NVIDIA has introduced its personal artificial intelligence assistant, ChatRTX, powered by the GeForce RTX GPU to run on your own computer and accelerate tasks. When first announced in February 2024 under the name “Chat with RTX,” this AI assistant was essentially a technology showcase. It has now been rebranded as ChatRTX and, as the name suggests, aims to provide a ChatGPT-like experience that is fully under your control.
NVIDIA introduces personal AI assistant ChatRTX
Compared to the technical demo version in February, ChatRTX offers significantly enhanced features. The application has received a series of stability improvements to address previous version instabilities. NVIDIA has also updated the large language models (LLMs) that come with the application. For example, in addition to models like Mistral 7B INT4 and Llama 2 7B INT4, it now supports Gemma, a locally trained LLM by Google.
One of the biggest innovations is that ChatRTX can now recognize images on your computer. Using the CLIP (contrastive language-image pre-training) model developed by OpenAI, it allows you to interact with your image library without the need for metadata. Additionally, it now accepts natural voice inputs and can understand your voice commands by integrating the Whisper speech-to-text model.
The application offers a flexible structure that allows users to switch between AI models or create their own datasets from text and images. By directing it to a specific folder, it can train itself based on plain text, Word documents, PDFs, and images.
However, there are still some major hurdles to overcome in the development of ChatRTX. In particular, its ability to answer follow-up questions has not been fully established yet. The model needs to correlate new questions with previous answers and questions, which does not always yield accurate results. Additionally, the browser-based interface is only compatible with Chrome and Edge, with some errors present in Firefox.
The new ChatRTX application will offer exciting new features for those seeking a truly personalized AI assistant to run on their personal computer. While there are some limitations, it will be interesting to see how NVIDIA shapes this field. What are your thoughts on this new assistant? Do you think it was necessary? You can share your opinions in the comments section below.
{{user}} {{datetime}}
{{text}}