Pretty obviously vibe coded (by which I mean the html header links to vercel.com), but I suppose if hank came up with the game mechanic on his own then it's... fine? Unsure how to feel about it....
Pretty obviously vibe coded (by which I mean the html header links to vercel.com), but I suppose if hank came up with the game mechanic on his own then it's... fine? Unsure how to feel about it. In this instance it's mostly unfortunate that the guy who claims to be a "big fan of humans" couldn't bother pay one or two to figure this stuff out for him? But like, how important was it that the game got made right now rather than like, 2 or 3 weeks in the future?
Also, the existance of the undo button seems at odds with the "par" system. If you can just undo your non-optimal choices then what's the point?
I mean at this point it's not really an either-or, if you were to hire a developer, of any quality, any experience level, they WILL use claude/codex in writing the app. Depending on style, they...
I mean at this point it's not really an either-or, if you were to hire a developer, of any quality, any experience level, they WILL use claude/codex in writing the app. Depending on style, they may read all of the output, or go full yolo, but they're definitely using some LLM tooling.
Agreed. I work at a software company and the question isn't if, it's how. The difference nowadays is between vibecoding and AI assisted coding. The latter is an experienced coder using it as an...
Agreed. I work at a software company and the question isn't if, it's how.
The difference nowadays is between vibecoding and AI assisted coding. The latter is an experienced coder using it as an extra tool in their belt.
cc @CptBluebear I can understand why you'd say this if you work in tech these days, because general programming work does seem to be sadly inundated with AI-required policies that result in near...
I can understand why you'd say this if you work in tech these days, because general programming work does seem to be sadly inundated with AI-required policies that result in near everyone and their dog using LLMs in that industry. But we're talking about games here, and that's a different industry. So:
Hi, I'm a developer, and I last used an LLM in 2023.
And since we're talking games, I'm very much not alone. There are a lot of indie developers I know, from social media, to personal connections, to local ones, who do not use AI at all. And why would we? We're making art, and we like doing it. There's no good call to offload something human and enjoyable to a machine like that.
There are absolutely plenty of people who could be contacted to develop a web game who would not use an LLM in the process. And if no effort was made to do so, I would find that very disrespectful to both my craft and the result's audience.
I would think some argue the art is in the overall game design and not necessarily just the code. Much like having really good character art doesn't make a game by itself, you need the rest of the...
I would think some argue the art is in the overall game design and not necessarily just the code. Much like having really good character art doesn't make a game by itself, you need the rest of the game too.
Don't misunderstand me, I'm not advocating for replacing art forms with AI. Far from it. I'm saying that it's fairly normal behaviour to supplement your code with AI without the entire project turning into AI slop or stolen valour.
Vibe coding, to me, is writing code without checking or understanding the output. Not whether or not it's used at all.
I'm going out on a limb and say that indie games probably have the most holdouts, I'll give you that.
Corporate gaming however, not just triple A, almost unanimously uses at least some form of it. And if they say they don't, I'm willing to bet one of their programmers slyly does it anyway. Plenty of worker drones that aren't doing it from a place of passion.
Yeah, I do agree with you broadly.1 I just feel it's important to remember that the seemingly-prevalent sentiment of "obviously every programmer uses AI" is hogwash, and plenty don't. It's not...
Yeah, I do agree with you broadly.1 I just feel it's important to remember that the seemingly-prevalent sentiment of "obviously every programmer uses AI" is hogwash, and plenty don't. It's not restricted to artistic endeavors or ethics concerns either — some folks just like to code, and don't want to switch to reviewing output all day.
And personally, even without my many other gripes with AI, I'd be one of them. While I can imagine a world where AI was a viable supplement to my work, it's not the world I live in and I don't think it's one I see coming anytime soon. The only things I want automated out of my coding are things Intellisense / autocomplete have already been doing for ages; I just don't need or want LLMs.
Bit of an aside, by the way, but I also wanted to say thank you for being such an understanding and friendly face when it comes to AI discussions. People like you are a breath of fresh air when it comes to talking to folks who use AI in work or elsewhere.
