Bonus article tonight because my newsreader is blowing up on Google’s new AI Genie. There is no link to the Genie AI service yet (not public yet), but you can google “Google Genie” if you want to get lost in the rabbit hole of tech rags. Just look at all these headlines, some of them worse than clickbait, bordering absolute stupidity. It’s like a hyper-orgasmic text version of (AI) Girls Gone Wild.
Come on, now, AI needs a reality check when all these different news sources are reporting with various stages of hyperbole about completely AI generated games from user text prompts. Like tomorrow or even a few years from now we’re going to see some kind of amazing 2D platformer game completely generated from Google Genie by somebody typing a paragraph of well-written text input.
(watch this happen, and then have to walk this back — sigh)
This article from Dataconomy is one of my favorites — not:
It’s this kind of intellectual dishonesty that runs most gamer readers for the hills. Lucky for PGM readers, we’re here to be the devil’s advocate of AI. There are some sane gaming-related publications that will report on just how nascent AI real world usefulness truly is in 2024.
If 2023 was the year of AI birth, 2024 is the year AI sleeps most the time, and also craps out bad ideas into some really smelly diapers. Yeah, if you’ve ever changed a baby’s diaper, you know the odor.
Tech companies seem to be going all in on AI and some of these ideas are terrible. Does the game industry need more games created by AI from text prompts?
No, no, no, no, no, no — a hundred times no.
There are plenty of terrible games out there flooding the market built in part by AI and completely generated AI games is not something we’re going to see taking off any time soon.
Why not?
Because there are too many freaking games already out there to play!
Us gamers are hopelessly outnumbered. We don’t need completely bot-created games. We don’t need to read blogs and news articles generated by AI. We don’t need websites completely crapped out by AI bots to fake search engines for attention and clog up meaningful search results for human beings. We don’t need more crap online pushing human beings away from engaging because they now have to play AI vs. human whack-a-mole. We have enough crap in the baby diapers to change already here.
Is the idea of games generated from text prompts completely meritless? No.
I am fascinated with the prospect of some AI generated code used in games for RAD (rapid application development) that can then be tweaked by human programmers. Some code is very repetitive, and AI could assist in this area. The biggest problem with AI generated games isn’t the idea itself, it’s the company that’s putting out this “Genie.”
Google.
Yes, said it. Google sucks for new ideas, because they have a long, established history of coming up with stuff or buying their way into something cool and then years later — sometimes not even that many years — closing these service down. No need for a list here, but — cough, Stadia, cough — they give up on stuff too soon. They will give up on whatever new fangled AI creation in favor of some other project before this matures.
Google just isn’t the company to develop AI-generated game software. They need to get their brains around search, because that’s the girl that brought them to the dance. Keep their search engine from being littered by trash AI content.
An AI generated service more interesting to this author right now is Rosebud (https://play.rosebud.ai/home). Am a little bit concerned about the waitlist thing. Waiting lists for tech stuff is a minefield. There seriously aren’t that many developers clamoring to make AI generated games. Getting people excited for something by creating either real or artificial scarcity can be a mixed bag. Am not sure what to think of that yet, so it gets only a mention, link and completely underwhelming screenshot at this time.
Rosebud doesn’t look any better than Google Genie, except that we can actually see something. I don’t know either service yet, but am interested in researching, learning more and seeing where it leads. How long will it take for 5,000+ people in front of me to get invited in? Need an AI text input answer for that.
I’ll check out Google Genie, too, when and if it’s ever made public. It won’t be as useful as the AI image generators from text, which are under fire for copyright violations. This is the only AI being even somewhat regularly used, and you can see it in the form of AI-generated images that are featured at the top of many posts here at PGM. This proves we’re not all anti-AI, but definitely anti-overhype.
AI is way, way, way overhyped right now. Is there potential for more real world AI usage in the future? Absolutely, yes.
Baby steps, folks. The AI infant is still mostly sleeping and crapping. We know where most of that waste goes.