comments
216
Joined
2 yr. ago

  • Upsides:

    • side quests other than the radiant ones are mostly cool so far
    • stealth archer isn’t so good that you can just play that way straight out, but the tree makes it looks eventually strong
    • zero g combat in a derelict space station was cool. I hope there’s more of that
    • base building seems fine, I’m not sure what it’s for, but it seems fine

    Downsides:

    • ship stuff feels bad. I don’t care about fast travel, but it’s just about the weakest ship-to-ship combat that I’ve played. Its early yet, though. Boarding a ship was cool at least.
    • combat AI is not good. Enemies never seem to take any initiative, they mostly just crouch behind wherever you found them
    • the setting has no… flavor? The factions feel like fallout analogues but without fun or verve. Maybe I just haven’t found the weird shit yet, but I’m not optimistic.
  • Fallout had a lot of clever environmental storytelling, vignettes in random buildings, drama happening via terminal logs throughout abandoned factories, etc.

    I hope that Starfield hasn’t gotten rid of that aspect of their world building.

  • https://www.understandingai.org/p/large-language-models-explained-with I don’t think you’re intending to be purposefully misleading, but I would recommend checking this article out because the pachinko analogy is not accurate, really. There are several layers of considerations that the model makes when analyzing context to derive meaning. How well these models do with analogies is, I think, a compelling case for the model having, if not “knowledge” of something, at least a good enough analogue to knowledge to be useful.

    Training a model on the way we use language is also training the model on how we think, or at least how we express our thoughts. There’s still a ton of gaps to work on before it’s an AGI, but LLMs are on to what’s looking more and more like the right path to getting there.

  • That’s accurate and doesn’t contradict the person you replied to. What they are saying is that SEO was never about fixing a search engine user’s problem; it exists to solve web host’s problem of “we aren’t getting enough ad revenue.”

    The same is going to happen with these LLMs once they rely more and more on searching the web: folks are going to find out how to poison the results in a way that pushes users toward their products/services/ads.

    SEO should always have been called index poisoning, because that’s exactly what it is.

  • Valve has done amazing work with Steam Play. Seeing how well the steam deck plays games convinced me not to put windows on my new rig.

    I don’t agree, however, that it just works. My graphics card needed a mess driver outside of the default repos for Ubuntu lts, and my gpu has been out for almost a year.

    I also have one high dpi monitor and one standard-ish dpi monitor, and scaling them independently, moving windows back and forth, and going into and out of full screen games all produce undesired behavior. It’s annoying enough that I now just use one monitor.

  • It has its moments. Season 3 and 4 are better than the first two. It takes a while for any show to hit the tonal balance that makes a good trek. One thing that I think the show does well is fill in the gap between “humanity is a bunch of selfish messy bitches who would prefer to annihilate each other rather than get it together,” and the humanity we see in future series. You’re watching them become less xenophobic, more open minded and inquisitive.

    Skip the finale, the episode before is a good finale.