This column was not written by AI…yet

The T-800 Terminator. HAL from 2001: A Space Odyssey.

Agent Smith from The Matrix. Ava in Ex Machina. The multitude of weird glue face robots that want to beat up Will Smith in I, Robot.

Film, and fiction in general, is littered with examples of computers who eventually outpace their human creators and thus apparently decide that we are no longer worth keeping around. Artificial intelligence exceeding our own.

The “A.I.” that is currently creeping into the public sphere is nowhere near the level of any of these fictional mechanical conquerors. In fact, I would greatly hesitate to deem it artificial intelligence at all. The programs now are still very much algorithms, only able to output answers based on previous data collected, rather than form a new thought. That being said, with programs like ChatGPT and others absorbing and utilizing the wealth of information available on the internet, there are times where it can certainly make a good emulation of actual artificial intelligence.

The progress in this sector over the last few years alone, at least from what we can see from a public standpoint, has been impressive.

I wouldn’t say I’m anywhere near being on the cutting edge of this technology, but I’ve certainly played around with it a bit recently. There are programs that can generate art, produce a five-paragraph essay on Isaac Newton’s three laws of motion, and some that can even simulate a conversation with another human being. I’ve got to say, I have been surprised at times and left to marvel at some of the results.

I think part of that is because a lot of this still has a long way to go. For example, a lot of art A.I. are terrible at drawing hands, sometimes adding an extra finger or mushing several together into some weird mitten-esque blob. The conversation bots would often get stuck in a loop, their somewhat limited memory resulting in them “forgetting” things already discussed. So when I finally came across something that was genuinely unexpected, but in a good way, I think much of that awe is a result from having to slog through so many less than stellar results to get to it. But compared to where it was even two years ago, the technology available to the public has drastically improved. Even in its current state, people are still able to do some incredible things with A.I., and we can already see some of the potential problems we will have to contend with cropping up. Students and even some online news outlets are using ChatGPT to write essays or articles for them. Creatives are left to wonder if their jobs will be replaced (one Netflix show already used A.I. generated background art for one of its new animated series). I bring up these things not because I don’t think we should stop developing A.I. but merely to note that we will need to examine how we wish to implement this technology, and do so sooner rather than later. We may not be at risk of being overrun by Schwarzeneggerlooking robots, but there are certainly issues to examine nonetheless.

A C ertain Point of V iew