• 0 Posts
  • 32 Comments
Joined 1 year ago
cake
Cake day: June 17th, 2023

help-circle





  • Very skeptical of that one.
    They’ve been trying to target amyloid for more than a decade, and it’s the first time I hear of it actually working.

    The treatment seem to have huge side effects (brain bleeding and swelling) and lead to patient death during the study.
    Elly Lilly is also know for marketing zyprexa as a treatment for dementia (despite inefficacy and increased risk of death). Which is IMO criminal, at best unethical. I’m not inclined to trust them at all.

    I hope I’m wrong and it works. Alzheimer is a terrible way to go.











  • Trying to find some that haven’t been talked about yet:

    Echo. It’s a fantastic experimental infiltration game with an AI that adapts to your way of playing. The setup is very impressive.

    Pathologic: one of the three playable characters (the Changeling). It’s a bizarre russian game, with an unique world, and messy gameplay. Can’t recommend it enough.

    Va11 Hall-A: chill bartending game in a cyberpunk setup.

    The Blackwell series: comfy, kind of amateurish point and clicks by Wadjet Eye. I like them very much.

    Transistor: weirdest game by Supergiant. You play as a redhead with a talking sword. I don’t remember much about it except that it was good.

    The Fall: (pushing it a little bit, since the protagonist is an AI, but I’ve always seen here as female.) Criminally underrated puzzle games, disguised as metroidvanias.

    Eliza (by Zachtronics): the only visual novel I enjoyed. It’s hard to explain, it’s about AI, burnout, whether tech dehumanizes people, and solitaire.

    Hedon Duology: for something completely different, it’s a slightly kinky retroshooter, with amazon Orcs fighting demons.
    It may sound a bit dumb, but it’s excellent. Huge levels, interesting worldbuilding, and a gameplay based on exploration, puzzles as well as shooting.

    There’s probably a ton more, but that’s all I can think about at the moment.


  • 10 years limit, absolutely non transferable, limited to human beings (not abstract legal entities) .
    Eventually extendable to lifetime of the creator if the work is still being developed, to prevent being usurped by copycats.

    I also believe that facilitating voluntary sponsorship (a la patreon, but without letting 10% get siphoned by leeches) is preferable to selling works. Especially since distribution is now pretty much free.



  • I can think of a way to help with the problem, but I don’t know how hard it would be to implement.

    Create some sort of trust score, where instance owners rate other instances they federate with.
    Then the score gets shared in the network. Like some sort of federated whitelisting.
    You would have to be prudent a first, but not do the whole task yourself.

    You could even add an “adventurousness” slider, to widen or restrict the network based on this score.