> My guess is that developing a new lethal virus or bacteria that is very infectious, is the easiest way, but even that requires a lot of high tech out of reach of most people.

You can do at home gene editing with open source software and have it synthesized into a bacteria for the cost of a nice meal for two (under $100), or viral vector for less than $500. That's in reach of anyone that can snatch a purse.

The super intelligent AI wouldn't ask you to do anything. It would find the right people and trick them into doing the necessary steps, whatever they are.

The fundamental difference between people who are worried about AI, and people who aren't:

* People who aren't worried will think for 30 seconds, then implicitly assume that the list of options they thought of in 30 seconds is comprehensive, and that they are well-informed about each of those options.

* People who are worried realize that a supersmart AI will think 10-100x faster. A dedicated schizophrenic could have it run for a good long while and generate a huge menu of options which humans never anticipated.

This is related to the concept of "What you see is all there is" (WYSIATI) in cognitive psychology. Our brains tend to quickly build a simple story out of the info we have available. Accounting for info which isn't available is much harder. I recommend the book Thinking Fast and Slow if you want to learn more.

You can't design an infectious pathogen without testing it.

It's got all the same problems as the concept of a dirty bomb did, only worse (dirty bombs aren't practical because handling highly radioactive materials en masse is both highly visible and will kill anyone trying to do it without the money and facilities).