NVIDIA is bolstering its experimental chatbot ChatRTX by grafting new AI models onto it. As a result, the latter sees its arsenal expand and its capabilities evolve considerably.
Introduced last February as “Chat with RTX,” ChatRTX was initially just a demo app. In concrete terms, the application creates a local chatbot server that can be accessed from the browser. This allows you to feed the AI with your documents and even YouTube videos, turning your machine into a powerful search tool that can summarize your content and answer your questions.
A supercharged, but still a bit heavy personal assistant
In terms of minimum system requirements to enjoy ChatRTX, there are no changes on the horizon. To take advantage of this, you’ll need an RTX 30 or 40 series graphics card with at least 8GB of VRAM.
Initially able to take advantage of the Mistral and Llama 2 models, ChatRTX can now count on new models: ChatGLM3 (bilingual model in English and Chinese), OpenAI’s CLIP (capable of generating textual descriptions from images, and vice versa) and Google’s Gemma. The latter has been specially designed in collaboration with NVIDIA and works wonderfully on solidly equipped PCs.
ChatRTX centralizes all of these new models and takes care of making it easy to run them locally. The interface seems to be quite intuitive and you can juggle different modules according to your needs: analysis of your photos, online videos or document summaries for example.
To use it, all you have to do is go to NVIDIA’s official website and be a little patient. In fact, ChatRTX requires no less than 36GB of free space to install!
Voice, the next interface with AI?
We can imagine that the results are more convincing than the first demo, where ChatRTX struggled a bit to perform some of the tasks it was asked to do. NVIDIA has not only improved the support of its model, but it has also equipped it with a voice recognition system. What? By integrating Whisper, an artificial intelligence model designed by OpenAI, specifically designed for speech recognition and transcription.
In this way, the user will be able to query ChatRTX, simply with the help of their voice. Could this be the new direction for chatbots? Break free from the need for the keyboard and simply listen to answer questions? It’s hard to say, but it’s possible that this initiative will lead the AI industry in that direction.