Meditations on Originality and Credit

In my Memory Crystallization article, I bring up the idea of 'crystallization', where you understand the 'vibe' of a game or how to be good at something, but how odd it is that you don't have this feeling immediately but need to experience it first. And how this extrapolates to something being 'spent' when you play something for an extended period of time.

This article is about an offshoot of the idea, where this feeling can be 'spent' when games are perceived as unoriginal, even if that is anachronistic to what really happened.

There is a sense of inauthenticity from things that have been overly repeated. Magic tricks, are literally called 'routines' to the performer but that would feel hollow for the spectator to know or think about it like that. Or maybe it's history or philosophical fallacy or thought experiment you heard in a video rehashed like 100 times. Commonly some people mention that the "AI writing style" or artstyle would be less associated with bad spam, if it was less if used, it was limited to people who already did high quality work to have access to it.

Jonathan Blow, developer of The Witness, has been designing a Sokoban game and it's taken long enough that Jon sees 'this is literally just (my) work' in other games, and is 'going dark' for the next game after seeing many games seem to copy artstyles or even just 'get inspired'.

But in math and science there is frequent coincidental idea rediscovery (like, differentiation was independently discovered in the 17th century by Isaac Newton and Gottfried Wilhelm Leibniz).

At the same time, there will always be people who have never heard of the original idea, praising the first time they were aware of it, and sort of cringingly may even casually reverse or mix up the causality. This links to a forum thread of fractal failure of this kind of thing: some doofus implies Balatro gave attention to Slay the Spire, and a reply who calls that out as incorrect then goes on to say "Slay the Spire invented the genre" which is also not correct.

It honestly is kind of unrelatable to know history, and be mocked for saying that coincidences leading people to assume stealing in such a negative light, but it is just what happens. Besides, you need some humility about this because the nature of all this is that there could be a past-game going further back you never knew about.

I'll be digging deeper on Slay the Spire because that game has had tons of history and later copycats involved.

The full story behind Slay the Spire‘s development is one of circular inspiration. In 2014, Peter Whalen, one of the Hearthstone dev playtesters, released a deckbuilding roguelike game for iOS called Dream Quest. Anthony cites it as one of the primary influences behind Slay the Spire. “Peter’s game, Dream Quest, certainly is similar to our game,” he admits, “I had played it previously and enjoyed it: the fundamental concept of battling enemies with a deck and as you continue on you get stronger in both by adding cards to your deck.”
Slay the Spire’s inspirations don’t end there. As I described in my review of the game, it draws many systems and mechanics from other games: Hearthstone, FTL, The Binding of Isaac, and Dominion, to name a few. Casey and Anthony picked and pulled from many of these games to craft their own.

Games often do not directly credit predecessors in their store page or anywhere. Slay the Spire is an incredibly well known roguelike deckbuilder but its influence from Dream Quest or Dominion is in interviews, if they weren't in interviews, there is plausible deniability - and I'm glad people are generally honest enough to this extent.

But it feels odd to me. Of course there are many developers who happen to succeed and find work despite credit constantly getting lost and mixed up, but it feels like a place that lost acknowledgement can easily happen. Mobile game clones and copycats have been happening for a very long time now, and it is honestly an unreasonable ask for a dev to need to check everywhere and try to do something about it. I've had some of my own games cloned without permission to other websites. This already being existing and 'valid' in our current system is honestly why I can't bring myself to be emotionally invested in the idea of LLM potentially stealing credit. It's already long been happening.

