Before artificial intelligence became a marketable buzz word, most games already included artificial intelligence (like NPCs) I guess when you have a GPT shaped hammer, everything looks like a nail.

[–] Mikina@programming.dev 4 points 11 months ago (1 children)

That's what I was reffering to. I'm looking for articles and inspiration about how to cleverly write NPC game AI that I'm struggling with, I don't want to see how are other people raping game deveopment, or 1000th tutorial about steering behaviors (which are, by the way, awfull solution for most of use-cases, and you will get frustrated with them - Context Steering or RVHO is way better, but explain that to any low-effort youtuber).

I've recetly just had to start using Google Scholar instead of search, just so I can find the answers I'm looking for...

[–] grenndel@lemmy.ca 1 points 11 months ago (1 children)

If you are looking into how to write game ai there's a few key terms that can help a ton. Look into anything related to the game FEAR there AI was considered revolutionary at the time and balanced difficulty without knowing too much.

A few other terms are GOAP for goal oriented action programming, behavior trees. And as weird as it sounds looking up logic used my mmorpg bots can have a ton of great logic as the ones not running a completely script path do interact with the game world based on changing factors.

[–] Mikina@programming.dev 1 points 11 months ago

Thank you! My main issue is that while I'm familiar with all those algorithms, its usually pretty simple to find how do they work and how to use them for very basic stuff, but its almost impossible to research into actual best practices in how and when to use them, once you are working on moderately complex problem, especially stuff like formations, squad cooperation and more complex behavior (where I.e behavior trees start to have issues once you realize you have tons of interrupt events at almost every node, defeating the point of behavior trees - which can happen if you're using them wrong, but no one usually talks about it at that level).

And I'm also dealing with issue that isn't really mentioned too much, and that is scale. Things like GOAP would probably be infeasible to scale at hundreds of units on the screen, which require and entirely different and way less talked-about algorithms.

I've eventually found what I needed, but I did have to resort to reading through various papers published on the subject, because just googling "efficient squad based AI behavior algorithm" will unfortunately not get you far.

But its possible that I'm just being too harsh, and that the search results were always the same level of depth - only my experience has grown over the years, and such basic solutions are no longer sufficient for my projects, and it makes sense that no-one really has a reason to write blog posts of such depth - you just publish papers and give talks about it.

Aside from the AI related keywords. I'm still salty about what the buzzword did to my search results.