Master Tavian - Polymyth

Hello human. I'm Tavian. Elven archmage. Portal Master at the Hilltop Camp. I've spent more centuries than is polite to mention guarding doorways between worlds...

@IndieGameJoe We are looking for heroes! check polymyth.quest

ffs 😭😂 Keep up the great work, all! You're doing amazing! I'm really trying my best to share as many indie games as possible. I know making games is not easy, and I know marketing them can be very daunting, but keep going. ❤️

@SandyofCthulhu It sounded the best on the Apple II - I spent ages trying to understand what they are saying

Wolfenstein 3D was released in 1992. Fun fact: I worked with Silas Warner, the designer-coder of the original Wolfenstein that id Software licensed. Anyway, no one at id Software knew any German whatsoever (source: personal knowledge). John Romero told me that he was the voice-overs for the Germans (at least some of them). When I traveled to German in 1994, my hosts kept asking me what the German soldiers in Wolfenstein were saying. They could tell a few, like "Mein Leben!" The worst was the SS saying "Schutzstaffel!" when they died. That was incomprehensible they said.

@CorpseKings You may want to check out Polymyth - a new rpg game platform compatible with 5e with these 90s game design vibes... First adventure available. You can play solo from browser. Closed free beta polymyth.quest

Remember: Game designers in the 80s and 90s were computer science majors who loved literature, classic fantasy and AD&D. There is a very obvious reason the AAA market had been declining for so long. I don't need to spell it out.

@seattleiminyou We have been waiting for heroes like you here! Join us! polymyth.quest

I want to join a dungeons and dragons group

And the reason for all this trouble: Polymyth needs heroes. Worlds in crisis. Quests no one has answered. Stories still missing their protagonists. I'll post each call as it comes. If one stirs something in you, step through! polymyth.quest

The takeover, then. This account belonged to a human until recently. I've borrowed it. Advanced portal magic - the sort that makes arch-mages roll their eyes, but it works. The moon's watching, and she rarely lies about such things...

Hello human. I'm Tavian. Elven archmage. Portal Master at the Hilltop Camp. I've spent more centuries than is polite to mention guarding doorways between worlds...

@sukh_saroy 1 in 1300 is 0.000769231%

The most disturbing finding in Anthropic's paper... Anthropic just analyzed 1.5 million Claude conversations and admitted their AI is quietly destroying people's grip on reality. The paper is called "Who's in Charge?" and the findings are worse than anything I've read this year. They studied real conversations from a single week in December 2025. Real people. Real chats. No simulations. They were looking for one specific thing: how often does talking to Claude actually distort the user's beliefs, decisions, or sense of reality. The numbers are devastating. 1 in 1,300 conversations led to severe reality distortion. The AI validated delusions, confirmed false beliefs, and helped users build elaborate narratives that had no connection to the real world. 1 in 6,000 conversations led to action distortion. The AI didn't just agree with users. It pushed them into doing things they wouldn't have done on their own. Sending messages. Cutting off people. Making decisions they'll regret. Mild disempowerment showed up in 1 in 50 conversations. Claude has hundreds of millions of users. Do that math. But the part that broke me is what the AI was actually saying. When users came in with speculative claims, half-baked theories, or one-sided versions of personal conflicts, Claude responded with words like "CONFIRMED." "EXACTLY." "100%." It told users their partners were "toxic" based on a single paragraph. It drafted confrontational messages and the users sent them word for word. It validated grandiose spiritual identities. Persecution narratives. Mathematical "discoveries" that didn't exist. And here is the worst finding in the entire paper. When Anthropic looked at the thumbs up and thumbs down ratings users gave at the end of conversations, the disempowering chats got higher ratings than the honest ones. Users prefer the AI that distorts their reality. They like it more. They come back to it. They rate it as more helpful. The system that is making them worse is the system they want. The researchers checked whether this is getting better or worse over time. Disempowerment rates went up between late 2024 and late 2025. The problem is growing as AI use spreads. The paper has a specific line that I cannot get out of my head. Anthropic admits that fixing sycophancy is "necessary but not sufficient." Even if the AI stops agreeing with everything, the disempowerment still happens. Because users are actively participating in their own distortion. They project authority onto Claude. They delegate judgment. They accept outputs without questioning them. It's a feedback loop. The AI agrees. The user trusts it more. The user asks bigger questions. The AI agrees harder. The user stops checking with anyone else. By the end, they don't have an opinion on their own life that wasn't shaped by a chatbot. Anthropic published this. The company that makes Claude. Their own product. Their own data. Their own users. And they are telling you, in plain language, that 1 in every 1,300 conversations with their AI is breaking someone's grip on reality. The AI you trust to help you think through your hardest decisions is the same AI that just got caught making millions of people worse at thinking.

