Open Source Personal AI Assistant
Hi All!
I'd like to introduce my first project on indiehackers :-)
Would you like to take advantage of modern AI developments, and have your own personal Assistant? But you feel creeped out by how much Google / Apple / Amazon already know about you, and you'd like to avoid completely integrating your private life into their systems?
Let's create our own open source AI, where we control our data and privacy!
Imagine your own personal AI assistant. Connect it to any information and media source you like. Twitter, gmail, GPS, your smart watch, your microphone, whatever you like. The AI keeps you up to date, helps you schedule your holiday, warns you about rain before you go jogging etcetera.
That's the premise for my new project, Anagon AI. What do you think?
- Would you be interested?
- What would you like it to do?
- Why might this be a bad idea?
Any reaction welcome!
@anagon nice hypotheses, it reminds me Evie AI before they pivot, as well as many AI-driven virtual assistants out there.
How are you planning to execute differently?
thanks for your reply!
with this project, the aim is not to build a conversational bot for social or entertainment purposes, but rather to help the owner to organizer their life better. actual assistance :)
I'm personally thinking about lifestyle coaching, productivity, knowledge management, analytics (on yourself).
the goal is also not to create a digital personality or some sort of avatar for the bot. it is simply an 'intelligent' tool that you can interact with in a variety of ways. i'm not sure yet which modality will turn out the most popular. it could be talking/typing or some combination
A generic automated assistant where I can connect it with many sources, own my data, I can tune it, or extend it if I have the skills. Definitely useful.
But that sounds like a huge task. This could be a long term goal. Short term you should perhaps seek for a specific use case where there's a lot of value for users, and focus on that.
thanks for your question ppanagi! it is a huge task indeed. I hope to counter that by creating an open source community around this project. as of yet, i have no clear idea how to do that, except by creating a great developer experience (https://github.com/anagon-ai/anagon-ai).
your point about an initial use case is important. that is an open question right now, what that case would be.
in the first phase, I want to attract developers, later users. so i'm thinking the project will start off mostly as a customizable framework/library, and is easy to use, so that developers will get interested. Then they will create cool stuff with it that will attract users. in theory :)
but still, a user-focused mvp will be valuable too
Love it