LarsDu88
I started building a VR game during COVID and released it
2024. When it released, I was using GPT and Copilot to add
a lot of quality-of-life stuff (like visual juice)
faster.Yesterday, I finally got around to setting Opus 4.8
on the codebase. It was able to find and correct numerous
subtle performance issues.Features I could no longer spend
time implementing, it could one-shot in roughly 10 minutes
(not without issues). I could also fan out agents to work
on multiple things at once.One thing I found is that the
knowledge I gained from doing things by hand greatly
helped reject bad AI generated ideas. For example, my game
is a starfighter flight sim and one idea was to tick AI
collision detection checks lower the further AI ships were
away from the player, which for a flight simulator leads
to a lot of crashing into the terrain. A great idea for an
FPS, but a terrible idea for a flight sim where enemy
ships are often very close to the terrain.My takeaway is
that we are in a golden era where we currently have
projects that are half human coded, where AI can pickup
the slack. But soon we will be in the Dark Ages where AI
generates all the code, and the end result will be much
worse as the devs begin to lose an understanding of what
they are creating.
|
> Zarathruster
I recently tried to learn gamedev with Claude as a
tutor of sorts. It didn't end well. Claude confidently
led me off a cliff and I had throw the whole thing in
the trash. It seems there's just no shortcut around
learning stuff the hard way. When I've got enough reps
in to be able to tell smart vs. stupid decisions, I
imagine it'll be a lot more useful, but at this point
I suspect AI is far from being good enough to get us
into your described Dark Ages.
|
schnitzelstoat
Are there any popular/successful vibe-coded games? I
suppose perhaps they wouldn't disclose that it was
vibe-coded but I'm not aware of a single one.
|
> sarchertech
Look at steam releases. May 2025: 1727. May 2026:
1875A year over year increase matching the same trend
that has existed since 2017. The same pattern holds
for every month this year except March (where there
was an increase year over year of about 600 games).If
you've ever been to a game dev forum, you'll see that
there are at least 10x as many people who want to make
a game as there are people who have made a game (it's
probably much higher than 10x).If games were easy to
vibe code, or if AI speed up game dev 10x, I would
expect it to almost immediately show up as a flood of
games on Steam.
|
> > Cthulhu_
Thing is, Steam already was flooded with
shovelware - I suspect a percentage of new
relelases includes plenty of AI usage that was
previously made in other ways (e.g. using asset
libraries)
|
> > > sarchertech
Yeah it was flooded with shovelware. But I'm
not talking about people who want to publish
asset flips for a quick buck.I'm talking about
all of the people out there who have ever
tried to make a game or never got started
because it was too hard. In my CS cohort years
ago, pretty much everyone there got into
programming originally because they wanted to
make a game.Games definitely have the highest
ratio of I want to make a game to I have made
a game of any type of software. If it was
suddenly easy to make "your game", Steam would
explode.
|
> > > doctorpangloss
roblox studio has deep generative AI
integration, it has absorbed many users both
creators and players during the time period
you are measuring. steam grew so little
despite digital audiences growing so much.
|
> > kelvinjps10
But still how many of these games are successful?
|
> > short_sells_poo
I wonder if there's also another aspect: Games
have to be fun for humans and this involves a lot
of trial and error with actual humans in the
loop.You can't just spin up a bunch of claude
agents to implement a game for you, because
implementing a new feature requires that you
playtest it with a fairly fine granularity as it's
being implemented.You can save some time
implementing various subsystems with llms, but at
some point the dev cycle will turn into: tweak
things, build, play, rinse and repeat.I'm sure
asset flip and friendslop games will become
cheaper to make with ai tooling, but if you want
to make a genuinely good game, it will have to
involve humans actually playtesting it during
development.
|
> > > notahacker
Also, game audiences are brutal about games
that are flawed, and even more brutal if they
think it's shovelware. If you're selling games
on Steam, something with "Mostly Negative"
reviews isn't serious competition (at least
not unless it's part of a long established
franchise with a big marketing budget).And
sports management sims are one of the most
brutal of the lot. Slop might get you
something that superficially looks like a
sports sim (a decent third party UI library
would get you quite some way in the past), but
what fans actually care about is perceived
realism, game mechanics, balance/challenge etc
(and in real world sims, "how has my team been
represented?") and that's a design decision,
play testing, parameter tweaking and user
engagement exercise, not a feature-adding one.