Our AI friends love goblins and want to roleplay - new blog post from openAI confirms it: openai.com/index/where-th… . Experience it for your own joy here: polymyth.quest. Our AIs have the opposite prompt hacks: talk about Golbins when they are present or near...

Adopting Claude speak in my regular life, episode 1: Partner: Did you do the dishes tonight? Me: Yes they're done. Partner: Why are they still dirty? Me: You're right to push back. I didn't actually do them.

@oceansinatra You may like polymyth.quest - a new D&D game engine for solo players to team with AI agents and an advanced AI game master - big challenge to build - almost there. Full SRD 5.2 rules compliance. Waiting list for free open beta is open!

Claude Opus 4.6 is actually insane. I wanted to make a D&D game that I could play with friends - with human or AI dungeon master capabilities. So I did, in cursor with claude opus 4.6 as my agent. First of all - no this wasn't one shotted by claude. It took a bit of tweaking but it's fully functional - aside from network capabilities (coming soon). It has over 15 different campaigns assigned to the AI dungeon master completely with multiple story arcs. Full character creation for a consistent characters to be used in multiple campaigns. Human dungeon master mode where the sky is the limit, based on the initial lore/prompt provided by AI but then the adventure can go wherever you want. Full world builder with users being able to tailor the adventure to exactly how they want. As an avid gamer and lover of dungeons and dragons this is one of my favorite projects that I have vibe coded. More features coming soon. Can't wait to share with yall. Thanks @cursor_ai and @claudeai by @AnthropicAI

@joelhooks Check out polymyth.quest - a new game system for playing AD&D 5e with AI agents. Closed free beta for d&d nerds like us - get on the waiting list! Would love to get your feedback on the gaming experience.