1. I will say I do think that code can easily be a kind of art, but that's nitpicking, and I do think it's fair to say that other aspects of game development tend to be "more artistic." Well, at least insofar as my feelings are concerned, considering how immeasurable and subjective that is.
Much like spoken language can be turned into books and poetry (therefore art), I think programming can too. But much like language can be turned into purely functional documentation, programming...
Much like spoken language can be turned into books and poetry (therefore art), I think programming can too. But much like language can be turned into purely functional documentation, programming can too. The intent matters.
No you're right, an absolute statement like "every programmer" should be avoided. I tend to overdo it on those since I always think it inherent that there are exceptions or even large subsets where the rule doesn't apply. Even so, when looking at the broader picture I think it's safe to say that at this stage most working folk are using AI in some form or another.
When the output is not made out of love or a passion for the job, it's easy to grab the shortcut. It really did come a long way, even compared to last year, where a lot of people see value in scaffolding with AI.
I also don't like structuring the layout of a business plan, but I'm good at writing out the content. I let AI do the former and do the latter myself.
Thank you for saying that. I don't particularly like the tech for its inconsistency, its general effects on the world (economically, ecologically, ethically) and I feel that politically we aren't ready for the lightning fast advances it's promising (both true and false promises).
But I also think it's here to stay. A system that can (cheaply) create foundational frameworks for humans to build upon is a value add when it comes to menial tasks.
Non-users will have to interface and contend with the output of AI no matter how much they don't want to. Unfortunately we don't have any hard and fast social rules on what to do when it happens so the absolutists and extremists shout the loudest.
I try to understand both sides of the argument and push back on either. I grumble and somewhat reluctantly use it myself and when it's done thinking I grumble and reluctantly accept that it can now finally create a genuinely useful document for my work while it burned a forest.
It doesn't feel vibe-coded, tbh -- like his other games, it's polished, works well, and feels fun enough to return to. My biggest criticism for this one is that it feels like one of those...
It doesn't feel vibe-coded, tbh -- like his other games, it's polished, works well, and feels fun enough to return to.
My biggest criticism for this one is that it feels like one of those test-tube color sort mobile games rotated 90 degrees, albeit with some extra mechanics to make things more interesting.
The game itself is fine, though some of the text feels like an LLM wrote it: The phrasing just feels off to me. "solver's line"? 100% agree. It's not a bad game, but it's definitely not original....
The game itself is fine, though some of the text feels like an LLM wrote it:
Like colors with Like. Pull from the right. Unscramble the board.
Matched the solver's line exactly, 2 under par.
The phrasing just feels off to me. "solver's line"?
My biggest criticism for this one is that it feels like one of those test-tube color sort mobile games
100% agree. It's not a bad game, but it's definitely not original. Feels sort of like going to McDonald's: yeah, it's a burger, it tastes exactly like you'd expect, but unless you're craving junk food or nothing else is open, there's nothing compelling that will bring you back.
He shared via Bluesky recently he had in fact brought on someone to build and maintain his games. https://bsky.app/profile/hankgreen.bsky.social/post/3my6i5ttivk2r
He shared via Bluesky recently he had in fact brought on someone to build and maintain his games.
This is such an annoying discussion to have. It doesn't matter in the least. Dude is just putting a neat little game out there. Now liking humans means you have to do things some inefficient way...
This is such an annoying discussion to have. It doesn't matter in the least. Dude is just putting a neat little game out there. Now liking humans means you have to do things some inefficient way just to pay someone? Should we scrap all construction vehicles and do stuff by hand? I get it, you dislike AI, but do we really need to point out every usage of it.
There are a lot of things that could be commented on here, as you've got a lot of assumptions in your post. But I can agree that this is an annoying discussion, so I'll be brief: AI ethics issues...
There are a lot of things that could be commented on here, as you've got a lot of assumptions in your post. But I can agree that this is an annoying discussion, so I'll be brief:
AI ethics issues make first-resort use of it careless at best; AI is frequently inefficient, especially at art, which games are; and our society considers work required to deserve life, thus paying work becomes a necessary good.
