ai16z Eliza

Penseur.ai is currently testing the creation of a set of agents that utilize the ai16z framework. There are two approaches that we're examining for integration.

The first approach is to leverage some fairly strong functionality that eliza has that virtuals protocol (VP) currently doesn't have. We could put a facade in front of eliza and provide all of the functionality to Penseur via that facade. There are straightforward mechanisms to do this given the way VP operates, but this is most definitely not architecturally sound and a terrible hack. However, it does provide an easy way to get the functionality out quickly and as such may become our short term approach.

The second approach is something that may take a little but more time, but will allow us considerably more flexibility - and that would be to put both VP and eliza in the background along with the other agents that we have in development. We would then use Penseur.ai as the orchestrator for agentic workflows on its own and then utilize the various frameworks where it makes sense. This has one strong and obvious benefit - it provides a single interface to the platform that we can expand across multiple frameworks and it also means that we don't have to have 500 different tokens. I am a huge proponent of stopping token sprawl. I think there are too many as it is, it creates all sorts of mechanisms for fraud and deceiving users with wallet/email/website hacks. This second approach is where we want to go - we just need to assess how long it would take to build this WITHOUT building yet another agent framework. We don't want to create yet another way to build an agent either. None of that is good for the community UNLESS you have some new and novel way to do things. Since we're not in this for the rapid cash grab - we will certainly investigate both paths from the perspective of securing the folks that are already holding tokens but making sure that we can continue to rapidly add functionality.