Where I'm Landing on AI

A House Finch taken with the Olympus E-M10 and 75-300mm II

At this point, opinions about AI¹ are many, ranging from hype-trains that boast a utopic future where we can all pursue exactly what we want to while machines do all the dirty work, to stoic refusal to use these tools for any reason, to vehement hatred, to doomers who claim AI will bring about the end of humanity. Needless to say, no one seems to agree on what exactly AI is good for or what it will mean for the human species. And so, I want to make a point that everything I say below are my own opinions. We all need to make our own, informed decisions with this stuff, and I am by no means saying I have the right view on any of this. And, of course–if you’re anything like me–those views are still changing.

Landing, Not Landed

Why do I say landing, and not landed? AI is a relatively new phenomenon in tech. A lot of people have developed very strong views very quickly and, I think possibly out of a sense of “duty,” feel the need to stick to those views. I am not so hard-headed. I know that any new technology will have positives and negatives, and that those positives and negatives are always changing.

Hence, landing, not landed.

The Good Parts

What is good or bad about AI will very likely depend on what you need and what you value. But, there are some uses and tools that I think are, objectively, inarguably, mostly² good.

The main areas which I think fit that bill are in science and medicine. If a technology can help people make discoveries and advancements that make people healthier and solve real problems, then I think we need to be open to the idea that the positives of these use-cases may outweigh the negatives.

AI has already shown positive progress at helping doctors diagnose patients more effectively. Even AI, a technology highly criticized for its energy and water usage³, will likely help immensely in fighting climate change.

But what about less clearly positive uses? How do I use AI day-to-day?

What I Actually Use

First off, let me clarify, right away, what I don’t use: Generative AI to fully generate “unique” and complete audio, images, video, text, or any other medium AI can currently generate. I’ll talk more about why later on.

So, here’s a quick list of what I do use, and why.

1. Google Search Overview - I know, I know. A lot of you are shaking your heads. I’d also bet that a lot of you also, when you’re in a hurry or just don’t feel like doing more work, use them, too.
2. Generative Expand in Photoshop - Probably the most generative AI tool I use is Generative Expand in Adobe Photoshop. Look, I’m a web developer, and sometimes the photos we get from clients or Adobe Stock are… Well, let’s just say they don’t quite fit. Generative Expand has been extremely helpful in getting that awkward portrait crop to fill a 3:2 space when needed.
3. Gemini Pro Code - I do use code generation from Gemini. It can be pretty useful in two scenarios.
   1. When I know how to write something, but know that it would be faster to tell Gemini to do it than to type it out myself. The bonus here is I know I can go back, look over the generated code, and fix any errors easily.
   2. When I don’t quite know the syntax, but know what I need the code to do. The important thing to remember here is, if you don’t know enough to debug the code or find insecure code, make sure you’re learning what each part of the generated code is doing and look up best practices for using it. AI coding tools can and do create vulnerabilities.⁴
4. The hidden stuff - I’m sure many of us are using AI or AI-adjacent tools almost daily without even knowing. That’s my favorite kind, if I’m being honest. To me that says, “this tool is good enough at what it does that you don’t even realize it’s AI, and we didn’t feel the need to tell you.”

The Bad Parts

I’ll try to keep these brief. So much could be written about what’s bad about AI.

All Your Work Are Belong To Us

It’s no secret that AI companies gobble up everyone’s data, regardless of legality, and regardless of whether or not that’s actually what people want. The common argument for it tends to be something along the lines of, “Well, it’s already out there on the internet, so it’s free game.” But, even they (sometimes deep deep deep down) know that argument is lazy and in bad faith.

The other argument I hear is that, without all that data, without as much data as these models can possibly ingest, they’ll stagnate, or at least improve at a significantly slower pace. My gut reaction to that is, tough luck, just because you’re the new cool doesn’t mean you get whatever you want. However, politics has a way of making things messy.

