I wanted to start a discussion around AI agents, where you see them being helpful and whether you're working on any. Didn't see much in search about this.
Ask HN: Agents? Are you making them?
1–4 of 4 posts
Re: Ask HN: Agents? Are you making them?
#2I am yes. But I really want to run it locally and until llama 3 their performance wasn't really there yet. But I'm impressed so far.
I'm planning to make it more proactive, like event driven. When a message comes in, or something happens in my house it sounds determine if I should be told depending on my status.
Just haven't had a ton of time to work on this and I'm not doing it professionally.
Re: Ask HN: Agents? Are you making them?
#3Yes, I connected claude 3 Opus with whatsapp and create functions that I use it as a support creating content to post on Instagram
Re: Ask HN: Agents? Are you making them?
#4I am. I'm working a on a AI Agent/Chatbot, with enhanced multi-modal capability [1].
It's still early, and since I'm still learning Python and Gen AI/LLM in general, it's a bit difficult. But I've got too fascinated to learn experimenting on building my own AI Agent or just Chatbot, that could helps me, first to learn, secondly to help me with learning about new models capability and LLM landscape.
And I also open source it, fully.
My mantra is learn by doing, and by doing, I also want to share with everyone in the open source community since most stacks and resources I use to built it is from GitHub.