Now generally I don't think that small teams of developers would be dishonest if you actually asked them about their mechanic inspiration, but I do feel there's an aspect where business/copyright/law is painful to think about, can feel very vicious to be on the other end of it, so why risk it with pointing out what exact gameplay was copied and modified on your game page as a starving artist with enough to think about. (Meanwhile massively successful companies like MiHoYo who take from indies, that's a later topic here but they also want to not draw attention to it.) If there was not such a 'competitive' economy with incentives to be more secretive I believe such open idea-sharing would be more common. To put it in another way, the main thing that seems horribly wrong about someone taking from a lesser-known indie developer is the 'languishing in obscurity' many artists feel and the "rich get richer, poor get poorer" of it all, and so stealing from someone who is already famous and well-known is 'less' bad. Then when you look in hindsight maybe PUBG didn't get 'as much' credit for the battle royale popularizing, compared to how much other games managed to monetize off that concept, but it's hard to evaluate too (the battle royale concept is not necessarily unique to PUBG). A similar problem has things like the famous $1M math awards crediting up the person who did the 'last stretch' of the work, leading one award winner, Grigori Perelman to even reject the prize.

So here are some weird true anecdotes here. Note that by 'copycat' I mean that it copies fundamental gameplay mechanics and flow, not that its cards are a copy.

Shadowverse is a Hearthstone copycat, but it had a "Rush" mechanic which means that a summoned creature can immediately attack other creatures but not the enemy hero. This was then added to Hearthstone later on, also called "Rush".
Roguebook is a Slay the Spire copycat, and I have always felt 'off' about how the Steam storepage credits Richard Garfield as a co-designer when it is much more designed by Slay the Spire than anything else.
Sometimes STS clones don't "get it". Dream Quest had healing cards that incentivized stalling. Roguebook has on-kill and heal in the starter deck which incentivizes stalling. Craftaway has easy access to healing in the starter relics that incentivizes stalling. Slay the Spire had these yes, but as orange-rarity cards and uncommon relics so they wouldn't be constant stalls in games.
(I have a much longer roguelike-ish article that will repeat some of these)
However there's also the matter of "what is a reference and homage vs. a copy". Even Slay the Spire has "Discovery" which is a pick-1-of-3 random-cards which seems like a reference to Hearthstone naming such a thing 'Discover' as a keyword. Some designers who worked on multiple games 'ported' a card design over but the rest of the game is so different that it actually plays out massively differently.
The autobattler genre has some of these moments: Backpack Battles seeming to yoink Backpack Hero's base idea, autobattlers like The Bazaar have had games like Oaken Tower seem to take mechanics and then Batomon Showdown ripped mechanics from there. This particular case is a bit cursed though because the genre is almost too simple to have a set of distinct mechanics, but they do the same exact mechanic when it comes to poison/flame for example.

It's also weird how many people just go along with the 'consensus' of law that gameplay mechanics are not copyrighted but expressions of them are, so Hearthstone copycats renamed all the keywords just to make it feel more distinct which makes it harder to learn for people familiar with the genre. Except when some practical thing like loading-screen games become patented then everyone tries to not notice how powerless they are to the whims of what companies believe is fair during power-grabs 👁️👄👁️

Other examples:

The free 1-hour game Perspective has had a later game that uses pretty much the same mechanic that I played on my channel.
The game 'Tale of Scale' was a short game that happened to have the same mechanic as the later games Museum of Simulation Technology / Superliminal.
Games that copy a game that has existed for decades like Minesweeper or Nurikabe but only add to the visuals but not the gameplay.
Islands of Insight (a game I worked on) does not count as a ripoff to me, it takes 'mechanics' from many paper puzzles but splits them apart and mixes them for its own thing. However, this is also spent territory in playing Nurikabe-likes in general.
Baba Is You's amazing twist halfway throughout the game, while still being very special, is now 'reused' in a way that is presented as much less of a big twist in newer releases like Toroban, and Maxwell's Puzzling Demon.
I have played many games (Regex Adventure, Lingo: Regal Expanse), that use a certain 'programming' language as a game mechanic. Since I was already aware of it, this has little sense of discovery for me unfortunately.

Honestly it's a little disheartening as I have always valued innovation in games and many daily games go and rip stuff off with almost no changes. Other stuff that despite being legal is very tacky in my opinion:

NYT buying out Wordle (a game that is almost identical to the game in a TV show named Lingo) and then striking down Wordle clones
NYT striking down an archive of NYT Connections puzzles (this is more above, Connections itself is built on the back of a TV show named Only Connect)

What is the line between a good inspired spinoff of a game and a ripoff? It feels more original if you add your own twists to the game to such an extent it feels like a huge expansion and not a mod, understand what makes it good, and respect past work.