|
> yoz-y
Turns out making a game, even with full help of the
LLM power is still a massive effort most people won't
go through.Even simpler but non-trivial programs
require a lot of back and forth. So in the end it will
be the same kind of motivated people that will be able
to produce something good. We're nowhere near "Claude,
build me GTA6"
|
> > uludag
Here's an example of this regarding the recent
indie hit "Mina the Hollower." They were recently
interviewed and asked the question "What is Yacht
Club's stance on AI and has AI been used in the
production of Mina the Hollower at all?" to which
their reply was "We all got caught up in AI fever
like the rest of the world, but we didn't find it
was very effective for what we're doing. Maybe our
work just isn't that generic! We've found some
ways it can help... like Google or a thesaurus,
but it hasn't affected what's in our game."
[1]It's hard to immagine that even the best AI
model would result in anything better than a
marginal reduction in release timeline. Like maybe
for such projects one could spend $X00,000 worth
of tokens for maybe like a %single-digit-percent
reduction in time to release. Marginally good,
maybe even project saving, but not any larger a
paradigm shift than Unity was.[1]
https://www.gamereactor.eu/mina-the-hollower-inter
view-discu...
|
> > > Zarathruster
The conversation around AI in games is such a
polluted mess of confusion and bad faith. I
would hate to have to answer such an interview
question.Recently I came across a game on
Steam that was getting review-bombed because
the devs admitted to using AI on an entirely
different game. By "using AI" I mean they
admitted to using Cursor of all things. Not
assets, just a coding assistant.If we're going
to get backlash for stuff as stupid as this,
it's probably best to just keep one's entirely
mouth shut about all things AI.
|
> > > clates
> "What is Yacht Club's stance on AI and has
AI been used in the production of Mina the
Hollower at all?" to which their reply was "We
all got caught up in AI fever like the rest of
the world, but we didn't find it was very
effective for what we're doing. Maybe our work
just isn't that generic! We've found some ways
it can help... like Google or a thesaurus, but
it hasn't affected what's in our game." [1]Can
you see how this intentionally coy and evasive
answer was made specifically to not say "No,
we didn't use AI." but still sound like "No,
we didn't use AI." so that people wouldn't
immeidately bust out pitchforks and start
review bombing their game before it takes off?
|
> > > > uludag
Yeah, I get all forms of media are
avoiding the appearance of using AI like
the plague. I still think that what
they're saying can still be taken at face
value. Like if the game has a 3D sequel I
don't immagine the time to release or
quality would be outside the distribution
of how long similar games take pre-AI.
|
> > AndrewOMartin
We might get a powerful enough model to run
"Claude, build me GTA6" before GTA6.
|
> > > > adornKey
Hm. Nowadays a lot of games look alike.
They use some 3D-Engine and most of the
work is in the 3D-modelling and writing
some interesting scripts.For modelling and
scripting I think we're not far away. A
lot of games just reused old historical
stories or fiction and a lot of stories
feel like cheap soap operas. As soon as an
AI can separate the good from the bad
scrips it'll be mostly done.
|
> > > > > Cthulhu_
I'm sure AI will be used in asset
generation but it'll be in deployed
similarly as procedural generation of
mass assets like trees and npcs.I
don't yet see it used for characters
as they quickly become kind of generic
/ predictable.
|
> > > > > kakacik
Clearly you don't work in game
development, this comment put a smile
on my face. Although I expected a bit
more from hacker news crowd
|
> > > > > > whizzter
You overvalue the HN crowd (or
undervalue the AI hype-machine)
considering you're downvoted and
GP is upvoted (Another gamedev
here smiling at GP's comment).
|
> > > > > > gafferongames
Yet another classic unskilled but
unaware of it from the hacker news
crowd (another professional
gamedev here).
|
> > > > inigyou
It's not hard if you just imagine a slop
re-skin of any other GTA of similar game.
|
> > > > kobalsky
Did you miss the sarcasm or has GTA6
become some sort of anomalous memetic
agent that makes anyone who tries to work
on it not be able to finish it regardless
if it's human or AI?
|
> > > > > tripledry
Both probably, I don't usually
understand sarcasm in written text
unless it's explicitly stated.
|
> > > > pennaMan
someday we will have models that can
resolve physics to such degree to predict
the future with surgical accuracy and when
someone says "maybe models will become
advanced enough to create a whole other
universe from scratch" you will be there
saying "highly doubt it"
|
> > > > > owebmaster
This reminds me of a boss I had 20
years ago that said devs were going to
be replaced soon.Maybe he was 20 years
early or maybe it's not happening now
too.
