Bye Bye, AI

Even if you don’t think this will be your kind of post, please read it. It’s an important and timely topic.

In fact, it’s a lot more important than I used to think it was.

When artificial intelligence first became a buzzword a year or two ago, I rolled my eyes at both the lovers and the haters. My tendency to question both extremes of anything is something that usually serves me well. At first, it seemed like AI was no different than any other contentious issue, and maybe it wasn’t, for a while.

Up until a week ago, the only AI I’d used was the image generator that came with the platform I use to run this site. It is quite limited in capabilities and often very buggy. The results are laughable. If I tried to pass them off as real, you’d be justified in never reading my blog again.

Even the slightly better AI images I saw shared on Facebook were still clearly artificial. (Please, stop sharing these! They almost always come from scammy accounts trying to get the most likes with the least effort.) They fell into the “uncanny valley,” a term that refers to the way we instinctively know when faces or other images are artificially created.

When it came to writing, it was the same story. Everything I’d seen was stilted and clearly not written by a real person. I experimented with one of the blog topics this site suggested to me, and the result was ok, but very formulaic and awkward. Again, if I’d posted it, you all would’ve known something was up and be well within your rights to be disappointed in me.

So, I wasn’t overly concerned. The uncanny valley looked pretty deep. The people who thought AI was on the verge of replacing human creatives seemed to be donning their sackcloth out of season. No business that took itself seriously could trust AI to satisfy their audience. That was my assumption.

Then, I had a school assignment that required me to study and write about the differences between a memoir written by ChatGPT (one of the most popular AI programs) and one written by a person. The assignment promised me that the differences would be marked. ChatGPT would produce a dry list of facts only—something that would make even the worst Wikipedia articles look like Pulitzer Prize nominees.

I typed in the prompt, confident that the results would be exactly as predicted.

I don’t exaggerate by much when I say I got the biggest shock of my life.

ChatGPT’s memoir, far from being dull and stilted, was clear and even pleasant to read. It even employed many stylistic techniques that the original author had favored. At times, the AI was much more elegant and communicative than plenty of human writers I’ve read!

Suddenly, I regretted my complacency toward AI. Maybe the AI that came with my web host couldn’t take anyone’s job, but ChatGPT could. Or at east people without an understanding of the writing field might think it could.

How AI Works

Now, how did this happen? How has AI climbed out of the uncanny valley so quickly? (at least when it comes to writing. I still find most of the images easy to identify.)

Because that’s what it’s designed to do. AI started life as the creation of a computer programmer, just like your email inbox, minesweeper game, or turbo tax software. What makes AI different is that it “learns.” It updates and improves itself without requiring intervention from the original programmer.

The reason I specify that it’s without the programmer’s intervention is because AI does depend on someone to grow. It depends on us.

Every time someone uses AI or a program or website that gathers data for AI, they teach it something new. AI has gone from writing like a robot to writing like a talented author because it has gorged itself on the millions of billions of written words available to it online, including those written by professional authors. It becomes better at producing images as it scans all the photos and art available to it through social media, including those shared by professional artists.

This is why I think the common attitude that AI is “just a tool” or something to use “just for fun” is dangerous. As it stands, there is nothing in place to keep AI within those parameters. Even the most innocent use of AI is helping it become stronger and stronger, and that only makes it more tempting for those who aren’t so innocent.

The Dangers

Replacing creative jobs is the most obvious, immediate, and talked about threat posed by AI. Right now, creatives seem to be the only voices questioning the way we approach AI.

For good reason. If a business owner doesn’t understand or appreciate the irreplaceable traits that humans bring to art, writing, or music, they could find the significantly lower price tag of AI irresistible. If computer generated content seems good enough, why mess around with expensive and temperamental employees or freelancers?

There’s plenty of reasons, which I’ll lay out in the next section, but there’s another danger I want to address that isn’t talked about as much.

Many people who are not creative and don’t ever work with creatives may find it hard to care about AI if that’s the only area it threatens. I get it. I’m a big believer in focusing your time and energy on things that closely affect you and your loved ones first.

However, AI is one of those things, or at least has the potential to become one.

