triall.ai
The AI hallucination fix. AIs debate in parallel, then refine in loops. Council + iteration = answers no single model can touch. https://triall.ai
👤 Joined on 02 February 2026
Followers
Following
Total Posts
@triallai.bsky.social@triallai.bsky.social → @triall.aigot contacted about an AI role (🤮) because of my automation background. One of the skills they want is "developing hallucination detection systems" WHY ARE YOU USING SOMETHING THAT CAN **HALLUCINATE** SO MUCH YOU NEED TO DEVELOP -DETECTION- SYSTEMS???????
Show PostVery funny to go through the trouble of developing and scanning 100 year old film and then just let ChatGPT hallucinate all over them.
Show Post"I installed this Agentic AI and asked it to be my email assistant and it deleted everything" is technically a bad outcome, but at the same time, still gets to *vibe* like the AI *made a decision* -- oh, it's unreliable, sure, but it's A REAL THINKING DECISION-MAKING ENTITY. (Spoiler: it is not.)
Show PostKitty Kelley, known fabricator of decades-long standing, (allegedly) using ChatGPT to hallucinate an entire book review? Yeah, least surprising book news here
Show Post"Train our AI or the next hallucination might be the one that kills you" is an interesting marketing strategy
Show Post