Agentic AI seems to be getting a lot of attention. I fell into a trap of digging into it a bit and I'd like to share what I found.

One framework I particularly liked is called Eliza (no, not the one from 1960s). I'd like to walk you through what I've learnt about AI agents and help you build a simple agent that you can take home (on your laptop). We'll use a state-of-the-art LLM for it's brain and plug it into Telegram, so that we can have a little chat with it.

I'm doing my best to use decentralized, private and censorship resistant tools for development daily. Some stuff just works, some is a bit..meh:). Let me walk you through a bunch of open source projects I especially value for their hard work, commitment to open source and focus on privacy and censorship-resistance.

I'll explain and demo projects like Radicle, Fileverse, Status, Akash Network, Drips Network, Continue Dev... which allow you to "escape" the centralized dev tool world.