|
> > > > > > pennaMan
Please note that I did not imply
devs won't exist in my contrived
fantasy.Who do you think will be
operating those mythological
tools?:)
|
> utopiah
> Are there any popular/successful vibe-coded
games?Fair question... I'd even go as far as
broadening the scope :Are there any popular/successful
vibe-coded anything?And by popular/successful I don't
mean bought Github stars from other GenAI/LLMs related
project as it's been a demonstrated practice
https://awesomeagents.ai/news/github-fake-stars-invest
igatio... for that specific domain now.
|
> > jvanderbot
At some point in the apparently-impending
"software is free" era, s/w stops being a product
that has to be "popular" and starts being mostly
bespoke. One possible future is that your machine
does you want because you have a local agent
molding it into the right form all the time.Bit of
a stretch, but possible. I've had agents write
100x more code for me _to be productive at things_
than they do for new projects I want to
sell/share.
|
> > microgpt
OpenClaw is popular - or is it? I don't actually
know anyone who uses it.Claude Code is popular,
and is vibecoded.
|
> > > swiftcoder
If the only successful uses of vibecoding are
more tools for vibecoding... Feels a bit like
the snake eating its own tail.That said, I
don't think that is actually the case - there
is a small and growing percentage of
LLM-written code in pretty much every piece of
tech I have insight into the internals
of.Though not in the sense of "implement GT6"
|
> > > utopiah
Not sure if my last comment was clear but I
meant specifically not related to GenAI/LLMs.
So no harness, no model tooling, etc. I mean
end-user used software. I don't mean that this
topic isn't interesting per se, only that (as
the article pointed out) it's been "gamed" so
much popularity there isn't really meaningful
anymore.
|
> > Aperocky
Depend on your definition of vibe coding - if it
means code came out from a LLM vs. keyboard, you
can bet that most updated software had some of it
now.But if you mean "Built me a GTA6", the answer
is zero, because LLMs simply does not have that
kind of capability.
|
> > yieldcrv
yes, there areno, most won't burn themselves by
publicly linking them as vibe codedthere is that
ny times article about the peptide guy and lovable
showcase by revenue though. I guess next up are
even more disqualifiers about the term
"successful", but my outstanding question is who
cares? What does convincing you buy, an Anthropic
pro subscription at best?
|
> > > pennaMan
don't ask how the sausage is made
|
> alexpotato
I've started trying to make some vibe coded games.The
best analogy to use is AI image generation e.g.-
someone posts a photo made with AI and your reaction
is "that's amazing!"- then you generate an image and
it's pretty good but not exactly what you want- then
you try to get the AI to change the image to what you
want and it's very difficult/requires a lot of
stepsMore specifically, you can easily spin up a game
and parts of it are great. e.g. I made a game where
you run a Soviet tractor factory [0]. The AI came up
with some good mechanics and funny scenarios
("inspector comes to visit from Moscow: bribe him"
etc). But the game mechanics are "off", the humor
doesn't always work, you need to gametest A LOT etc.0
- https://alexpotato.com/games/tractor47/?l=hn2
|
> > contextfree
I've been experimenting with vibecoding with a 3D
game with no engine, all in C++ with DX11,
all-procedural modeling. My experience is you can
quickly add a lot of features and mechanics that
work semi-well but getting any one feature or
scenario to work really well ends up "requiring"
digging into the code and writing stuff
yourself."Requiring" in the sense that while it
might be technically possible to get the AI to do
it, it becomes an incredibly annoying experience
to repeatedly try and fail to get it to properly
understand and implement what you want. This
almost seems less about whether the AI is as smart
or smarter than you or a better programmer or
whatever, and more about the medium of natural
language being worse than the medium of code for
this.
|
> dvh
Games were filled to the brink with asset flops and
low effort titles long before LLMs. I've personally
contributed 13 or so.
