I was hopping on here to see if I could find an old blog where I listed off game ideas I once had and figured I'd post something to catch you guys up on what I've been up to in the past 7 years but also to post my thoughts on the current AI situation. I don't know any programmers IRL to talk to about it and it would be nice to hear some thoughts that aren't just ridiculously pro or anti AI. Because realistically, it's never going away but it also has many problems so neither side of the extreme opinion is very helpful on how to navigate the issue of AI for a normal everyday person trying to navigate a changing world.

First up, I'm still posting the occasional YouTube video on my channels, though real life issues keep getting in the way or ruining my motivation so the channels never been consistent enough to go anywhere. https://www.youtube.com/@GamePlague. I also have a bluesky since I stopped using twitter a couple years ago: https://bsky.app/profile/kabob.pycosites.com

I'm working on a few game projects right now but like a lot of people I'm struggling with how to balance actually doing my work myself vs the hope of ever accomplishing anything in a world where AI will soon be able to create a better game than 1 year of me working full time but done in a single day for a few dollars of tokens.

I hadn't been paying too much attention to how far coding AI had come until a few months ago but since then I've experimented with Claude Code, Codex and Gemini to get a feel for where they're at and I've seen enough to know that it's probably coming sooner than we'd all expected. Just like how image/music generation felt like they came out of nowhere because they progressed so fast from niche experiments to mainstream usage.

Like look at this, I heard the newest Claude release was really good at 3d modelling so to see if it was true I had it make a 3d model for some Pyco Mundi creatures that I'd been struggling to realize visually for years, it took a matter of minutes and a few cents of tokens to make this one: https://claude.ai/artifact/FX2iaw6x3vfmB73xBaczpb. I could probably spend a few hundred dollars and make the game of my dreams that I always assumed I'd need to get rich and pay a team to make. Right now my plan is that someday maybe I can pay an artist to make proper illustrations/sprites using that model as a better reference than my attempts to explain it using words could have ever been. But realistically I know I've been waiting 10+ years for that day to come and there's nothing right now to suggest it ever will. The one time I tried to commission some art the person ignored me, now I can't afford to.

And really, game design for me has always been more about the design than doing the actual implementation of the idea. I'll world build every single little detail and love every second of it but actually making even a tech demo requires so much time and motivation and then on top of that there's always the real doubt of my ability to ever afford hiring someone to do art/audio. I like programming but I've spent 20+ years programming and have very little to show for it.

So it sucks that this thing exists that I could use to realize all my dreams in a realistic time frame but it's also the thing that is going to destroy the game industry that I had always hoped to one day join. Plus the high risk of reputational damage from even touching AI in a final project on top of all that, despite the fact that a majority of game studios are using at least some programming AI at this point because it's so ridiculously efficient. And then time and time again you see people with no qualms about this stuff just diving in and hitting it big with no regard for the consequences. I'm practically a nobody online, I don't know why I worry so much about my nonexistent reputation but I do.

So here I am wondering where I should draw the line on everything, given the future I see us racing towards. I don't want to be one of the people burying my head in the sand and hiding from the inevitable, refusing to take part until I'm like a grandparent who refused to ever learn how to use the internet.

Right now my line is I don't mind AI helping out with programming, after all you can't even google a programming problem without getting an AI solution handed to you. It definitely made a big difference in my ability to mod minecraft, traditional google search was mostly worthless at finding out how to do things in forge. And I didn't want to ask for help because you'd either get ignored or get a bunch of attitude from people who think you are stupid for even needing to ask.

Now if I google for help doing custom liquids in a specific version of forge I can actually find the result instead of a thousand forum and reddit posts that are rarely the right answer and also always in an older version of the game where the API was completely different. I was able to finally finish the version migration I had started years ago and then finish a bunch of features that had been on my to do list for even longer.

I've seen some people compare AI programming to the difference between writing modern programming languages vs machine code. We've always been moving towards reducing the amount of actual work taken to implement a programming task. It's an interesting comparison but I still think with 8 billion people on the planet that the process of making things becoming this easy and accessible is going to have permanent consequences that can't be compared to any technological revolution we've seen previously.