Brewster Rockit by Tim Rickard for June 14, 2023

So things can get more bollixed up in the same amount of time.

To err is human to totally mess up takes an AI

Why waste time?

If machines destroy us, it won’t be because of their malevolence, but because of their incompetence.

AI is nothing more than computer memory being reinforced by successful trials. Optimization of a response. Just now the hardware is better at that algorithm. Beep.

Sometimes AIs make mistakes. Sometimes the “mistakes” are built-in by the programmers.

Jane Austen made the observation in Pride and Prejudice that people will forgive their own poor execution of a task if they can take pride in the speed in which they do it badly.

When you consider that an AI is trained on materials created by human, chosen by humans and presented to it by humans, is it any surprise that it does what humans do? Garbage in, garbage out.

Dr Mel has hit on the real problem with Artificial Intelligence. It lacks the delay in time to realize that an ERROR and might kill me. But to Artificial Intelligence killing a few or all Humans is not a real problem.

To err is human. To actually own up to it… Ai? Human? Neither???

“This post was not generated by AI!”

(◔˳ʖ◔)♫

Faster than politicians?!? I doubt that…

And more convincingly. AI chatbots give an initial disclaimer. After that it is free to authoritatively state the occasional bits of absolute rubbish. It will apologize politely for said rubbish, when corrected, but repeat the same rubbish if asked the next day.

Agatha Crum looking at computer printout remarks, “it would take a thousand accountants working non-stop for a thousand years to make an error of this big.”

That was a very human thing to say Oldbot.

Sometimes being/going slow is helpful – so that you can fix or undo things before they’re completely and/or irreversibly messed up.

And artificial intelligence carries with it presumed authority which must not be questioned.

Like destructive human beings, computers also lack morals. Except computers are much, much faster at it than we are.

Artificial Intelligence is Impossible:

The First Rule of Computer Programming proves it:

You Cannot Program a Computer To Do Anything You Do Not Know How To Do!