I don’t want it to do things for me. It very often eliminates my ability to vet the information I want at a glance.

I also think it’s basically being weaponized as a propaganda machine and surveillance tool and when you add that to the fact that it’s so convincingly sychophantic, that’s just outright dangerous.

It also cuts out meaningful context. A summary of what is likely already a summary isn’t helpful if you plan to actually use the information for anything. Part of how the human brain recalls information is by reading and understanding details and then parsing which details are important enough to remember for recall or use later vs which can be looked up.

Summarization by AI very often removes that context and because this is a learned skill that skill is likely to get rusty over time.

Additionally to that some people use it as a way to offload their need to think and that’s dangerous especially with the feedback loop of agreement it gives. I have seen several examples of AI sychophantically “agreeing” with whatever the user says regardless of whether it is factual or not.

And then of course there’s the fact that it’s often marketed as intelligent with the ability to tell truth, lies, and so on. That marketing isn’t true to how it actually works and what it actually does.

When you add to that the training data that stole the work of actual people, I just don’t see this as a cost the benefit ration where Generative AI can win.