This article seems more of an effort to self-promote using a recent topical warning from Elon Musk (and a number of other people involved with AI) as the excuse. Nothing in your article addresses man’s ability to program machines to cause harm. AI in all its forms, and it doesn’t have to be in any way sentient, can be harnessed for harmful purposes — regardless of the level of sophistication and learning it may possess.
Humans don’t need AI to make machines dangerous, so before shouting down Musk’s warnings, you should take a good hard look at firstly history to see how humans have always weaponised technology , and secondly what it would take to lose control, even temporarily, of a weaponised machine with the ability to continue on a runaway trail of harm and avoid or resist attempts to stop it.
That’s the point, rather than the one you address which is the futuristic sentient Asimov-type machine with the embedded Three Laws of Robotics, which most would agree is still a long way off.