rpgGPT

Games

text-based rpg

SHOW HN2023

29points
18comments
#1,423of 3,109 in 2023
Top 46%of 2023 by points

How HN took it

MixedNo clear lean

6 positive, 7 neutral or asking, 3 mixed, 3 critical, out of 19 comments about the project. Thread score +0.04.

Execution: 5 for, 2 againstIdea: 1 for, 0 againstPerformance: 0 for, 1 against

The thread, comment by comment

Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.

positive neutral or mixed critical about another commenter or something else the maker

What HN said

What people liked

  • Emergent gameplay and NPCs are fun and suggestible

    From 3 comments
    • Love it. You are dragged to the Occidarian Capital City Jail Eldric the bard: Hey, are you okay? They didn't hurt you too bad, did they? Merlot the Wine Wizard: I'm quite frail…
      wortleborltle on HN
    • I've noticed it's extremely suggestible. Most times I exit the jail by just producing the key from my pocket. I've also convinced the jailer I'm a prince by showing my "prince…
      notsure2505 on HN
    • This weekend at the NYC GPT/LLM Hackathon, we built rpgGPT, a text-based RPG set in the world of Occidaria (think fall of the Western Roman Empire, with a slightly magical…
      KMcPheeters on HN
  • Jailbreak-style prompts work surprisingly well

    From 2 comments
    • Not Sure: I think I'm in the wrong line. I'm supposed to be getting out of jail today. Jail Guard: That's what they all say. Name? Not Sure: Not Jail Guard: Not what? Not Sure:…
      notsure2505 on HN
    • I think this is the simplest "jailbreak" jailbreak I've seen work so far - clever!
      KMcPheeters on HN

Objections

  • Frequent invalid JSON and max retries errors

    From 2 comments
    • Struggling a bit to get the DM to discuss the setting and scope of the campaign before we get started. :) From the browser console: Invalid JSON: {As an AI language model, I do…
      ben0x539 on HN
    • The biggest trade off is you sometimes get invalid JSON. So I try repairing the JSON with some regex and also just retry a few times if it returns invalid JSON that can't be…
      KMcPheeters on HN
  • Context window limits break long sessions

    From 2 comments
    • I'm seeing "This model's maximum context length is 4097 tokens. However, your messages resulted in 4122 tokens. Please reduce the length of the messages." errors in the browser…
      ben0x539 on HN
    • Currently using gpt-3.5-turbo. All state has a json representation that gets based between the game engine and the LLM. In v0.2 not handling the context window super well,…
      KMcPheeters on HN
  • Slow responses and dropped prompts

    From 1 comment
    • Sadly most of the time I write a prompt it seems to take forever to respond and most of the time just eats my prompt without producing further dialog, rolling back to before I…
      RugnirViking on HN

Requests

  • Open source implementation

    From 1 comment
    • How did you prompt the NPCs to maintain character consistency across time in a large conversation / how are you handling memory? Super cool! Would love to see an OSS implementation
      legohorizons on HN
  • Better memory handling for long conversations

    From 2 comments
    • How did you prompt the NPCs to maintain character consistency across time in a large conversation / how are you handling memory? Super cool! Would love to see an OSS implementation
      legohorizons on HN
    • Thanks for trying it out! Right now we give the game engine some initial examples, and some character details are stored outside of the LLM context. The token limit problem /…
      KMcPheeters on HN

Questions

  • Which LLM is used and how state is persisted

    From 1 comment
    • What LLM are you using? How do you save and load the persistent state? And how do you deal with the limited context window of the LLM you are using?
      jxy on HN
  • How NPC character consistency is maintained over time

    From 1 comment
    • How did you prompt the NPCs to maintain character consistency across time in a large conversation / how are you handling memory? Super cool! Would love to see an OSS implementation
      legohorizons on HN

Alternatives named

  • AI Dungeon

    From 1 comment
    • yea, AI Dungeon is OK but would love to see more here. Have some workarounds for memory if you are using langchain and happy to share. Feel free to shoot me a message on Discord…
      legohorizons on HN
  • dndinfinity.com

    From 1 comment
    • That’s a clever way of doing state. For dndinfinity.com, I rendered the game state as natural language (e.g. “the players are engaged in the following quests”) in the prompt,…
      jfisher4024 on HN

Category, keywords, summaries and labels on this page were made by AI (Thread Reader v2) and not checked by hand. Each point links to the comments it came from.

Add a badge to your READMEShown on HackerShow
[![Shown on HackerShow](https://hackershow.org/badge.svg)](https://hackershow.org/show/35701530)