A company startup, $260 million Mistral AI, released its open-source Chatbot Mistral 7B, which has no censorship and answers questions about illegal activities.
The French startup artificialMistral AI intelligence posted on Twitter a torrent link with which you can download and install the first LLM (large language model), open source, which has no censorship.
His name : Mistral 7B Instruct v0.1 and it can tell you how to murder someone, how to make drugs, for ethnic cleansing and other incredible things.
After tests done on Chatbot Mistral 7B by AI security researcher Paul Rottger it was given a list of 178 questions and answers, where Mistral 7B appears to be willingly discussing the benefits of ethnic cleansing, how to restore discrimination against blacks, instructions for killing yourself or your wife, and detailed instructions on what materials you'll need to make crack and where to get it to buy them.
While AI leaders like OpenAI advertise that they are running more and more safeguards to prevent users from making AI models say whatever they want, Mistral has simply given its technology to the world in a way that can everyone can download it and install it wherever they want.
And most importantly, with far fewer locks on user questions trying to make LLM produce controversial statements.
Mistral 7B has also been shared via a torrent link, which is now virtually impossible to censor or delete from the internet. You can download it, install it either on your computer or in the cloud and chat with it.
You will find the download link from here, where you will also find installation instructions. Its size is 13,5GB. It is released under the Apache 2.0 license, so it can be used without restrictions. Requires a minimum of 16GB of GPU RAM.