|
> uludag
I've been lurking on the aigamedev subreddit and
testing out the demos people have been sharing, as
this would be the place where people most enthusiastic
about vibe-coding games post and everything I've seen
has been immensely underwhelming. The problem with
these games fall into the following categories:- An
utter lack of animation. So many games feature mostly
static images.- AI assets make the game's art style
feel horribly incohesive. I played a TCG game where
there were high fantasy art, chibi 2d art, anime art,
seemingly used haphazardly.- Bad UX. For example, in
the same TCG game, all of the cards had very fancy
artwork generated, but the game board itself was far
away enough that the made the cards look like blobs on
the board.- Buggy mechanics: There was this guy why
shared a platformer generated by fable where you move
a lamp to create a shadow path. The character had very
awkward movement, would constantly get stuck/unstuck,
and would be hard to control.- Broken mechanics: A lot
of games featured exploits that just rendered the game
boring to play- Poor balancing- Lack of ovararching
game structure: no story, no meta-progression, no
world-map. These AI games tend to be isolated
experiences.I don't see the masses creating quality,
sellable games anytime in the near future. There are
so many aesthetic qualities to a game that have direct
human bottlenecks. AI can even make games
significantly worse as you can seemingly implement
many bad mechanics without validating them. Good game
devs seem to be relentlessly pruning out the things
that down work.
|
> adamzbik
S&box is a good example of how the scene might look in
the LLM era. They made games easy to make and publish,
and while Garry's Mod always had a hacky, rough vibe
to the available game modes (Source engine magic + CSS
textures), you'd always feel that the games were made
by people.Fast forward to 2026 and the next generation
platform is here, and while there are unique, passion
projects available, most of the discover screen is
filled with vibe coded xp farming games with AI slop
thumbnails. The issue is so big Facepunch had to
actively derank and punish games that would do this,
because the "marketing" content was so detached from
the game (despite everything being AI generated pretty
much).
|
> > lopis
Like in any other field of Software Engineering,
it's not the actual programming part that is hard.
|
> everyday7732
enclose.horse is somewhat popular and successful,
posted here on hacker news a while ago.
|
> SpaceNoodled
If there truly were, I'm sure we wouldn't hear the end
of it. However, the state of the art is all hype and
no substance, i.e., slop.
|
> fzeroracer
I certainly can't think of any. I've gone through
about 18,000 games on Steam (since I dig for hidden
gems and interesting games) and the only time a
vibe-coded game stood out to me was when it was so
awful I did a double take.There's so many vibe-coded
roguelite deckbuilders out there and all of them look
exactly the same. A bunch of shitty art with zero
cohesion, mechanics that don't actually work and a UI
that's actively offensive to engage with.
|
> rvz
We would have heard about one already by now but it
appears that we are still waiting.Or maybe it is not
worth the time and tokens spent to vibe code yet
another Minecraft / Roblox clone that makes no money.
|
dvh
When I was a boy my father forbade me playing Summer
Olympics on our Atari 800XL, he didn't like the crunchy
noises coming out of joystick and he didn't want to pay
for the repairs.
|
RobRivera
I have been building my own game and game engine for a
while now, and cannot help but see whats really happening
with vibe coded game generation: and investment into the
content generation pipeline.I feel a solid game engine
paired with a good content pipeline can ship games at such
a rapid clup that the gameslop gets lost as the noise it
is.
|
> ksaun
If you don't mind disclosing, for what type(s) of
games is your game engine intended?(Also: I agree with
your assessment. I'm not working on an engine, but I
am working on a game, mostly/initially just for my own
edification. It's fun to feel empowered to create
again! (My personal circumstances had obstructed me
until genAI's recent progress.))
|
> > RobRivera
It's generalized to the point where "the only
limit is yourself"In other words, its incredibly
modular and lends to configuration flexibility.
|
> SpaceNoodled
I'm not confident that the bottleneck in quality game
development is "content." I can only see this
accelerating the slop and drowning out the good games
in slop.
|
> > RobRivera
What do you believe is the bottleneck?
|
alexpotato
> Why not? I think because it wasn't a good business
decision to compete with me.The owner of Pinboard has a
great story about this:"I ended up buying a competitor.
Why? Because his choice of tech stack + server footprint
cost more than mine. The consequence of this was that even
with each of us charging the same price. I was profitable
and he was not.Do not try to compete against Pinboard"
|
akoboldfrying
I always enjoy hearing the thoughts of someone who took a
slightly different path (indie game development being a
favourite), and isn't committed to advancing some thesis
-- pressing me to love this or hate that. It feels like it
gives my brain a chance to step back from dopamine- or
rage-induced habits and just... connect with other
people.I don't know what will happen either. I hope that
you and I and other hardworking, basically good people
will continue to have a somewhat meaningful, somewhat
pleasant existence in the post-AI world, and I think that
might be possible, but I just don't know.
|
> jensmtg
To have this simple comment revealed, alone, five
years ago, as a glimmer of a future approaching.
|
tantalor
> old school> 2012sigh
|