John IV

I've been building Confluence on @spawn for almost a month now, and I am getting close to publishing. After a lot of iteration and polish it is almost time to get some actual player feedback. I can't wait to see what crazy combos people pull off.

I have mixed feelings on the new Claude Code rule against recreating existing games. Clones of established IPs are bad for the optics of every dev using AI, so I'm not sad to see them go. But some of these random people with those assets aren't just cloning. They're releasing unique games and mechanics that the IP holders could have built themselves. Players would have loved them, but those companies refuse to take any risk, and their bureaucratic bloat keeps anything new and exciting from being released.

Most people still think of AI as an advanced search engine. I am building a wizard dueling arena without writing a single line of code with an AI that debugs itself and ships its own updates. We are not the same.

When it's time for a wizard battle, but you've got to skedaddle.

If there was a section on Steam explicitly for games developed with AI and another section with games that weren't, you would really just have 2 sections that had games developed with AI, but the only difference is one is honest and the other isn't.

This sentiment is true across all technical fields. Every single company is incorporating AI into their workflow to some degree, and the software engineers that are being cucked by the companies they work for are begging to get access because they too know that if they don't they will be left behind and their jobs will be in jeopardy long term. For game dev in particular, devs use it quietly and choose not to disclose it because they know that the antis will come out in droves to protest the use of a tool instead of judging the games they make based on their merits.

Epic just confirmed AI is used to build Unreal Engine. So does every UE5 game need the AI label now? If yes, the label ends up on everything and tells players nothing. If no, somebody has to explain where the line actually is.

Sakurai says most AI-generated games will copy systems we've already seen, and that people with their own ideas will matter more than ever. He's right, but that was true before AI too. Most devs take inspiration from games that they love, or take systems/mechanics that have worked in the past. The same applies to artists and musicians. Imitation is the sincerest form of flattery. AI can make the process more efficient, but it can't tell you what makes a game fun, unique, and "feel" good to play.

Next Fest wasn't buried by AI. 35% of the games disclosed AI and only 2 of the top 25 demos did. Players walked right past the slop. What decided the top 25 was having an audience before the fest started. That was true before AI too. He's right that discovery is getting harder for unknown indies. But blaming AI lets everyone skip the part where you build an audience before you launch.

Confluence has come a long way and I have managed to finish 4 of the elements I have been working on so far: fire, wind, water, and lightning. Here is a compilation of combat between a bot and myself for testing. #ScreenshotSaturday

This is one of the most intellectually dishonest takes someone can have on this topic. Games should not have to disclose their AI use simply because for whatever reason, there is a horde of people who rush to shit on any project that uses AI without giving it a fair shot. The gameplay and the experience is what a game should be judged on, not the tools used to create that experience. If that experience was created with the help of AI, you shouldn't feel less than because of it. Little by little there will be breakout successes of games that confidently state their use of AI. It will become normal and expected because they are tools to be used for a purpose. Sure you will have some slop along the way, and there will be people who benefit from it that do not deserve it, but this will be the exception, not the rule. Inevitably when that happens, people with these radical positions will fade into irrelevance where they belong and good games will surface to the top like they always have.

Just when I thought Jaben couldn't be more based. Here's your crown king 👑

People are acting like AI invented shovelware. Steam was already putting out around 19,000 games a year before anyone was vibe coding, and most of them were dogshit then too. The good ones still found their players.

My game Confluence is all about blending magic together, but I figured I would showcase each element on their own, starting with fire. Here is a mock fight that I had with a bot. This took multiple attempts to get a win and the bot beat my ass a lot . I'm bad at my own game, but maybe when I publish there will be some prodigies in the @spawn community that will go crazy.

Everyone's worried the market will get flooded now that anyone can make a game. Most people are consumers to their core. Finishing something still takes hundreds or thousands of hours, and the vast majority will never put that time in no matter how easy the tools make it.

I've hit a pretty major milestone in the development of Confluence, my Spellbreak inspired mage arena game. I have created enough spells and interaction to put together a fight simulation between bots and it is coming together pretty well, but these bots are only scratching the surface with what they can do to make spell combinations. Being able to build something like this out of nothing has been amazing. A big thank you to @spawn for building such an awesome platform.

Nobody asks if a house was built with a nail gun or a hammer. They ask if it’s a good house. Games are no different.

Building with Claude, Blender, Godot and Meshy/Tripo works well and I’ve been able to make a ton of progress that way. But describing something in plain English on @spawn and watching it get built in front of you, then playtesting it right there without jumping between 4 different programs, is a completely different experience.