As AI’s ability to generate images, words, and even sound becomes more and more sophisticated, it can produce a lot more than just art. It could be used to simulate authority figures and convincingly disseminate bad information.

To be clear, I don’t have any particular political parties or entities in mind here. Passing what is fake off as real is dangerous on any level of public or private life.

Even if this seems like some sci-fi nightmare, far, far away on the horizon, we can’t ignore the fact that it is well within the realm of possibilities given the current trajectory of AI.

AI vs. Human Creators

I hope I’m stating something that is already obvious to my audience, but even the best computer-generated content still shouldn’t replace what humans can do. Whether you can point it out or not, there will always be something missing from anything produced by AI.

Part of the reason I never got into the AI controversy was because I adopted an argument I still hear from many people:

Art (be it pictures, stories, or songs) is inherently human. Without human emotion and creativity, it will always fall flat. Even if people try to replace artists, it won’t work for very long.

That’s true, but what my ChatGPT assignment made me think about is how many things we create that aren’t clearly art.

A piece of writing meant to simply inform or explain without significant emotional impact may not fall into the category of art. That leaves it unprotected if the only human trait we use to combat AI is emotion.

But humans are much more than emotion. There are many other things that give us the upper hand.

We can handle uncertainties and gaps in logic. We make decisions that aren’t strictly logical and see connections and patterns that don’t follow straight lines. We follow hunches and take risks.

AI can never truly learn any of that, although its ability to approximate it may become more and more alarming. And, as unfortunate and maddening as it is, there are a lot of people in this world willing to settle for that approximation if it makes their life easier or less expensive.

You may have noticed that I said some creative pieces are not art. What’s the difference? I think creativity simply refers to our ability to come up with something completely new. I don’t love the term “creative writing” because all writing that a person produces out of their own head, even if it’s well researched and strictly factual, requires a certain level of creativity.

AI cannot create in this way. I’ve been careful in this post to use words like “generate” or “produce” instead of “create” when referring to AI.

While human creatives are influenced by the work of others, they can reinterpret and repackage that information on a more fundamental level than a computer can. We are equipped with that mysterious CPU called “personality,” which is something inherent to each of us that no one else has a duplicate of. That’s the filter that allows us to turn the information we gather into something entirely new.

AI can only reorganize the same data it has been fed. While it can consume a nearly infinite amount of data, it can never create something it hasn’t been fed.

That raises a whole ‘nother ethical dilemma. AI is basically eating up everything we create and making it available to the whole world without credit or charge.

To put aside the philosophical and psychological side of this discussion, there is a simple, logical reason humans are a better choice than AI. Computers are not the target customers or clients of any business, real people are. So, it seems more than reasonable that real people would do a better job attracting and serving those clients and customers.

What Can We Do?

The more I research AI, the more passionate I become about stopping its incursion into every part of our lives and careers. At the same time, I also feel more and more helpless.

AI has become a kudzu vine.

Yet, if we all pick up a pair of shears and a bottle of RoudUp, we might stand a fighting chance, small though it may be.

Here’s what we should do:

• Resist the allure of AI. Unless something changes, using it simply as a tool or just for fun does nothing but increase its abilities in other areas.

• Opt-out of anything that gathers data for AI. This may be as simple as changing a few settings or as complex as switching to totally different software instead. Not to make a statement, but to protect your digital property from being sucked into the whirlpool.*

• Speak up. Many people and companies use AI because they don’t know the dangers.

• Support human creatives. It really means a lot to know we have your support. Even if you can’t see it, we do pour a lot of time and effort into our work. It’s a career just like any other.

AI depends on everyone who is using it. It doesn’t care what your intentions are. It is simply a program designed to chew up and spit out ever-growing amounts of data. Whether you use it to create cat memes or plans for world domination, the results are the same. It has learned a little bit more about humanity that it can use to imitate us.

*This is a challenge. And it's something I'm not able to draw a hard line on yet. I use Adobe, Instagram, Facebook, and even this very blog hosting site. All of these use AI, but there are really no alternatives. Right now, I'm just doing what I can to avoid directly using AI and keeping my eyes open for anything better that comes along.