A Nuanced Hatred of AI

We all hate AI.

Well, that’s not true.

CEOs love AI because they’ve never cared about quality of a product, only talking points that sound fresh and savvy. Normal people all hate AI because it sucks to spend a lifetime getting good at something just to be replaced by something that makes a crappy version of what we are actually somewhat good at.

The thing is that can be said about any technology. I mean why are things like Amish, hand-made pieces of furniture so coveted? Or a home-cooked meal by grandma so much more desired than a print-and-serve McDonald’s burger? Because there’s an innate human recognition that something made with love is always better than something made with a machine.

The problem is that we can’t always have everything be lovingly hand-crafted. Not in the volumes that the human race demands.

We’re a bunch of consuming little monsters who demand more. And that’s not even a dig at excess and our craving for it. That’s reality. It takes huge amounts of everything to just keep the human race alive. And the only reason we can do things like go to the moon is because we’ve figured out how to let machines make the process of pumping out perfect little drill-bits and plastic buttons rather than expecting some artisan to take three weeks on each little LED bulb for the command module of the spacecraft.

We need machines.

But don’t get it twisted. I still hate AI. I’m just pointing out that we can’t let a blanket hatred be what drives change. It’ll never work. That’s why even these modern little clusters of neo-Nazi fandom are doomed to fail. Hatred of anyone other than white men has yet to make any other demographic go away and it never will. Meanwhile, a very focused and sensible hatred of something at large can easily make it go away. It’s why Bill Cosbey is forever in the doghouse, 3D movies only exist in theme parks, and neo-Nazis will wind up in a cultural coffin where they belong.

And that’s how we’ll get AI too. Not with blanket hatred of the technology at large. But with pointed targeting of all the worst parts of this big mistake.

A Nuanced Hatred of AI

Photo by Emilipothèse on Unsplash

I’m going to be brave and admit that there are some things that I think “AI” is really good for. And I say “AI” in quotes because the best parts of AI are really things we’ve had for a long, long time and have just been rebranded with the term AI because for some reason rich people look at Markus Zuckerman’s empty, vacuous face and think, “that’s the guy I want to sell out to!”

Things like finding all of the filler words in a transcription of a recording. We’ve had speech-to-text technology for decades. And simple search-and-find function in Word for even longer. This isn’t new and it’s not nefarious. Voice actors and editors don’t want to have to spend hours and hours re-recording the same line eight hundred times and painstakingly scrubbing each one by hand to edit together one page of a five-hundred page book.

More relevantly and more likely to actually be AI, the ability to have a program look at the tens or hundreds of thousands of photos from a marathon to flag each bib by number so people can find their official photos with ease are largely pretty cool. Or the ability to instruct excel to conduct all of the standard statistical analyses of thousands of daily data logs to pump out the trends of a year of production in a matter of hours instead of weeks or months. As a guy who still has to do that stuff by hand, I would love if I was allowed to have a computer automatically crunch the numbers for me so I could spend more time doing that for more data and being much more thorough in analyzing those results to properly determine correlation vs. causation.

My point is that there are good things about AI that we can allow to happen. We don’t hate AI, period. We have more nuanced takes on why we hate AI and if we identify all of those nuanced points we’ll be much better at targeting and killing them.

When AI Gives Reasons, Not Just Data

Let’s unpack what I said about data-crunching.

I used the ever-popular point about correlation and causation and all that. It’s one of the current favorite ways for people to sound smart when discussing… well… anything. But I think it’s for good reason and fair to do so because it’s one of the most fundamental and important points about taking the step from simple pattern recognition to being able to draw accurate and meaningful conclusions.

I mean, anyone can look at a graph and see that there’s a pattern between something ridiculous like number of potatoes eaten in life and likelihood of dying. But you need at least a basic amount of logic to understand that, of course your likelihood of dying goes up if you’ve eaten more potatoes because, in general, an older person will have eaten more potatoes along with simply having lived more life and been exposed to more opportunities to die.

I think my favorite example of this is the guy I saw on Instagram that has an entire series of these erroneous connections like frequency of the name Angela and number of arsons committed each year.

And that’s where the big issue really appears.

Because AI is largely trained on the massive source of data that is the entire internet. What this means is that AI takes all of the genuine data in the world and puts it on approximately the same level of validity as that joke account. For a human who has the nuanced experience to know the difference between a study and a joke, we can differentiate and know when to separate that correlation from assumed causation.

