Skip to main content

Browse GamesGame JamsUpload GameDeveloper LogsCommunity
Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines
qingisbuilding2 days ago(+1)

I'm not an expert in this domain but I'll pass to someone I know who is to see if they would be willing to give some feedback. 

as someone with only a passing interest, my main thoughts are:

- "biological neuro basis for novel network architecture" is an enormously oversubscribed field with relatively few good results - you lampshade this a bit but i think the pitch audience will be looking for concrete evidence your thing is exceptional because the baseline-crank-rate is so high. ("aeroplanes are great, but they'll fly even better if the wings flap!") 

- I'd love to hear more about your snake methodology. with a lot of skepticism to overcome (both about your specific domain and also the general state of ML research experimental techniques) you're battling priors of "did they make a mistake with the experiment" 

- the paper you linked to seems like a concrete and modest improvement to a specific technique. the research gap to get from there to Atari games seems pretty huge but I don't have info on where you are with your other puzzle pieces - being able to visually parse the games is useful but also the least interesting piece compared to the actual RL when it comes to success at the Atari stuff

- why snake, and why Atari? why are these useful benchmarks as your milestones towards your field-changing ambitions? AFAIK none of the original deepmind Atari techniques turned out to be of long term importance, so why is it a good proxy for what you are trying to achieve? 

Reply
extenebrislucet2 days ago (1 edit)

Thanks for the feedback! To address some of your points:

Yeah. Bio inspired is a graveyard of cargo cults with mediocre ideas, poor implementations, and zero results. ML people look down on the field for historically validated, if not universally correct, reasons. The wings don't need to flap, but they DO need to twist -- iykyk

I could write about the snake implementations specifically, I mostly put that aside to just focus on Atari 100k as it significantly more publishable/relevant. 

Yes, the paper is a refinement of a specific algorithm that makes it more applicable to always-on, always-learning systems, really it's just a thing that I built for my own purposes, then realized that it was actually a novel contribution, and worth writing up, for the practice if nothing else. Absolutely the actual brain to play Atari games is *significantly* more complex than just this algorithm.

Why Snake, why Atari? My impression is that the real limiter we're hitting in the field is the inability to run continuous, adaptable intelligences on smaller, edge hardware. So, my internal roadmap is Atari 100k, then the harder version of Atari 100k (take off the training wheels that are typically part of the benchmark), then drone racing and harder video games, then different types of robotics and significantly harder video games. Eventually: practical brains that can be deployed into robot bodies like Optimus or Unitree, and produce actual value output in the real world. The video games are a ramp that leads to physical environments. I think. The original Deepmind Atari solutions didn't matter in the end because they weren't solving in a generalizable way, and they were using way too much compute. Afaict.

Reply
AboutFAQBlogContact us
Copyright © 2026 itch corp · Directory · Terms · Privacy · Cookies