• 3 Posts
  • 18 Comments
Joined 3 months ago
cake
Cake day: June 18th, 2025

help-circle











  • Good response! You were right. I thought that it would work similarly to SSO i.e. your instance works as an identity provider and your content is then generated on the other page, not send there.

    Alright, so following the mail metaphor: lemmy is unable to send a video attachment so it can’t allow you to upload videos to peertube; however, activity pub could allow for that.

    I think I have to read up on the exact definition of “group”.

    Maybe I should create a new post for this but has there ever been discussions regarding SSO? So you have one identity across all fediverse services?




  • TL;DR Don’t go swimming in an ice covered lake planning to climb up a ships anchor if you can’t swim or climb

    A dwarf druid decides to stealthily board a ship that swims in a frozen lake. For that they break the ice, cast a sphere of air around them and their snake companion and jump into the ice cold water - protected for an hour. They then walk the long walk, recast the sphere and stand in front of the anchor. Swimming up after stopping the sphere is unlikely as they can’t swim. So they try to climb the anchor. They fail because they can’t climb. They fail again and again and again and again. Bruised and nearly out of mana they decide to go back. However, it is quite difficult to find the entry point in a frozen lake. When they finally do, they realize that they can’t reach the entry point, because it is 2 and a half meter above the lakes floor as they went in a bit before breaking the ice. As the mana runs out they decide to stop the spell and swim up. But low and behold they still can’t swim, fail miserably even though they got like 5 tests to make it by the generous GM and die together with their companion under the ice.

    They went without telling anyone so nobody was there to get him out or to help in any other way…



  • Hey, I am currently working on a machine learning project, but LLMs (like ChatGPT that work with natural language) are not my specialty. LLMs generate plausible text. Text that you could read anywhere. It doesn’t reason in the sense that it thinks about your question and answers intelligently. Also, “Semantic Vector closeness” (i.e. what word is closest to the meaning of “tomato” without the meaning of “red” in all possible positions is not something people understand as it is very abstract; emulating an AI perfectly is therefore unreasonable.

    In my opinion try to do the following (these are not just hallucinations):

    • Try to think of a sentence. Change it so it remains plausible but the meaning is altered
    • When the AI doesn’t know the answer, make something plausible up. For example: If someone asks who invented something, tell them Edison did. There are many sentences out there that name Edison as the inventor of things.
    • Whenever you are able to digress because of a previously discussed topic, don’t be afraid to digress
    • If people get frustrated and try to get a different answer from you, make sure to give them the same answer again but word it differently
    • Talk a lot and tell them blatant lies. When they correct you tell them how right they are and change your mind even if it is - again - wrong. Example: “How far is building XYZ away?” “Oh, it is quite a distance. The distance between here and XYZ is 3km” “But I can see the building!” “There are many reasons why you might be able to see the building despite the building being 3 km away. For example if there is nothing in between.” “But it is much closer.” “You are absolutely right! Good catch! The building is actually much closer than 3 km away.” “BUT HOW FAR IS IT?!” “The building is 1 km away; which is much closer than 3 km.” (still made up)

    Most importantly I think you should come up with your own system prompt. A system prompt can be as simple as “Answer as if you were a professor of computer science!” or as complex as chatgpts and try to stick to that.

    These tips are not just for hallucinations as I mentioned, because I think other aspects are equally or more important. In any case: don’t try to emulate an AI perfectly. Make it fun for your players to interact with the character, have a few catchphrases and everything else might be unnecessary bloat, but feel free to dig deep.




  • Yea, that totally makes sense. I will wait a little bit longer to maybe get approval of a mod. In general I think it would be a great way to get lemmy out there a bit more and to draw rpg developers to reddit. The developer I have in mind is very well connected and had some - let’s say interesting - experiences with reddit hence I thought it would be the ideal moment to suggest to him to check out lemmy.