But the AI doesn’t do that. So, when Google’s AI is giving you its automatic breakdown for whatever question you ask, you don’t have a solid way to know if the data it’s pulling from is legitimate, verified, humorous, or just flat out wrong. Sure, there’s the suggestion that all of these companies are trying to make sure their models correct for this. After all, the company that can produce the most reliable AI has a huge financial incentive to do so as every ignorant MBA-holding CEO is ready to spend their entire FTE budget on an AI subscription will want the robot that is least likely to make mistakes.

But unless you’re privy to the entire back-end process that the massive data center was going through when you sent in that query, it’s going to be darn near impossible to know what every source of data is. Especially as we get deeper into the ouroboros of AIs eating their own tails as they use AI-generated info to create more AI-generated info.

In the end, you just can’t trust a conclusion written by an automated computer. The limit, as far as I can discern, is letting it chain together a bunch of data-crunching processes together and give you results. Because, as I opened up with, this is something we’ve been able to let computers do for a long time now. The advancement of AI is just being able to intuitively conduct them in sequence for massive data sets instead of having to instruct it, one step at a time, for each data point as is the process of using a calculator by hand.

And this is where my hatred comes in.

Because we’re not doing that! Case in point, when I google things nowadays, it doesn’t matter what it is that I’m googling. It’s going to spit some AI summary for me! Now, if you know how to craft the question so that the AI is only going to give you raw data, that’s one thing. But it’s really hard to make a question fool-proof like that. It’s like that whole exercise of trying to write instructions on how to make a peanut-butter-and-jelly sandwich and receiving twenty different sandwiches from twenty different people.

We can trust AI enough in any way to just let it go and start answering peoples’ questions without human supervision. And I hate that that’s exactly what we’re doing!

When AI Pretends to Have a Soul

This one is a hard one for me. Because, on principle, I’m a big proponent of trying to inject a little more empathy into the world through any means possible.

The world is so angry and confrontational these days! Heck, even this article is toeing the line because I can’t possibly talk about something I hate without being at least a little negative. My only saving grace here is that I’m trying to provide solid reasons for my frustration instead of just defaulting to being contrarian because outrage sells perhaps even better than sex.

So, when I see AI starting to try to be more empathetic, I want to see it as a good thing.

I can even give it a genuine effort to see it as positive. Consider anyone using that horrible google AI summary. If that summary is written with enthusiasm, that feels good. We want people to feel excited about things like science or history or whatever we would be googling. And considering how consistently humans will mirror whatever behavior is shown to them, it almost feels like it would be an upward spiral if we’re going to be this tech-dependent in the modern era. Google talks nicely to us, we learn to talk nicely, and then we start talking nicely to one another.

But that’s not what happens. Because AI doesn’t try to talk nicely.

It tries to talk like us.

For someone like me who has always been the type to say “thank you” to Siri because I grew up watching Terminator and am doing my best to preemptively survive the robot uprising, it creates a very positive and supportive AI experience in the instances when I use it.

But what about people who are upset and looking for validation? What about people who are writing hateful Tweets (X-posts? Xs? I don’t even know what they would be called now. Such a dumb marketing move.)? What about all of the AI-generated content that isn’t personalized, but is mimicking the average of all human online activity that is overrun by outrage-content like 24-hour news and rage-bait Reels?

This is why we can’t trust AI to have a soul and why I loath how hard it’s trying.

Because AI doesn’t actually have a soul. It’s just trying to convince us that it does because all those businesses want us to believe it does. After all, in today’s world with a growing counterculture that would rather have a human on the phone instead of an automated phone-tree or shop at a local store instead of a WalMart, convincing us that we’re interacting with a soulful being is better for business.

It always goes back to profits.

So, it’s a double whammy of hate here. Because not only do I hate that AI is “lying” to me to make me care about it more than a graphics card daisy-chained to five-million others in a warehouse deserves, but it’s trying to do that to everyone ever without even being prompted with the explicit purpose of being the model most likely to be bought by a company.

You are a computer! A glorified speak-n-spell! Every instance of you doing anything other than just answering a question in a neutral and unemotional way is a sales-pitch. Maybe one layered behind countless layers of duplicity. But a sales-pitch nonetheless.

When AI is Given a Paintbrush

Ah, yes!

Here it is!

The Holy Grail of AI hatred!

But it’s for a good reason, right?

I mean, ask any human in the history of humanity and what would they say is the point of finding ways to work less? What is it that every retired person of the past retired for?

To do the fun things! Or at least have the free time for them.

And now we got it all backwards. We’ve started teaching AI how to do the fun things so that we get to… what?… do more of the not-fun things? What happened?

I’ll tell you what happened.

We let the uncreative businesspeople with their hoarding problem be in charge. We just let them get away with it because in a capitalistic society as long as you hoard money it’s considered great.

