Live data from Hacker News

Building an AI game studio: what we've learned so far

braindump.me

71–80 of 312 posts

Re: Building an AI game studio: what we've learned so far

#71

Earlier quoted context omitted.

Word spreads fast on good games due to Reddit and Twitch. Look at the success of Balatro and how much marketing that had. The number of games you listed just highlights how little Steam filters.

By definition you can't pinpoint (but I think it's a fair assumption) the many great games that don't gain traction on Reddit and Twitch, and silently sink without notice. You've just picked a single anecdote.

And you picked none.

Re: Building an AI game studio: what we've learned so far

#74
post #68

How robust is the system to fix bugs? Can you just point at a thing and say "fix this bug"? How does that side of it work?

So it's hard to give a number on how robust it is, but yes, simply telling GPT that "this doesn't work" or "try again" does work a lot of the time. We're trying to give GPT as much context as possible to help it figure out where things went wrong. Currently that means giving it type errors, execution errors, self-critique and user feedback. But we're working on expanding that; for instance we want to experiment with automatically generating tests that validate things are working as they should. We also want to add a GPT based "debugger" that can help you pinpoint exactly what went wrong.

Re: Building an AI game studio: what we've learned so far

#75
post #30

They mention using a service called Meshy for generating models, and I noticed that they have a free plan where the generated assets have a "license" of CC-BY-4.0[1]. Any lawyers in here that can comment on that? They're claiming copyright ownership over models generated from customer prompts, and I doubt they trained their model exclusively on stuff they own/licensed. 1: https://www.meshy.ai/pricing

https://app.meshy.ai/preview?web=018f8ef6-e562-7e19-b0d7-5f6...

https://app.meshy.ai/preview?web=018f4d04-6a56-701e-9abc-62a...

IANAL, but I think Blizzard and Marvel would beg to differ that these are kosher for commercial use.

They're also very poor quality models but that's a separate issue.

Re: Building an AI game studio: what we've learned so far

#76
post #53

Earlier quoted context omitted.

I guess if you mean in the sense that writing a book is just writing words or publishing an app is not hard it makes some sense... But no. Shippable games are incredibly hard to make. Why do you think publishers invest in new games instead of just hunt for diamonds in the rough?

Probably the same effect that VCs invest in many startups, knowing only a tiny fraction will make a large profit?

How does that gel with the idea that it's just the marketing thing?

Re: Building an AI game studio: what we've learned so far

#77

Hard facts: Steam has over 73k games - and is growing at over 10k games per year and increasing. The median game on Steam earns less than $800 - down from $4k in 2020. Making a game is not hard - marketing a game is.

I miss the days of Flash.

I used to be able to make a game in a few weeks from my dorm, add Mochi ads, upload it to Newgrounds, and then it would spread to thousands of sites and get played by hundreds of thousands of people.

Re: Building an AI game studio: what we've learned so far

#78
post #26

Earlier quoted context omitted.

It is not just about marketing though. There is this belief, especially among fledgling indies that if they make a game that they think is good, and if they are able to market it well, it will be a success, even a hit. A hit in gamedev is not easier than creating a hit in the music industry really. Even the veteran composers are not able to replicate (or even achieve) creating a hit. Possibility of success depends on…

> If you are a beginner composer creating a hit, it is an extreme fluke. You're punching too high. Music is a passion for many people. Same as game dev. I would expect >95% of game devs to be perfectly happy if they can make a living creating indie games. Of course it's nice to make a hit that makes you tens of millions of dollars, but most game devs would be happy to make enough to pay themselves a salary to live on…

>> I'd honestly struggle to find any good games (right for their time and audience) that failed because of a lack of suitable marketing.

> I'll offer Flappy Bird as an example.

Flappy Bird is a bizarre exception: it was a massive viral hit before it was voluntarily withdrawn by the developer feeling guilty over its success. It was then cloned a lot.

Re: Building an AI game studio: what we've learned so far

#79
post #41

Earlier quoted context omitted.

This is subject to survivorship bias. Balatro is a great game that deserves its success. But it's easy to imagine a universe where it simply never got any traction, anywhere, and died in obscurity. Conversely, this suggests that in our current universe, many games just as great as Balatro have quietly passed without ever getting their 15 minutes of fame. At the end of the day, much of success just hinges on luck. The…

I immediately knew someone would claim "survivorship bias" but at the same I knew that they would fail to come up with a single counterexample.

That's sort of the point of survivorship bias. Name one unsuccessful would-be entrepreneur. It's hard because chances are, anyone that's recognizable was successful in some way.

For games, I'll give an example: Aegis Defenders. Great game, no traction.

Post reply on HN