multi agent AI collaboration problem solving protocols? (beyond MCP?)
Details
What if AI could negotiate with other AI…?
this is a very important topic in AI... how do AI's "work together" and what is the specifics of API and multi agent AI problem solving collaboration protocols?
TOPIC: multi agent AI collaboration problem solving protocols?
(beyond MCP?)
DATE: April 29, 2026 (Wednesday)
TIME: 5:30 PM (PACIFIC TIME ZONE)
WEEKLY FREE OPEN PUBLIC MEETING
We’re hosting a 20–30 min rapid session + open demo
on something slightly unhinged (but maybe inevitable):
→ Multi-AI collaboration protocols
→ Human-in-the-loop decision layers
→ A new kind of “currency” built on real + imaginary value
Yes… imaginary numbers.
No… not imaginary impact.
We’re exploring several real world example systems like:
• liarz.org
• lyrz.org
• liez.org
• blnq.org
AutoGPT
LangChain
CrewAI
Custom MCP could connect through an API where AIs:
coordinate tasks
validate each other
negotiate outcomes
and escalate to humans when needed
Think:
“APIs for trust between intelligences”
⚡ Agenda:
• 5 min — The weird idea (why AI needs protocols, not just prompts)
• 10 min — Architecture sketch (agents, validation loops, human override)
• 10 min — Demo (early concepts + experiments)
• Open — Discussion + wild ideas
Questions we want to explore:
Can AIs form consensus without central control?
What does “trust” look like between models?
How do humans stay in the loop without becoming bottlenecks?
Could value systems include both measurable + “imaginary” (contextual, inferred) components?
If it sounds like the future… join us?
Drop a comment or DM if you want!
Thanks!
Ash. (speaker)
773-321-8181
PS: We can allow up to 100 people! However we are hoping for something like 10-20 people.
;)
PS: feel free to invite other!