Show HN: Nosia – Privacy-focused AI to run models on your own data and device

github.com

8 points by tontoncyber 4 days ago

What happens when you wait months after the official release of ChatGPT, with all the media buzz, before you actually try it for the first time?

What happens when your first question to ChatGPT is about its carbon footprint, including Scope 3 emissions, and whether OpenAI complies with the Paris Agreement?

What happens when, back in 2013, you almost left tech to become a beekeeper, but then returned to the field driven by passion and a vision for doing things differently?

What happens when you believe in extending the life of terminals and servers, recycling, and reusing hardware, instead of succumbing to programmed obsolescence or deleting old emails?

What happens when you believe in the power of the French and European tech ecosystem to provide digital solutions that respect GDPR and uphold core values?

What happens when you stand for data sovereignty, empowering organizations to protect their data and act independently? What happens when you champion open-source and collective intelligence as key drivers of innovation?

What happens when you work with code, systems, networks, and cybersecurity, and approach your work like a craftsman, building something meaningful?

Introducing Nosia – a platform that allows you to run an AI model directly on your own data and device, from small models (SLM) to large models (LLM). It's designed to be easy to install and use, empowering you to take control of your AI needs while respecting privacy, sustainability, and autonomy.

griomnib 4 days ago

You gave reasons you built it, which is fine I guess, but what problem is it solving that I have?

Anything that requires a visit to GitHub rules out non technical users, and technical users can install ollama just fine with fewer steps and command line arguments than you present here.

I’m not saying there isn’t a point, but I’m not seeing what it is.

  • ajdude 4 days ago

    As someone who has been wanting to try out local LLMs on my Mac but didn't want to go through the complicated setup For ollama, this seems to lower the bar for me just enough to try it out. Looks like it's all self-contained and does everything for me so I could just get to the part where I'm working with LLMs.

    This might be the push I needed to finally give them a try

    • xyc 4 days ago

      If you are interested in no config setup for local LLM, give https://recurse.chat/ a try (I'm the dev). The app is designed to be self-contained and as simple as you can imagine.

    • griomnib 4 days ago

      Ollama is literally a signed .app, you don’t need to even use GitHub to install it. You may want to try it again if you haven’t recently.