If you ever want to make a tenuous connection between seemingly disparate elements in a world of loosely connected AI ideas, do this:
Watch the Boston Dynamics videos of the robot dog (or dogs tag-teaming the task, depending on which video you watch) opening a door without assistance on YouTube. Admire the tenacity of these guys (yes, I’m guilty of anthropomorphism, but it’s all I can do when it comes to reacting to anything robotic that has been purposefully designed to evoke emotional responses from human beings by mimicking physiology that is familiar to me).
Their focus, their determination, is impressive, even more so when being forcibly opposed by a human “handler” (no doubt a developer or engineer testing the integrity of the software). Closely study the complexity of moves involved in the robot achieving its goal of, first grabbing the door knob, then turning it, then pulling on the knob, then parking a foreleg in front of the door so that it will not again close while the dog’s mouth-like claw then swivels around and further opens the door so that the rest of its robot body can make its way through. If you’re not astounded, maybe a bit mortified, you should be.
Next, after watching these disturbing and intriguing videos, read through the comments…all of them…and see what your responses are to the responses. How many comments sympathize with the robot, how many with the human? Are you surprised? Terrified by what you read? Fascinated? Is it a mixed bag of dread and fear that a robot apocalypse may be coming coupled with longing and hope that maybe everything will turn out okay?
Now go to Wikipedia’s article entitled Existential risk from artificial general intelligence and carefully read through the warnings. There are many. As with the comments that accompany the YouTube videos, also comb through the references this article has used to build its content. It is also impressive.
It’s not just you. Many are kept awake at night by the scenarios that abound, some of which will surely come to pass.
Now, go find an engineer involved with AI (any engineer will do) and have them look you in the eye while reassuring you that there’s nothing for us to worry about. Make sure he doesn’t blink in his response.