A lot of us care quite a bit about these factors, so this is disappointing at a minimum.
Personally I like the undo. I can try out one option of how to proceed, and if it leads to a stuck position I can back out until I see another option, and just slowly explore different moves until...
Personally I like the undo. I can try out one option of how to proceed, and if it leads to a stuck position I can back out until I see another option, and just slowly explore different moves until I find the best answer.
Take away the undo, and I have to figure it all out in my head beforehand.
Another fun entry into the hankgreen-verse! I really like 4x3 and this one too - though the 'tools' feel unnecessary. 4x3 has 0 chill, the 3 tools here are a little cheaty. Which I'm saying as...
Another fun entry into the hankgreen-verse!
I really like 4x3 and this one too - though the 'tools' feel unnecessary. 4x3 has 0 chill, the 3 tools here are a little cheaty. Which I'm saying as someone who has been squinting at my first 'real' 4x6 for fifteen minutes.
I did really appreciate the warmup! Taught the rules very well. Fun time waster!
Edit: here's 4x3 btw in case people are interested. It's a clever word-combo one!
Pretty obviously vibe coded (by which I mean the html header links to vercel.com), but I suppose if hank came up with the game mechanic on his own then it's... fine? Unsure how to feel about it. In this instance it's mostly unfortunate that the guy who claims to be a "big fan of humans" couldn't bother pay one or two to figure this stuff out for him? But like, how important was it that the game got made right now rather than like, 2 or 3 weeks in the future?
Also, the existance of the undo button seems at odds with the "par" system. If you can just undo your non-optimal choices then what's the point?
I mean at this point it's not really an either-or, if you were to hire a developer, of any quality, any experience level, they WILL use claude/codex in writing the app. Depending on style, they may read all of the output, or go full yolo, but they're definitely using some LLM tooling.
Agreed. I work at a software company and the question isn't if, it's how.
The difference nowadays is between vibecoding and AI assisted coding. The latter is an experienced coder using it as an extra tool in their belt.
cc @CptBluebear
I can understand why you'd say this if you work in tech these days, because general programming work does seem to be sadly inundated with AI-required policies that result in near everyone and their dog using LLMs in that industry. But we're talking about games here, and that's a different industry. So:
Hi, I'm a developer, and I last used an LLM in 2023.
And since we're talking games, I'm very much not alone. There are a lot of indie developers I know, from social media, to personal connections, to local ones, who do not use AI at all. And why would we? We're making art, and we like doing it. There's no good call to offload something human and enjoyable to a machine like that.
There are absolutely plenty of people who could be contacted to develop a web game who would not use an LLM in the process. And if no effort was made to do so, I would find that very disrespectful to both my craft and the result's audience.
I would think some argue the art is in the overall game design and not necessarily just the code. Much like having really good character art doesn't make a game by itself, you need the rest of the game too.
Don't misunderstand me, I'm not advocating for replacing art forms with AI. Far from it. I'm saying that it's fairly normal behaviour to supplement your code with AI without the entire project turning into AI slop or stolen valour.
Vibe coding, to me, is writing code without checking or understanding the output. Not whether or not it's used at all.
I'm going out on a limb and say that indie games probably have the most holdouts, I'll give you that.
Corporate gaming however, not just triple A, almost unanimously uses at least some form of it. And if they say they don't, I'm willing to bet one of their programmers slyly does it anyway. Plenty of worker drones that aren't doing it from a place of passion.
Yeah, I do agree with you broadly.1 I just feel it's important to remember that the seemingly-prevalent sentiment of "obviously every programmer uses AI" is hogwash, and plenty don't. It's not restricted to artistic endeavors or ethics concerns either — some folks just like to code, and don't want to switch to reviewing output all day.
And personally, even without my many other gripes with AI, I'd be one of them. While I can imagine a world where AI was a viable supplement to my work, it's not the world I live in and I don't think it's one I see coming anytime soon. The only things I want automated out of my coding are things Intellisense / autocomplete have already been doing for ages; I just don't need or want LLMs.
