Artificial
Agentic operating systems will need an audit layer beneath the AI
I had an interesting conversation with ChatGPT about what an agentic operating system might look like and the trust problems that would come with it. Below is a compiled summary that I had ChatGPT construct for this post. The full conversation is linked at the bottom, though the first few prompts are about the singularity before the discussion moves into operating systems. I don’t think an agentic OS would literally replace the desktop with one giant chat box. More likely, the OS becomes i...
Read the full discussion on Reddit
This article was aggregated from Reddit. Click to join the conversation.
View on Reddit