What’s my beef with AI?

I’ve spent my entire 30+ career working in Information Technology. This was born out of a love of computers when I was a kid. Not so much as a “computer science” type of thing, but because I like gadgets and fun and useful tools. I was tinkering with things like old-school BBS systems, building my own computers, and learning to program, all at a very young age. So despite my other leanings towards professional ministry, I ended up in a career using technology.

The timing was opportune since I started my career right as the internet was beginning to develop. I was able to take my inquisitive nature and parlay it into being the most useful person in the room because I could create websites, and eventually web applications. My career eventually progressed into management in the last decade, but my love of gadgets and tinkering has never gone away.

One could be forgiven for thinking that this means I would be all in on Artificial Intelligence, but in fact the opposite is true. I’ve actually become quite critical of the technology and more specifically the ways that it is being applied. There are many things that I genuinely like about the technology. As well as areas where I’m really excited for what it can bring to the table to help human beings. But so much of what comes with the technology also makes me feel everything from frustration to confusion to sadness.

Let’s start with what I like. Most people don’t realize that tools in the AI space have been around for a long time. Machine Learning is a technology that has existed for decades, going all the way back to the earliest modern computers. Computers are great when it comes to math and numbers, and using the power of a computer to analyze massive sets of data has been a huge boon for all technology. Even in my current job, my employer has multiple systems that have been using machine learning for many years to help do predictions that would take a human beings orders of magnitude longer to do. That’s a really great use of a type of AI.

Similarly the use of Large Language Models to extrapolate trends and patterns in large quantities of textual material is amazing. The advent of the internet means that humans have access to pretty much the entire corpus of human knowledge. Trying to dissect and distill that information into something useful is a great use of a tool like LLMs. I know there are many projects out there that are analyzing textual reports from field offices and technicians and then alerting people when too many trends are detected in their service notes. This can help with proactive troubleshooting, and potentially save a lot of time and money (and maybe even lives).

Even using LLMs to backend web searches makes some sense as trying to sort through all the junk out there online can be daunting. I think we’re slowly learning how best to implement some of these tools, and how best to use them to surface links to deeper information. I do worry that too many people rely solely on the quick summaries that AI provides on tools like Google Search, but they are at least trying to show the chain of custody of the information and hopefully that will improve over time.

I don’t even object to the occasional hallucination. Humans should always be questioning their tools, that’s something that’s been true since the first computers. No one should be making life-changing decisions without actually digging into what they’re trying to learn about. Verify, verify, verify is the mantra whenever it comes to AI.

I don’t even object that much to using AI for coding. Having been a software developer for many years having a tool that can handle all of the repetitive tasks that I would spend hours copying and pasting is awesome. Coding languages are just ways that we communicate with computers to tell them what to do. Using natural language is just the next step in the evolution of programming languages.

I can be content with all of this. Maybe toss in a personal assistant tool that can help organize a busy schedule and you’ve got computers that are supporting and helping humans as we do human things. Where I end up bristling is when we start talking about generative AI.

There are certain things that I believe are core to the human experience. Emotions like love, anger, and sadness. Silliness and laughter. Even self-contemplation about our place in the universe and deep philosophical questions of life. But perhaps one of the greatest human traits is the ability to create, and specifically create things of beauty.

Art is as old as humankind. Just a few hours north of me are pictographs that are thousands of years old. Drawings on rock using primitive inks and tools, expressing an image or thought that someone wanted to share with the world. Even when the brutality of a hunter/gatherer existence demanded constant vigilance and work, people took time to create. It’s a truth of human existence that has permeated every people and culture in our history.

Now though we’re bombarded with “art” that has been generated by some chatbot, complete with all of the telltale stylings and feel of some “thing” trying to will it into existence… just a bit too hard. As I browse social media I’m bombarded with images, videos, and songs that are devoid of the spark of human creativity. Human faces that are just a little too real. Posters that all look the same with far too much busy-ness and no thoughtful layout. Essays and stories that are bland and repetitive. Music that sounds too perfect until you listen closely and realize how flawed it truly is. It’s maddening to see so many places deciding that relying on this type of creativity is easier than actually working to create something that is true and real to the human experience.

Perhaps the biggest issue I have is that all of this generative AI is built upon a foundation of theft. AI cannot create the way that a human does and can only generate art that is a mush of what real humans have been creating for millennia. It can’t come up with new ideas, or develop innovative ways of putting different shapes, colors, words, and sounds together. It can only take.

One might say, “But isn’t all art derivative? Doesn’t every artist just learn and copy from those that came before?” That’s certainly true, and every artist who is looking to grow in their craft will constantly be exposing themselves to the art of others. But how artists create new work from this history is more than just a regurgitation. It’s a blending of the life experiences of multiple people. An interplay of the souls of two unique human creatures that will never exist again… ever. That’s something that no computer can ever duplicate. There is no depth of soul, expression of passion, delight and whimsy, or weight of sadness in what a computer creates. It’s simply the combination of 1’s and 0’s that seem to make the most sense based on all the work that has been fed into its training data.

Perhaps I’m just too deeply embedded in this. As someone who is a photographer and writer, seeing people constantly post fake images and rambling social media essays really rubs me the wrong way. I want to scream “If you want photos just text me! Maybe I can come help you with some images that aren’t stolen from folks like myself who spent a day in the woods getting the best shots we could.” And I certainly understand that for many people taking the time to properly create media assets, such as hiring a designer and paying for their time, etc., feels like it’s too much hassle when their purpose isn’t to create “art” in the first place. But the more we normalize that behavior, what are we doing to the future of creation?

I’ve been spending a lot of time lately visiting a dear friend at his studio in a building dedicated to artists. This is a place dedicated to the pursuit of the uniquely human endeavor of creation. People pouring their passion, hope, despair, and whimsy into something that will hopefully bring joy to others and outlive their time on this earth. They don’t deserve to have their work stolen and amalgamated into a soul-less system of data bits and algorithms, only to have something new and dispassionate spit out for anyone who can fire up a ChatGPT chat window.

So ya… that’s my beef with AI. I could also go into all the issues with constant data center development, water usage, and other environmental concerns. But this essay is already too long. What I want to leave people with is the truth that AI is a powerful and amazing tool that can help/assist/enable humans to live life, learn, and be more productive. But let’s stop trying to ask AI to replace what makes us uniquely human.

I would bet that you have people in your life that are more creative than you. Even if we’re creative people ourselves, there’s always people around us that are creative in different ways (or are on a different level than we are). Take a moment and think about who they are as a person and how it gets expressed in what they create. Then think about how there are billions of people in the world creating something new every single day.

There’s more than enough creativity in the world to go around, we don’t need a silicon god to be a part of that.

Jamison's avatar
Jamison

Adventures in the second half of life

Leave a Reply