Bit of an aside, by the way, but I also wanted to say thank you for being such an understanding and friendly face when it comes to AI discussions. People like you are a breath of fresh air when it comes to talking to folks who use AI in work or elsewhere.
1. I will say I do think that code can easily be a kind of art, but that's nitpicking, and I do think it's fair to say that other aspects of game development tend to be "more artistic." Well, at least insofar as my feelings are concerned, considering how immeasurable and subjective that is.
Much like spoken language can be turned into books and poetry (therefore art), I think programming can too. But much like language can be turned into purely functional documentation, programming can too. The intent matters.
No you're right, an absolute statement like "every programmer" should be avoided. I tend to overdo it on those since I always think it inherent that there are exceptions or even large subsets where the rule doesn't apply. Even so, when looking at the broader picture I think it's safe to say that at this stage most working folk are using AI in some form or another.
When the output is not made out of love or a passion for the job, it's easy to grab the shortcut. It really did come a long way, even compared to last year, where a lot of people see value in scaffolding with AI.
I also don't like structuring the layout of a business plan, but I'm good at writing out the content. I let AI do the former and do the latter myself.
Thank you for saying that. I don't particularly like the tech for its inconsistency, its general effects on the world (economically, ecologically, ethically) and I feel that politically we aren't ready for the lightning fast advances it's promising (both true and false promises).
But I also think it's here to stay. A system that can (cheaply) create foundational frameworks for humans to build upon is a value add when it comes to menial tasks.
Non-users will have to interface and contend with the output of AI no matter how much they don't want to. Unfortunately we don't have any hard and fast social rules on what to do when it happens so the absolutists and extremists shout the loudest.
I try to understand both sides of the argument and push back on either. I grumble and somewhat reluctantly use it myself and when it's done thinking I grumble and reluctantly accept that it can now finally create a genuinely useful document for my work while it burned a forest.
Regardless, it's appreciated.
It doesn't feel vibe-coded, tbh -- like his other games, it's polished, works well, and feels fun enough to return to.
My biggest criticism for this one is that it feels like one of those test-tube color sort mobile games rotated 90 degrees, albeit with some extra mechanics to make things more interesting.
The game itself is fine, though some of the text feels like an LLM wrote it:
The phrasing just feels off to me. "solver's line"?
100% agree. It's not a bad game, but it's definitely not original. Feels sort of like going to McDonald's: yeah, it's a burger, it tastes exactly like you'd expect, but unless you're craving junk food or nothing else is open, there's nothing compelling that will bring you back.
He shared via Bluesky recently he had in fact brought on someone to build and maintain his games.
https://bsky.app/profile/hankgreen.bsky.social/post/3my6i5ttivk2r
This is such an annoying discussion to have. It doesn't matter in the least. Dude is just putting a neat little game out there. Now liking humans means you have to do things some inefficient way just to pay someone? Should we scrap all construction vehicles and do stuff by hand? I get it, you dislike AI, but do we really need to point out every usage of it.
There are a lot of things that could be commented on here, as you've got a lot of assumptions in your post. But I can agree that this is an annoying discussion, so I'll be brief:
AI ethics issues make first-resort use of it careless at best; AI is frequently inefficient, especially at art, which games are; and our society considers work required to deserve life, thus paying work becomes a necessary good.
A lot of us care quite a bit about these factors, so this is disappointing at a minimum.
Personally I like the undo. I can try out one option of how to proceed, and if it leads to a stuck position I can back out until I see another option, and just slowly explore different moves until I find the best answer.
Take away the undo, and I have to figure it all out in my head beforehand.
Another fun entry into the hankgreen-verse!
I really like 4x3 and this one too - though the 'tools' feel unnecessary. 4x3 has 0 chill, the 3 tools here are a little cheaty. Which I'm saying as someone who has been squinting at my first 'real' 4x6 for fifteen minutes.
I did really appreciate the warmup! Taught the rules very well. Fun time waster!
Edit: here's 4x3 btw in case people are interested. It's a clever word-combo one!