Locked OpenAI research on the potential of the new GPT-4 version and how they need to limit its potential for malicious purposes.

[dEEpEst]

OpenAI recently launched an investigation of the potential of its new version GPT-4.

In the study you can see its potential for malicious purposes. OpenAI worked to restrict all those responses. For example:

As you can see GPT-4 becomes macabre without the restrictions of OpenAI.

You can download all the research at this

This is the hidden content, please

• Sign In
• or
• Sign Up

, it is not wasted.

[Boom Diggetty]

I'm glad they didn't leave it in the hands of the major corporations to take control of and capitalize on.  Putting it in the hands of everyone to be able to train and learn and benefit and even profit from equally just seems fair and reasonable right???