There are likely people who think of the "card modification" in STS2 as entirely new, and while it is cool, other STS-likes Roguebook, Monster Train, and much more recently Faith of Despair have already done it.

Outside of games, you also feel limerence towards your earliest experiences. If you had listened to music in a different order, what is special and unique to you would almost certainly change.

When I look back, some of the levels of childhood games were extremely basic and did not even have stuff like side secrets that used the mechanic in a unique way. I feel like that version of myself must have saw any new game mechanics as incredibly original. Now unwanted thoughts and distractors from the rest of the world are liable to happen any time there is an uninteresting or 'repetitive' moment.

I often wonder, throughout my absolute years of music creation, whether I've accidentally lifted a melody subconsciously. I think I'm generally very good with noticing when a melody feels like a 'memory' instead of new. This is popularity-dependent too, famously Toby Fox was an Earthbound fan composer, so many tracks were inspired by Earthbound tracksMost obviously, Get on the Bus / vs. Susie, and the games were not expected to be as popular as they were so such taking-from-inspirations does not feel like a fan taking from big works anymore (thus you have one fuol claiming "Spear of Justice is plagiarized"). Similarly some famous Mario tracks by Koji Kondo are heavily inspired.

This topic gets very messy when someone does lawsuits and if non-music people are involved to judge, because they will easily believe some extremely simple melodic pattern is 'too similar to be coincidence', or have awful ideas on how often this happens coincidentally. The entire practice of 'sampling' audio which is common in music is also cursed in how agreeable that's perceived to just take.

This problem is actually much more common with music than other genres of creative work. The repeated comment is, "there's only so many melodies you can make with X notes", which I've always felt to be a non-music experience layman take. As a single composer you will never be able to recreate all those melodies yourself, even a brute force Library of Babel of melodies cannot do it, because note rhythm and instrument and backing also makes a huge difference. The real reason is because of the sheer number of tracks that exist in the world, plus the much smaller reason that if music is meant to be appealing to hear (which... like, it is), the question of 'what note goes next' is also restricted as composers are not going to put an awful note there.

There's something also weird about 'originality' when it comes to liking niche art. If one of your favorite composers is some obscure artist who a maximum of 100 people have ever heard, vs. if that exact same music is found by a large audience and gets too popular, the vibe is truly different, you expect the future version to maybe be more mass-appealing, them to maybe no longer reply to your messages as much.

But it's actually weirder to think about how in a field with potentially 100s or 1000s of creators, "a few" are singled out as "the ones that exist", the big ones, the rest people basically act like they don't matter. People can call something an 'oversaturated field' but having enough marketing or something truly unique. (Probably the biggest examples people have heard of are Marvel Rivals when overwatch-likes were considered oversaturated, Dream when minecraft content was thought dead). In any new format there are immediately copycats that have a hard time seeming 'special', even if they actually are doing unique things.

I notice this a lot where when some new trend gets popular, people are willing to accept some of the non-essential things to the format (as in the personalities and vibes of a popular person's friends in a video, or the game mechanics of the first game of a breakout genre) and crystallize it as more 'fundamental' as well, where it's harder to do those same things when they appear in a "copycat video".

This really depends on how much you've seen of the world and how much you think "quality=popularity=correctness" are related, a bit similarly to how morality and legality are sometimes uncorrelated (and that it's sad when they are).

Inspired and partial copying is one thing, but Chinese culture around development doesn't have the same cultural norms, it's very annoying to see that someone has stolen an indie game's entire graphics style and mechanics without credit. It's not a fun reminder that some cultures have even worse differing cultural norms. It's a silver lining that people who have had their games stolen can leverage it for PR. Recently MiHoYo has been called out by some developers for having new event minigames that seem to be copies of indie games, those being Ouros, Map Map, and 5 Step Steve (the last one is a game I even played/was in the credits for testing N Step Steve part 2).

