Just as dumb, that is.

Someone just did, for the sake of demonstrating the concept, something I wrote about years ago, that is transform a robotic dog into a guardian.

A guardian that shocks experts because it is armed with a sniper rifle:

Some say that “this crosses a moral, legal and technical line, taking us to a dark and dangerous world." An expert on autonomous weapon systems, instead, was much less pessimistic (emphasis mine):

“Given that the system is controlled by humans, and other countries have developed armed, remotely-piloted robotic ground systems, I am not sure this system raises any new ethical questions beyond those raised by [military drones]"

The weak spot? Just ignore the four legs

Let’s look again at that reasoning. The part that this “dog” isn’t more or less ethical than drones that can kill people is almost flawless. Assuming that even these “dogs” are only accessible to the military and the police, of course.

The "controlled by humans” part, instead, is the weak spot.

If it is controlled, everything that a human can control remotely with digital technology, another human can hack, sooner or later. Before selling this to SWAT teams, please make sure they would be happy if their opponents got control of the robodogs remotes.

Making the robodogs autonomous would just make thing worse. Because a robodog with a sniper rifle attached would be nothing else that a four-legged version of what “experts” called “smart guns” two years ago, and as such would have exactly the same problems: in a nutshell, it could not work, and would be a cretin idea even if it did. See why here.