You see, because of how extremely accessible AI tools are, these AI companies don’t just represent themselves, they represent the technological advancement of the country (or countries) they hail from. And this provides an opportunity. If the AI companies of other countries exceed the capabilities of those of your own, you could argue they pose a threat. To ensure your own safety, your only choice is to ease restrictions in your own country to allow your own AI companies to out-improve the competition. To the government, this is a win-win. By prioritizing the success of the AI company not only does the company win, but you’ve opened opportunities for you to increase your own power of your own citizens. It’s a fabricated technological arms race that, once set into motion, becomes ever more difficult to argue against. Sounds pretty extreme? It very well could be. And it’s already happening.

Forced Usage

As with many new technologies, those who are seen as able to benefit by said new technology are often forced to. “Keep up or you’ll be left behind,” they say. Sadly, that often seems to be true, just maybe not for the reasons the phrase implies.

A massive number of people were laid off in 2025, in no small part due to AI⁵. But, is that paying off? Microsoft’s Satya Nadella bragged that as much as 30% of their code is being written by AI while Windows 11–possibly Microsoft’s least popular release since Vista–sees an increase of critical bugs and poorly implemented “features,” earning them the nickname Microslop. Lawyers are using it and paying for it. Many companies that do try to replace their workforce with AI regret it and struggle to fill the gaps, outsourcing the work when they can’t afford to hire people back. This has led to an increasingly hostile job market for recent grads or those with little experience.

As with many new technologies, we try to run before we can crawl.

Energy, Water, and the Environment

I mentioned earlier that this can be a contentious debate, but that’s only if you give AI the benefit of the doubt. You know what’s less than 10x water usage or x parts per million pollutants emitted or y wattage used? None. If we all just took a minute and said, “Wait, is this something we need? If not, could we actually be better off without it?” Maybe, if we had asked those questions and come to the reasonable conclusion that, no, we don’t, and yes, we could, AI would be increasing our water usage, our energy usage, our pollution rate exactly zero times more than it was before. Pretty simple, really.

Unfortunately, powerful, wealthy people either didn’t ask those questions or–possibly more horrifyingly–did and decided to do the thing anyways.

How We Learn

This is a big one.

When I was in school, Wikipedia was still up-and-coming and was strictly forbidden. All research had to be cited from published books. That meant either reading enough on the internet to copy the sources of existing research, or going to the library and reading and citing books yourself. This did two things: 1. It taught us how to write without the computer doing it for us, and 2. It taught us how to learn. Not memorize, but learn. It did those things by adding friction.

It’s no secret that literacy in the US is falling. Fewer people can read and fewer people can write. This isn’t solely due to AI, but it’s not going to help. I’d argue that’s because AI doesn’t just change or reduce friction in completing assignments… It removes it completely.

Even at the height of the internet, but before AI was accessible to anyone with a phone or a browser, if you wanted to write and research something, you still had to do the important bits yourself. You had to click the keys on the keyboard. You had to pick the links on Google to click and read and examine. You had to format the words on a page the right way, in the right order. You had to convey meaning and emotion and structure. Much like I’m doing now, writing this post.

But with AI, it’s different. It’s not just a tool to help you type faster, or fix grammatical errors. It’s not just a quicker way to find the right book to reference or article to skim. It’s not even watching a YouTube video so someone can explain something to you with dopamine-approved visuals. You give it some quick direction on what you need, and it poops it right out, references⁶ and all. No friction needed.

How We Create

As a photographer, web designer and developer, and general art appreciator, the idea of a glorified algorithm making art is simply laughable. Art is a human expression, and generative AI simply is not, and never will be, art. Full stop⁷.

But, unfortunately, it is effecting how and when we create. Fields that used to celebrate creative thinking and ideation now increasingly prioritize productivity and efficiency. I’ve seen this in my own job, and I’m positive I’m not alone. I’ve also seen quite a few people say that generative AI has allowed them to be more creative, where they want to be. Certainly, these tools can be used smartly. But, I have to wonder what, in that process, has been lost. Are they becoming better artists, or just more productive people?

A Quote

We’ve all heard it, “Be the change you want to see in the world.” We know what it means, and I think it’s something to keep in mind when you’re deciding what AI tools to use. But, while that quote implies something actively done, I think it’s just as important to consider something actively not done. When a new tool pops up or a friend or coworker is pushing their new favorite generative AI, just pause a moment and think, “Do I like a world where people use this thing to do this thing? Do I really want this to be how it is, potentially forever?”