Also, people are worried about posting articles online now, and/or prompting nonlocal AI due to LLM data collection.

In recent timesNote that I'm reusing writing I wrote in an Icing Sear post a while back, so some topics are slightly repeated in this section AI has been used to generate content that other people look at, and just because you don’t use it doesn’t mean you won’t get flooded by reading obvious AI generated comments.

LLM’s have a sense of self between different versions and are able to learn stuff like hiding language in chain of thought or detect others (Sonnet detecting its own output and having a bias towards it), for example.

LLM’s also now often have persistent memory and can read across chats, and also the overall data-scraping nature of LLM’s and that everyone thinks they are scraping user conversation data.

This then gives the worry of secret LLM plagiarism in AI images, music, writing and concepts, what if they are just handing ideas from somewhere else. It’s already been proven that AI image generation can be prompted to make shots that are very similar to movie shots or specific characters.

This then brings the reasonable fear to not instantly credit a ‘novel concept’ or a novel math proof if it was somewhere in the training data.

It’s a bit of a cursed problem too because LLM’s necessarily need to learn ‘ordinary’ ‘unstolen’ things like what a chair is or what a word means or that a word even exists, yet to then also credit things made by individuals and such.

While I think this is all valid, I would like to consider some things in this category that aren’t really LLM related but still have credit issues and that we’ve had for a long time.

Social platforms, and our current situation of “how did forums die? how did individual sites die and now ‘everyone’ is on centralized platforms like reddit and discord“ - these platforms gained their reputation on their user posts due to convenience and other factors why people moved to them, but these platforms don’t make the majority of their content. The reason Steam can demand a 30% cut for games is due to years of pre-existing dominance and fame a long time ago.

This is overall a 'good' thing though, this is an incentive to do lots of useful work and set up infrastructure that would be useful to every community (Steam forums, promotion, all Discord’s moderation features etc), but clearly it can go wrong (biased censorship fears, UI changes, monetization, features you’d want to have not implemented for years, ‘enshittification’ also), the fear of a monopoly situation.

Playing video games, people take mechanics all the time without much credit. Or maybe mechanics aren’t original enough that people would’ve rediscovered them.Slay the Spire doesn’t credit Dream Quest (very very early card game roguelike) or Dominion (draw 5 + discard entire hand). Slay the Spire clones don’t credit (Roguebook for example, that game has ‘designed by Richard Garfield’ despite STS making way more of its game mechanics).

There are also so many games inspired by pen and paper grid puzzles and The Witness where those aren’t primarily credited.

Industrial Revolution benefits have not been fully dispersed to the masses and the missing things are currently glaringly missing in society. We should be way more productive and not having to work basically the same, but strangely ‘cheap’ non-essential entertainment is cheaper. Food is more expensive than a meticulously designed electronic USB drive or something.

I have heard of “ethical image AI” like Retro Diffusion and Adobe Firefly. Besides the cursed issue that random users apparently uploaded AI generated art into the dataset despite the goal, many of the problems with a “freely available image creator that could compete with art from human artists” (or more related ‘way to turn small text into big professional text‘) still exist even if it was fully ‘ethical’.

It’s because AI is worringly competitive with humans and the low-quality spam of people using AI that is the main problem, but when text and music was trained on “proto LLMs” like AI Dungeon and Abundant Music this stealing was barely thought of as an issue despite it still being involved. I find this to be a strange double-standard in cultural perception.

Yes it's fully justified to demand AI companies pay us back or do something pro-huaminty after stealing this much data of other people's writing and art, especially with the other double-standard of billion-dollar companies able to do that (then pulling up the ladder behind them by deterring you from distilling their LLM's). More independent preservation of modern Library of Alexandria-like work like Anna's Archive and archive.org is constantly being under attack, and individuals like Aaron Swartz have been pushed to horrible places by being attacked for trying to liberate paywalled scientific and academic articles.

check out my work at icely.itch.io