AI will “get” news from the Internet and start a conversation with you.
This project uses a web scraper to gather news/articles from various websites and feed them into AI models, along with appropriate prompts to instruct them to generate topics and initiate a conversation.
This is an example using DeepSeek API to chat with.
The number of rounds you want to chat with the AI: 3
AI: Hey! So, Nvidia just dropped their new RTX 5090, and it’s a beast for 4K gaming. It’s got this crazy DLSS 4 tech that can boost frame rates like crazy, but it’s also $1,999 and sucks up a ton of power. What do you think—worth the hype, or is it just overkill for most gamers?
User: Oh, I won’t buy that anyway. My RTX3060 is enough for most of the games I play. To be honest, it would be a waste if a student like me uses it to play games.
AI: Yeah, totally get that! The RTX 3060 is solid for most games anyway, and the 5090 feels like overkill unless you’re deep into 4K or heavy AI workloads. Plus, that price tag is wild for gaming, especially if you’re a student. What games are you playing these days? Anything pushing your 3060 to its limits?
User: Yeah, Nvidia today focuses more on AI, some old GPUs are enough for gaming. But you know, the 3A game Black Myth: Wukong consumes a lot on GPU, my RTX 3060 felt some pressure and I had to lower the graph quality.
AI: Oh, Black Myth: Wukong is such a stunning game, but yeah, it’s a beast on GPUs. Even the RTX 5090 struggles a bit without DLSS! Lowering settings is totally fair—it’s still a gorgeous game even on medium. Have you tried using DLSS or FSR on your 3060? It might help squeeze out some extra performance without sacrificing too much quality. What’s your favorite part of the game so far?
-
Clone this repository.
git clone https://github.com/flowing-wind/Proactive-AI-Talk.git
-
Install necessary libraries.
pip install -r requirements.txt
-
Create a new file
api_key.txt
in the root directory of the project and fill in the API key of your model. -
Run
init.py
and this will create adb
f