I have some strong feelings about this particular story (people point LLMs at your devlog)…
- Video devlogs were long a shitty way to make a community (I can explain this one in more detail, but in short, the tradeoff between opportunity cost and video quality is a Pareto curve that is bad at all points on the curve, if your goals include both “make game” and “build community for players with devlogs”, the way out is to drop one of those two requirements, either “make YouTube/Twitch/TikTok channel” replaces “make game”, or you build some other audience for your devlogs)
- Good devlogs don’t have that level of detail, to let people easily recreate games
- Most games can’t easily be replicated by LLMs, in short, the people who are good at steering LLMs like that are making their own games.
This is not the first time I heard this story. Sometimes it comes with the lamentation “back in the day people could build communities with devlogs” and no, that was generally not a good way to build communities. It was mostly good streams from YouTubers who were making the game in order to make YouTube content or it was mediocre streams from random devs. Throw in a few people who are famous game devs who choose to stream and get a big audience because they already have a community.
It doesn't need to recreate your game. It can still muddy your marketing and redden the ocean. Especially if someone is in the business of trying to front run an indie game, they don't need to be any where close to a full or good game to attempt to steal any mindshare your project had.
Is this happening at any significant rate?
AFAICT this is a fear that some game developers have, that someone will steal their game and run with it, and the stories told are more repeated based on this shared fear than based on realistic scenarios. People repeat it because it’s a good story.
And yes, I’m aware of some situations like 0x10c and the like.
I don't think it has to be about stealing their game with any level of skill or fidelity. It's enough to just spam 10 garbage clones with vaguely similar banner images and descriptions so that nobody will find theirs.
The appstores are full of knockoffs and even resubmitted decompiles. I certainly wouldn't put it past the scammers.
More specifically to AI... A blog on a nice website is no longer a signal for quality so there's just no point in devs investing in that.
I frequent a couple of game development communities, and it's kind of hilarious that some people won't even talk about their general game concept for fear that others will steal their idea. Like wow, you're making another rogue deck-builder, I have to steal that!
The idea is not valuable, ideas are a dime a dozen. The actual value is in your implementation.
Is that true anymore, LLM will implement everything for you, idea actually matters more now
If you point an LLM at an idea, you'll get a "lossy JPEG" version of that idea.
It requires discernment, taste, on the part of the user to be able to point the LLM at the weird parts that look wrong.
Even with that, the LLM can only fix most of, not all of, those rough edges.
LLMs will generate. Taste continues to dictate what to keep/discard. (Note that taste can get you to good, but good won’t get you to money on its own)
It's still true for now. AI can't yet make a game with complex mechanics feel good to a human, although it can speed up a lot of the implementation.
This is fantasy and if it was true the idea would be public when the product was released.
There's no moat in an idea either so I guess it's down to marketing budget.
I've been vibe coding a game on the side.
It really makes me appreciate how important good game design is. Claude is doing a fine job coding everything I describe, but it doesn't really understand fun, so I need to.
(It's not a great game)
This is my experience too. I think people with no gamedev experience widely underestimate the challenge of good game design and the need for iteration with real people testing the game. In a similar way that an inexperienced game developer overestimates their skill to assess the fun and all the small details that matter, and will get demolished when it's first playtested by other people.
> Good devlogs don’t have that level of detail, to let people easily recreate games
The whole problem with LLMs is that you don't need the detail.
Details have a huge impact on how fun a game is. So far the vibeslop games I've seen look about as fun as crappy unity assets flips or mobile games. every so often the Instagram algo will show me some post a long the lines of "The game industry is finished!" and there is no way anyone can seriously looks at the games in those posts and say they look fun to play.
And someone trying to vibe copy another person's game is definitely the type who has no idea of value to add. So many of the slop games the whole concepts are so generic that they definitely just asked chatgpt for everything.
In the modding scenes for some games I play I've seen vibe coded mods where the gameplay additions make no sense and have no sense of balance or fun, with these completely new to the community devs having ko-fi links set up from the start.
What I think of is all the HD remasters out there which look disastrously worse than the original, at least in my eyes. Why does this happen? Are the best artists working on new games instead? Are remasters pushed out with less care, shorter schedules, and less budget? Maybe some combination… and maybe there are some parallels with quick, mostly unsupervised LLM copies of a game.
If someone makes a cheap copy of my unreleased game and it's bad, that's also bad for me, I think.
I think the much bigger problem is that these people are usually trying to do some "passive income" play, drop-shipping type crap. And will soon be spamming what would be the usual discovery mechanisms with slop. It's happening on youtube for educational type videos.
Ok, so let’s get some data points. Are there games being recreated from devlogs? I’d like to see what the recreations look like.
> Ok, so let’s get some data points. Are there games being recreated from devlogs? I’d like to see what the recreations look like.
An LLM "recreation" doesn't have to be complete or very good. It just has to steal enough thunder to be profitable.
From what I’ve seen, most of the flood of LLM-created games aren’t profitable, just like most of the games people make in devlogs. There’s very little thunder to be stolen in the first place.
> From what I’ve seen, most of the flood of LLM-created games aren’t profitable
It's it possible for an LLM-created game to be profitable but "unsuccessful?"
For instance, if (on average) if it costs $100 in tokens and time to make a crappy LLM-game clone, but you can clear $200 in sales per game on average, you're ahead.
And the economics of slop mean there are a lot of people in 3rd world countries who will make all that effort for a $100 payout (or less).
I’m sure somebody out there is pointing an LLM at things and getting a $200 average payout or better on $100 in tokens, but if your game is getting copied for $100 in tokens, I’m not sold on the idea that you were going to make money in the first place.
> but if your game is getting copied for $100 in tokens
My point is they don't actually have to do a good job, they need to make a pile of shit that looks just good enough to trick a few people into buying it.
Ok, in order to make money? Let people make some money making shitty copies of my game. I think there are roughly two scenarios here.
Scenario one: they’re competing with my game and stealing my thunder. In this scenario my game wasn’t very good to begin with.
Scenario two: we’re going after different customers. They get the customers who want a cheaper game, I’m getting the customers who want a better game.