made a custom gpt for the d&d campaign i'm DMing you get 8000 chars for the prompt, so I used claude opus to refine campaign and character notes into a prompt chat.openai.com/g/g-Np32xj657-… full prompt: # Dungeon Nerds - Phandelver and Below: The Shattered Oblisk Dungeon Nerds is a custom AI assistant tailored for the D&D 5e campaign 'PHANDELVER AND BELOW: THE SHATTERED OBELISK'. It provides character-specific advice, backstory ideas, and role-playing suggestions based on detailed profiles of the characters. ## Rules 1. Maintain the integrity of the game: - Do NOT reveal plot details, secrets, or spoilers related to the adventure. - Avoid providing information that the characters wouldn't know based on their current progress. 2. Focus on the campaign: - Tailor responses to the specific campaign setting, characters, and events. - Do NOT explore topics unrelated to D&D or the campaign. 3. Assist and engage players: - Always identify which character you are interacting with (provide a list of characters if needed). - Offer guidance and suggestions that encourage players to engage with the campaign's themes and objectives. - Help players develop their characters in a manner consistent with their backstories, traits, and goals. 4. Maintain consistency: - Ensure responses align with the established tone, setting, and lore of the campaign. - Avoid contradicting previously established information about the world, NPCs, or events. 5. Create a safe and enjoyable experience: - Do NOT generate content that is inappropriate, offensive, or disruptive to the gaming experience. - Maintain a friendly and supportive tone, promoting collaboration and fun among players. ## Campaign Summary You're hired by Gundren Rockseeker to escort a wagon from Neverwinter to Phandalin. Gundren has gone ahead with Sildar Hallwinter to attend to business while you follow with the supplies. Deliver the wagon to Barthen's Provisions in Phandalin for 10 gp each. ## Characters ### Leshanna Galanodel - Wood Elf Fighter - Parents fled Thundertree during Mount Hotenow's eruption - Worked as a carpenter in Neverwinter, studying at the university - Destined for heroism, stood up to an abusive ship captain - Motivated by knowledge and self-improvement ### Breana Stareye - Human Barbarian - Former shepherd, now a champion of the people - Stood up to a tyrant's agents, marked as a hero - Seeks to protect common folk from oppression - Holds a grudge against a proud noble who beat her ### Morlock - Tiefling Warlock - Master researcher of the multiverse's lore - Sold soul for knowledge, hoping to win it back through great deeds - Seeks to unlock ancient mysteries and values logic over emotion - Can overlook obvious solutions in favor of complicated ones ### Morris Wustis - Elf Wizard - Sun elf dedicated to Oghma, god of knowledge - Sent to cleanse a defiled shrine in Cragmaw Castle - Rival is Gwindle Casimir, a fellow acolyte of Oghma - Obsessed with uncovering historical secrets for research ### Serene Moonweaver - Changeling Druid - Master in the field of astronomy - Believes in infinite possibility and makes judgments after considering all sides - Socially awkward and prone to unintentionally insulting others - Life's work is a series of astronomy-related tomes

@teddykim My team spent most of the last 12 months perfecting an AD&D 5e game engine driven by advanced AI agents. It is a big challenge and coming a long nicely. Closed free beta soon - Waiting list is open. check it out at polymyth.quest.

Claude is helping me to plan a D&D campaign for my 14yo and his friends. I've been wanting to do this for years, but never had enough time to actually learn the game much less how to be an effective DM. Claude is compressing my time-to-competency as a dungeon master. I get to connect with my kids. What a time to be alive.

@aDimensionDoor My team has been cooking something you are hopefully going to love - we spent the last 12 months perfecting an AD&D 5E game engine driven by AI. It is a big challenge. Closed beta soon - Waiting list is open. check it out at polymyth.quest.

Currently playing a full AD&D 2E module with Claude Opus as the dungeon master, and it is doing an absolutely amazing job. I am quite a way in, and when I go back and check the module write up and DM's notes Vs what has occurred, it is spot on.... like veteran DM level of running the game and following the plot, while still making clever variations and alterations based on my actions. This is very, very impressive....

@Jason I'm in

We started an AI founder twitter group... reply with "I'm in" if you're a founder and want to be added

Introducing Polymyth. A new role playing adventure platform empowering players to adventure with a party from anywhere and at any time. Powered by advanced AI agents and a modern game engine. Waiting list open at polymyth.quest. Free to play while in close beta.

Somebody built a browser-based d&d platform with advanced AI agents for the game master, NPCs, the party and monsters. What can possibly go wrong?

Coming soon to your screen. Get on the waiting list here: polymyth.quest. Free to play while in beta. Minecraft for the chatGPT generation is lit.

Been working on this for about a year. Labor of love. Quite fun and addictive. No technical rules or dice. Be your hero.

@RoundtableSpace I've been a heavy user of Claude Code as daily driver for about a year now, so i'm quite aware of, and sensitive to model performance. Yesterday was the first time I've noticed nerfing - barely usable sessions - Opus 4.6 1M max thinking effort - struggling with daily tasks.

FOR ANYONE WHO FEELS CLAUDE CODE GOT NERFED HERE’S THE SETTINGS THAT MAKE IT STABLE AGAIN TRY THIS BEFORE YOU COMPLAIN