Good guy AI?

Illustration source: https://www.memedroid.com/memes/tag/terminator

Artificial stupidity (AS) may be defined as the attempt by computer scientists to create computer programs capable of causing problems of a type normally associated with human thought. Wallace Marshall (as quoted by Ray Kurzweil)

Should we as a species be more optimistic or fearful about the rapid evolution of Artificial Intelligence? 

On one of our weekly therapy sessions (AKA hike), a friend and I reflected on how robots and androids and computer programs are portrayed in the movies. It was pretty easy to come up with a list of villainous technologies in films like 2001: A Space Odyssey; Terminator; RoboCop; The Matrix; and Blade Runner. 

But do a search for movies in which tech plays the good guy and the going gets a little tougher. We all love C3PO in Star Wars, the titular character in Wall-E, and Data in Star Trek. Many science fiction movies have robots as supporting characters - often in comedic roles. An interesting character in terms of good or bad is the Samantha character in Her. Perhaps what makes Samantha hard to categorize so simply is that she really does seem emotionally human. But a program as fully good as the Terminator is evil? Good luck finding one.

The media has set us up to be highly suspicious of machines who are as smart or smarter than we are.

I’ve been re-reading Ray Kurzweil’s prophetic book The Age of Spiritual Machines from 1998. Kurzweil begins the book by asking readers to tackle some pretty tough questions about what the press is now bandying about rather flippantly: How do we know we exist? What is consciousness exactly and do other species have it? How does the development of machine intelligence fit into the overall design of evolution?

Less existentially, he demands, we should simply start with the question “How do we actually define intelligence?” That’s more complicated than one might initially think.

Now defining stupidity is not that hard. I’ve completed that task a number of times (See Seven stupid mistakes teacher make with technology), “I use stupid under fairly constrained conditions. To me, a stupid act has a degree of willfulness about it and is serious. Making an error once is ignorance; making the same mistake multiple times is stupidity.” In short, ignorance is doing the wrong thing due to lack of knowledge; stupidity is having the knowledge, but doing the wrong thing anyway.

It may well be that in order for ChatGPT and other AIs to fully be taken as human, there may be a need to build in AS (as Marshall defines in the opening quote of this post). Until then, you can be assured that the Blue Skunk is fully human-produced - due not to its content’s brilliance, but to the dumb ideas it contains. 

The next iterations of AI should be interesting, indeed.