15
votes
Let me tell you why only a dumbass would get into NFT games
Link information
This data is scraped automatically and may be incorrect.
- Title
- look what you made me do: a lot of people have asked me to make NFT games and I won't because i'm...
- Authors
- GB 'Doc' Burford
- Published
- Nov 19 2021
- Word count
- 7025 words
A long, profane rant about why, as a game designer, NFT's are a bad idea.
[...]
[...]
I don't really think the game design parts of this argument make all that much sense, because this pretty much already exists. There's a gigantic CS:GO weapon skin market, for example, where CS:GO skins are scarce, people pay massive amounts of money for them, the market is surprisingly liquid, and people use them for money laundering and avoiding regulations (e.g all the CS:GO skin online casinos). Lots of very popular and successful games that have mechanics or even revolve around FOMO of scarce digital items!
It may be against his personal aesthetic - a very understandable position - but clearly people eat it up, and it's not like only shady developers do it (Valve is a reputable developer by most metrics!).
That's a very very common feeling for people to have in both games and real life!
The more convincing one is just... yeah, Valve did it with a database and some business logic on their backend, why do you need a blockchain. What about items that persist between multiple games? Surely the developers of those games can collaborate? If they're like actively hostile and trying to screw over the other I don't see why they'd want to share digital doodads at all, so for this premise to exist they must want to collaborate?
Well, that's for public blockchains. For a group of closed individuals (like a handful of game companies), a private hashchain where you do consensus with BFT is both efficient and not the worst idea, although again I'm not quite sure why you'd collaborate with a company you trust so little that you think they might fabricate your shared digital items.
Part of his argument seems to be that it's just hard for different game companies to collaborate even if they wanted to, due to things like IP licensing issues, different game mechanics (otherwise they'd be clones), improvements to game technology over the years, differences in business models making it unclear who pays for what and how to split revenue, wanting to control how their own game evolves, and so on.
There are certainly lots of games with mods but mods aren't portable. The question is how to come up with standards so that any game engine that implements the same standard isn't basically a clone as far as game mechanics are concerned.
The CSGO skins thing is talked about directly in the article, and a specific situation that happened around them in the professional scene too. He actually says the same thing you do here, that the NFT approach is exactly the same thing that happened with the CSGO skins, specifically like the scandal that happened with them:
Lots of games may be exploiting this FOMO and generating money out of it but they're not exactly well-respected for it. I'm not sure how the success of capitalistic exploitation trumps the argument of capitalistic exploitation being a pretty shady thing to do and is easily exploited from the shadows.
The actual argument here isn't about the viability of NFTs in games, it's about the ethics of it. The full paragraph of that quote there actually has the thesis in bold:
A consortium of game companies running a private chain for managing digital game goods might be pretty cool. New game studios can opt into at least respecting the rules of the chain, and can eventually run mining infrastructure if they feel they wish to help secure the chain.
If it's a private chain you don't need to use PoW, and instead can use one of the traditional algorithms for the Byzantine Fault Tolerance problem like pBFT.
TBH, if it’s a private chain, you might just go for a more standard DB and not waste all the energy.
The two most important aspects of a blockchain compared to other DBs are 1) all transactions are recorded so no data is lost, and 2) it is built for a situation where you need to trust the system, since you cannot trust all the actors, so the votes are divided up. If you have enough votes, you can still rewrite stuff, so a centralised chain where the votes are all in the hands of one (or very few) is in effect very close to just a standard DB.
The hypothetical with a private chain is, say, that there's a 6-7 large group of major game companies (EA, Activision, etc). They trust each other, since they're big and with reputation to lose and things that can be sued.
They want to add in a smaller dev to their virtual doodad ecosystem. But they're worried about said smaller dev counterfeiting rare doodads or whatever. Probably just wouldn't want to work with them, but if they did, so long as the unknowns are a small enough proportion of the voters (which can be done for any N smaller devs just by giving the big boys more votes), they can trust in the system that if they allocate 20 doodads that the smaller dev can make, that will be how many doodads they mint.
Probably contrived but a genuine niche.
In terms of energy, it'd be on the usual scale of "doesn't matter" - pBFT and it's brethren aren't particularly expensive. Whether or not one of those game servers are written in python or go would be a bigger detriment.
There are actual production databases that effectively use a hashchain + pBFT to maintain consensus on the DB journal across multiple nodes like https://bedrockdb.com.
Does Bedrock DB really use a Byzantine algorithm to achieve consensus? I was able to confirm that they use Paxos but I haven’t found anything about them using Byzantine Paxos.