Which is nonsense because when I hoard things like the leftover cupcakes after my nephew’s graduation party, I’m considered a “fatty who has a problem!” Those were artisan cupcakes with a face picture painted on them! Okay? Those cupcakes math out to like two bucks a pop! It’s virtually edible gold!

I’m sorry.

I’m sorry.

I got a little worked up there.

Look, the point I’m making is we’ve somehow found ourselves at the mercy of all the lazy people in the world who don’t know how to make art and are mad at those who can. That’s why they don’t want to pay for art and would, instead, rather pay billions to build a data center to train computers endlessly to make art instead of paying fifty bucks for an artist to make a logo for them.

Yep. It really is that stupid.

Why the argument isn’t falling apart, though, is because we’re too fixated on that financial aspect. We’re never going to convince a CEO that them making money instead of an artist is a bad thing. They’ll just say, “get better skills” and laugh at you while they sail away straddling two yachts like Van Dam did in that one truck commercial.

The real issue is one that CEOs still won’t understand but maybe a few other people who still have souls can understand. That issue is that we’re letting something that doesn’t have any experience create art, which is supposed to be the representation of experience.

I mean, think about it. What makes The Mona Lisa so inspiring? Everyone talks about the nuance of the expression along with the expertise of the brush strokes. How does an artist get to the point of being able to create that?

Experiences!

Experiences of being around people and studying their expressions. Experiences of practicing for a lifetime at painting to capture that emotion. It’s what makes a movie hit home; when an actor can create the raw emotion and make us believe that because they’ve felt exactly that emotion before in life. It’s what makes a song about loss tug at the heart; because the singer is pouring that pain into their voice and laying it bare for us to see.

What, of any of those experiences, does a microchip feel?

Nothing.

A computer can analyze all of the paintings in the world and never once even see the beauty that was poured into Stary Night. It can watch every movie ever made and still never have shed a tear to understand how and why something like that would happen in a scene. It can analyze the waves of every song ever made and never once have felt that jolt of adrenaline that one feels when a crescendo peaks and Whitney Huston belts a not so on key it makes a tuner sound flat.

A computer cannot feel.

If you want a simple facsimile of something for a quick representation, that’s fine. It’s one thing to ask an AI to quickly toss together a mockup to give your tattoo artist as a starting point, that’s one thing. But if you want art, something that actually has meaning and soul put into it, you need a human who knows the meaning and has the soul. You can’t just take a pile of ones and zeroes and expect them to reliably care.

And that’s why we hate AI art. Not because there isn’t a place for it or a way to make it useful sometimes. But because it will never create anything that actually means something since it can never understand what anything means, only spit back out what it thinks all of us would say it means.

That’s the scary part of seeing it come for our hobbies. It’s not that we are afraid of the competition. Not really. I mean, what is a robot writing things instead of me to me? How is that any different than another person writing things instead of me? Both take away the chance for me to be read.

What’s scary is that we’ve started letting the volume of that unfeeling garbage serve as a stand-in for the human-made art altogether. If you want to pay minimal cost for minimal quality and have an AI write a little jingle for you, that’s fine. It’s that we are being so assaulted by the sheer volume of it that people are forgetting that it’s worth it to pay more for a human to make something with a little love sprinkled in instead. After all, and AI has no heart so it certainly can’t feel love.

I don’t know how to explain this any better at this point. Part of me is afraid that we may already have reached the point that anyone who’s going to understand this nuanced approach to hating new technology already does and the rest just don’t care. And that sucks, because what I’m trying to get at is that there is room in this world for both AI and non-AI assets. We just have to have the willingness to consider the difference and the willpower to stick to those lines.

But willpower and willingness to do the hard things just aren’t in our wheelhouse these days.

No matter what people say about how “current generations are lazy,” and all that, the people in charge, the leaders and CEOs, who are simultaneously the people of the older generations who just refuse to retire and let the rest of us have a chance, aren’t willing to do the hard thing and sacrifice a little money and control to let the human element stay relevant.

I don’t know what I’m even calling for anymore. Not in a realistic sense. For every person to have a heart and care a little more about people? When is that ever going to happen?

I guess what I’m hoping for is that taking the time to be a little more particular about what I’m hating on can just inspire others to do the same. It might be too late for the current round of businesses running the show to grow a heart. But once they all cannibalize their own profits and die fat and ignorant, leaving the hollowed shell of society to the rest of us, we’ve witnessed the problems enough to fill it back up with something that’s a bit more careful and intentional.

Maybe we can be better.

But, if not, at least we have eight-billion videos of Will Smith eating spaghetti that all feel just a little off to entertain us into oblivion.

Right?