yeah some friends of mine in game dev are a bit doom and gloom about all this and I need to keep reminding them that the hard work is making the million decisions that communicate the vision, the code/assets were never really what people were buying your game for.
Even assuming an AI could do literally every technical part of game dev someone (for now) needs to sit there and go "nah this part doesn't feel right" and "that boss 2nd phase opener attack is too long" etc.
until AI can also answer those questions and intuit what is good and bad taste there's still stuff for game devs to do.
"minecraft in skyrim" is cool for about 10 seconds till you think about actually sitting down to play it.
They're getting surprisingly good at identifying why things don't seem right. I came back to an old game project that I had shelved due to the core game loop not feeling right and I decided to just ask Astra to analyze the gameplay and figure out what was wrong and it was able to provide a good direction to take it that also works well with the game mechanics that were already in place.
The new Terra 6.1 is pretty good at doing user testing as well via computer use, going through all the features and noting any bugs or inconsistencies.
It feels like we might be on a trajectory that takes us to a place where someone can just say "I'm bored, make me a new game" and then provide minimal feedback like "this isn't fun, make it better" and get an entertaining game without a lot of iteration. Of course we'll still have games that come from human ideas as well but vibe games seem like they're here to stay.
loading story #50268912
loading story #50269527
Yes, Opus 5.5 can build out quite a bit on its own that is super impressive when you first look at it. But spending just a couple minutes will make you realize dozens of things that are off.