I was under the impression that distributed databases don’t usually guard against Byzantine faults. Paxos or Raft is usually good enough.
I still fail to see why your usecase needs a blockchain.
(sorry for being terse, on mobile and it' late)
At this point blockchain has become too charged to be a useful phrase.
Why do we want a hashchain? Immutability, that way once everyone has a record up to X point, everyone can be sure that, if the hashes follow each other, everyone has the same records. That's not too strange - that's in effect part of what a branch in Git is. If you just have a list of records, it becomes easier to make mistakes. Let's say we have the records "insert A; mutate B; insert C". If one node just gets "insert A; insert C" - they can easily tell that something has gone wrong, because once you start hashing things they don't match up (without B as the predecessor the hash for C would be wrong).
It's a useful property, and practically free on modern computers to run the hashes.
Why a consensus algorithm? Well, I mean, this is the definition of the byzantine problem, no?
This type of problem comes up in many areas (although the non-byzantine family of paxos algorithms is more common given that it's usually more "one or more machines in a group is dead" rather than actively malicous).
Together it's a set of fairly common technologies, that's efficient to run, which accomplishes attributes that are desirable for the situation (easy to verify correctness, append only, some resilience to attackers) of inviting entities you don't fully trust to add to a journal.
Perhaps you could think of it like a Github repo with an overly complicated PR system - the master is the hashchain that everyone agrees to, and to do pull requests the set of maintainers have a somewhat complicated voting system to agree to merge or not merge.
I understand the concept, but thank you for explaining the Byzantine problem to me – I haven’t found the time to look that one up and it’s been popping up often.
The issue I have with it all is that while it may solve the problem, the question is whether it’s 1) an actual problem in the use case and 2) it is the only and best solution.
E.g. going back to your chain of doodads:
Why would IndieDev need full access to the DB of doodads? How does the blockchain prevent IndieDev from “counterfeiting” assets (they can still copy and change a pixel for it to have a different hash value)? Why would they even need a vote in the internals of the DB? Especially if their vote is pondered down.
If BigGamesGroups don’t trust them, there’s many other ways for them to keep the DB and assets in check, and giving IndieDev the needed read and limited write access via an API to a normal DB. You can set as many limitation or permissions there as you want.
If the issue is with IndieDev not trusting BigGamesGroups, then with their (esp. pondered down) minority vote, their influence is still negligible. Perhaps the idea is that all the IndieDevGroups together could have more votes than all the BigGamesGroups, but I don’t see that scenario IRL in the first place. And even if, it would probably result in a fork. Again, nothing here that the blockchain prevented or made easier.
Maybe the issue is that we want to make sure that the transfer of such rare and expensive doodads is trusted. And here you just need to look at all the BTC/ETH/NFT/… hoaxes and scams out there – the blockchain not only does not prevent them, but the immutability actually facilitates them and makes any remediation much much harder. You can trust that the transaction happened and what it was, but pretty much everything around that is wild wild west.
I’ve been there when Bitcoin Pizza happened, and while I agree that blockchains are a very powerful technology (if implemented right!) I’ve yet to see an actual problem that a blockchain solves better than pre-existing solutions.
When it comes to GitHub, it’s totally possible to rewrite Git history. As long as people don’t clone or fork before, don’t notice or just don’t care enough about it. (For an actual blockchain example see also The DAO, also interesting from the voting, transaction reliability, scamming, etc. PoVs).
Also in general, when we’re talking about serious load (even those NFT JPEGs), it is often not on the blockchain itself, but on a separate medium and the transaction in the block just points to the load (e.g. via URL).
Again, I think you're letting "blockchain" bleed through.
For instance, yeah, you can rewrite history in Git by just accepting a different chain. And so you can in this scheme! You just need a majority of votes to take the new chain. But it means that for a given chain, you can't rewrite history, and provides an error detection for the integrity of said chain. Additionally, the chain is only really used as a journaling mechanism - it doesn't mean you can't do undo things. DB journals are also append only for the purpose of ACID compliance, yet as you have noticed, there is no issue undoing things in that scheme because "undoing" a transaction is business logic - these are disconnected concepts.
After journaling the business logic records of "X did Z to Y", they would store the records with whatever database they want. That doesn't really have anything to do with each other. The chain record is on a higher level of abstraction in this case, since we're not trying to use it as a general datastore as in a blockchain currency.
