You may as well buy a shooter game starter pack or whatever that can save you >1year of coding, no llm needed.
Code is not a hard part.
Making mechanics fun and good assets is what is hard and takes forever.
Sure you can use llm to write a generic game, but its easier to find same game on github and just use that code, why would you write it again with llm.
Having a good and semi unique idea, is a rare. If I had a great game mechanic idea, the rest would be trivial.
Say you do get a good game loop together that you feel will be successful. You will also now need to loop in art teams for artistic direction, music, character design, etc. A good game loop isnt enough, it needs to be presented in an equally interesting and unique way.
Finally, there is the risk. There is a massive time investment in making games, and you are catering to an audience that is not only accustomed to pirating but finds it morally righteous to steal your work. This is why app developers prefer to make iOS apps. The customers are accustomed to paying and have little interest in pirating.
post-launch and even before that, your job becomes paying and convincing streamers to play your game constantly in the HOPE people start to notice it.
All of this stress and work to hopefully just make an ok amount of money. I have so many excellent games in my steam library by indie devs that gave up after one or two very successful games. And I doubt it's because everything was going so well.
Just look at something like ludum dare and all the top entries (out of thousands of games submitted) are all usually quite polished given the timespan.
1: Even with AI, it's a lot of work to make a full game. When most people think "I have a cool game idea", they're usually imagining something polished and non-trivial, possibly 3d. You could make a short text adventure in a few days with AI, or a very simplistic 2d game, but anything more ambitious (like 3d) is going to take a lot more effort.
2: Releasing on steam requires you to pay $100. I imagine this is a substantial deterrent for "3-day projects", unless you think it'll sell $100 worth.
3: There's more to game development than creating assets and writing code. The author of the article recreated an existing game, which sidesteps one of the most difficult parts of gamedev: design. Creating a compelling game is surprisingly difficult. Granted, you don't need a compelling game in order to release on steam, but I myself have made many prototypes over the years which I've abandoned because the idea just wasn't as interesting as I thought it would be.
4: I've made a few prototypes with AI assets, and one issue I frequently run into with image generation is: it still takes a fair amount of work to generate the same character in different poses, facial expressions, outfits, etc.
5: There is still considerable prejudice against using AI to make game assets. I think some people (myself included) are hesitant to release a game with lots of AI generated assets at the current moment, for fear of public backlash. Eventually that will calm down and it will become more socially acceptable to use AI to generate game assets.
I am bullish about AI improvement over the next decade, and I think we'll gradually see all of these issues resolve themselves as AI improves. But at the present moment, it's not quite as easy as the article makes it seem.
Also the idea that a dev who could making a game in 24 hour would create something professional and polished in 3 days is a joke. The answer to “where are all the games” is simple: LLMs don’t actually make a huge impact on making a real game.
Which is really easy to argue it's more down to Unity + successors making game dev accessible as it starts in 2015.
No huge spike since Claude code got released or anything like that.
Not really. The jump from 2023 to 2024 is bigger than the jump from 2019-2022 in raw numbers and 2020-2022 in %. So the jump of 3 to 4 years happened in a single year.
The open secret is that they might not start coding or building assets until the start time, but they have spent a lot of time thinking about the ideas before then (even when the "theme" isn't known before hand people tend to make ideas fit theme with tweaks), which just speaks to the "code is not the bottleneck" thesis.
There are a bunch of games made using heavy gen ai but it is usually for art and dialogue. Most players can tell quickly and drop the game. Games are fundamentally creative things and most interesting art work was not done in 5s with a prompt.
All of these take time and many of them are iterative processes where you might not even know if it fits or is right before multiple tries.
Also the jump in 2024 is only around 10-15% more games than we would have expected from the previous trend. Assuming all of that is directly down to AI, I wouldn’t call that an explosion.
From what I’ve seen, most of the growth was in NSFW shovelware and was just people noticing a business opportunity. This also explains why the number it takes in 2025 isn’t showing similar growth.
No we're not. Use Wayback machine or whatever and this year is 1k+ ahead at the same date.
https://web.archive.org/web/20240822090931/https://steamdb.i...
>Also the jump in 2024 is only around 10-15% more games than we would have expected from the previous trend. Assuming all of that is directly down to AI, I wouldn’t call that an explosion.
How many games do you imagine can be released per day even with the help of current Sota LLMs ? Nevermind the fact that you have to pay $100 to distribute your game on Steam. You're not making a game you'd pay $100 to distribute in 3 days, LLM help or not.
But fair, exploded is probably overstating it.
>How many games do you imagine can be released per day even with the help of current Sota LLMs ?
Given the number of people who want to make games—if code is the bottleneck, and LLMs can really make you hugely more productive, I’d expect to see an actual explosion.
My experience is that neither of those assumptions are true though.
Game development is not a zero sum game. There can be multiple bottlenecks or difficult hurdles.
>and LLMs can really make you hugely more productive, I’d expect to see an actual explosion.
Well growth was double the previous year. Maybe you might not call that an explosion, it's still a very noticeable uptick.
I think it’s much more likely that LLMs don’t actually boost productivity all that much.
This is mostly from artists themselves. Most people are ok with even fully generated content if it's fun and interesting. Anyway, solo developer cannot afford even single artist, musician, writer.
That's where LLM helps. I recently tried GPT5 for story telling. Gave it a single image (in a bar, women, man, and a gun) and asked for a short story. Then asked for the next part 6 times. Every time at the end for illustration. The result was consistent and readable. Images generated had even similar faces. Remember, that was a problem with earlier models. I'm sure this will be used fill Amazon's bookshelves.
But if you mean copying an already successful PC game with AI Slop assets and putting absolutely no thought into what makes the game good, then you probably should work in a field you actually care about instead