My vibe-coded "AI News Radar" pivoted quite a lot over the weekend. I’ll pause it here – for now. But this is where I am after 1 full day:
The AI News Radar – Version 3.8.2
Imagine having a personal assistant that not only aggregates the latest AI news from your preferred sources (aka RSS feeds) but also analyzes it to highlight key trends – all running locally on your machine, ensuring full data control. This is precisely the vision behind my "AI News Radar" project.
What can it do?
It’s a flexible dashboard designed to help me stay ahead in the fast-paced AI sector:
Dynamic Information Hub:
- Visual Radar: Displays news from the last 24 hours as interactive dots – the closer to the center, the more recent. Hovering provides details and highlights the corresponding article in the news list.
- AI-Powered Trend Analysis: A locally run LLM via Ollama analyzes aggregated headlines to identify up to five key trends incl. confidence scores and supporting keywords.
- Interactive News Exploration:
- I can chat about current news and trends.
- Clicking on a trend panel in Column 1 displays all related articles.
- Curated News & Feed Management: Scrollable list of all collected news items including category information provided by the feed.
- Add & Manage Feeds: Streamlined management of RSS sources.
Technology Stack (all running locally):
Python (Flask), Ollama (for the local LLM), HTML5, CSS3, JavaScript (ES6+), feedparser, requests (soon), BeautifulSoup4 (soon).
Why this Project?
With rapid AI advancements and information overload, it’s crucial not just to be informed but also to recognize relevant patterns and trends early on. The AI News Radar is my ongoing experiment to create a tool that enables exactly this: a personalized, AI-augmented view of the AI world that operates locally and respects data privacy. The ability to interact directly with an LLM that contextually accesses your specifically curated news corpus – and will soon "read" full article texts – opens up fascinating possibilities for deeper, individual engagement with the subject matter. It’s a step away from generic chatbots towards specialized, context-aware AI assistants for your own desktop.
This project remains, in typical "AI Logbook" manner, practical and always aiming to make the potential of local AI applications tangible and understandable.