Why should something like this make it to the front page?
Because agent memory is a real issue if you tried to build any agent. But yeah a universal protocol by a small player will not solve it.
Universal Memory Protocol – a shared format for agent memory
31–40 of 46 posts
Re: Universal Memory Protocol – a shared format for agent memory
#32Earlier quoted context omitted.
Yes https://github.com/space-bacon/SRT I can read any models every thought. No one cares. Not the narrative.
Saw this shared a few days ago, skimmed it, didn't understand it. See it again now, another skim, still don't understand. I think it could use a ELI15 or something.
It’s the babel fish from hitchhikers guide to the galaxy that can also be the pov gun.
Re: Universal Memory Protocol – a shared format for agent memory
#33Re: Universal Memory Protocol – a shared format for agent memory
#34You are at stage “Memory Architect”. https://delightful-marigold-803f7f.netlify.app/
So this page is just to shame people using LLMs and to make you feel better?
More to point out everyone has the same LLM discovery patterns. But none of the methods seems to have observable, measured improvements over the basics and tend to regress back.
Re: Universal Memory Protocol – a shared format for agent memory
#35[dead]
Re: Universal Memory Protocol – a shared format for agent memory
#36Earlier quoted context omitted.
So this page is just to shame people using LLMs and to make you feel better?
More to point out everyone has the same LLM discovery patterns. But none of the methods seems to have observable, measured improvements over the basics and tend to regress back.
The tech is barely two years old and people are exploring the usage.
I think it's better to keep an open mind instead of claiming to have seen the pattern. Indeed, there is a lot of hype and delusion but there is genuine progress and it's clear as the sun, anyone claiming anything else is equally risky of being delusional at the other end of the spectrum.
Re: Universal Memory Protocol – a shared format for agent memory
#37* …tools, UMP does for memory - negotiated operations over a portable, signed, bi-temporal record … * What is a bi-temporal record? I don’t think I’ve heard the term before and I’d love to learn more.
IIUC, the most basic version is when you have a log where every entry has both “date added” and “effective date,” so you can add stuff to the log retroactively. For example, “the user just informed us yesterday that they moved last year” -> address date added=yesterday, date effective=last year
Thank you!
Re: Universal Memory Protocol – a shared format for agent memory
#38[flagged]
Re: Universal Memory Protocol – a shared format for agent memory
#39Earlier quoted context omitted.
So this page is just to shame people using LLMs and to make you feel better?
More to point out everyone has the same LLM discovery patterns. But none of the methods seems to have observable, measured improvements over the basics and tend to regress back.
> everyone has the same LLM discovery patterns
there's so much discussion about AI and "everyone" might just be reading the same things.
this is a good indication not to treat this particular phrasing of a discovery path as canonical
Re: Universal Memory Protocol – a shared format for agent memory
#40Not sure if this is the right abstraction: The recall seems to need a search term.
But would it not be more sensible to assume that the full conversation (+ system parts) CAN inform the recall and some neural network picks the right memory bits?
So my fear would be that something like this, if adapted, drags the development into a local optimum that is hard/impossible to get out of.