The difference is that this assumes complete trust in that BigGameGroup as a collective by both the members of the BigGameGroup and the indie devs. The private chain scheme would only require implicit trust of however large amount of members the BFT solution chosen requires.
It also requires that, at some point, a single entity, presumably either one of the big corporations or a new group formed by it, actually maintains that API on a day to day. Whereas, for transparency purposes, each member may wish to run their own instance. In which case, you could just have everyone run a Paxos BFT algorithm without the hashchain layer but you have to store them somewhere, seems like a light lift to run and store a SHA256 hash per record for data resilience reasons.
Again, it is a contrived situation - the more likely situation in reality is something like Valve workshop, where Valve manages the digital doodads and indies just gotta roll with it. If companies want to partner on a deeper level that Valve doesn't trust... then they don't add them, because it's probably not that much of a value add to add potentially hostile companies to the ecosystem.
There's always bedrockdb - hard to say it's straight up better than a distributed MySQL database, but it does provide different benefits and is battle tested.
Yeah sorry, I wrote "mining" there without thinking too hard about what I was writing. Indeed you don't need to use a PoW algorithm here so you don't need to mine.
I'm curious about what the point is of posting these sorts of articles on this site. I did a simple search on here about NFTs, and almost all of the sentiment toward it is negative. This is another negative post, this time profanity laced, about NFTs. Is the goal to bring out arguments in favor of NFTs? I don't see what the purpose of posting content like this is.
I find a lot of finance stuff fascinating even if I wouldn't participate myself. For example, I'm a big fan of Matt Levine even though he often writes about crazy stuff, sometimes literally fraud. I guess it's sort of like how movies about bank robberies are popular?
Also, sometimes the crazy stuff involves edge cases, which tend to raise philosophical issues. That's kind of fun too.
Cryptocurrency is at the intersection of finance, computer algorithms, and people doing dumb stuff. As a spectator, what's not to like? (Well, mostly a spectator. I do have a small amount of Ethereum that I'm not doing anything with.)
Game design and the game business is another thing I'm interested in even though I play very few games most of the time, and I'm quite out of the loop. So I thought this rant was interesting and don't mind the profanity too much.
But admittedly, the author does hold cryptocurrency in contempt. If someone thinks the author is making bad arguments I think it would be understandable to be upset over it. I'm not sure I really want to get into an argument over whether it truly deserves contempt, though?
And then there is the question of what's worth posting on a general interest site, which I'm quite unsure about. I just thought it was interesting.
Thanks this actually makes a lot of sense to me. I just wished the author had used more novel arguments, but fair enough, and some of the discussion being generated is pretty interesting. Cheers!
A lot of people on Tildes have financial interests staked in crypto and NFTs, so these sorts of articles are a pushback to the positive sentiment driven by those members. It's also a pushback to the general excitement in the tech industry over crypto.
Could you quantify this? I took a search on Tildes and found almost universally negative sentiment.
Where is this positive sentiment? It's not on HackerNews nor on Lobste.rs. News articles outside of the cryptocurrency space routinely dump on NFTs and cryptocurrency as a whole. I guess if you hang out on cryptocurrency Twitter you'll find a pro-NFT bubble, but you can probably find a pro-X bubble for anything on Twitter.
I'm wary of repeatedly posting articles that dunk on the same viewpoint and are largely agreed upon by a community. At its worst, this sort of behavior creates an out-group (in this case, folks in the NFT world) that can be rallied against in the in-group (Tilders?) to create a sort of group identity and echo chamber dynamics. At its best it's rehashing the same viewpoints over and over again and crowds out more interesting/nuanced discussion.
In pretty much every other technical site I'm on, cryptocurrency discussions are largely low-quality emotional rebuttals to low-quality articles criticizing cryptocurrencies which are posted because posters know the site agrees with these dynamics. I'd like to avoid those circular dynamics on this site if possible; they don't lead to good or illuminating discussion.
I think a lot of the positive sentiment might be outside our filter bubble. We keep hearing about it in the news and prices keep rising. Or fluctuating, anyway.
I suspect it's not so much that our bubbles are negative, but their bubbles are positive. It seems like outside of these specific communities, the general sentiment is skeptical of cryptocurrencies, and outright hostile to NFTs.
Rightly so, in my opinion.
This discussion from August is what first comes to mind for me personally when I think of positive crypto sentiment here on Tildes. There were three different users defending the NFTs, and while they were certainly in the minority, they received a fair amount of votes and support, more than I would have expected. I wouldn't call the general feeling about crypto on Tildes to be universally negative, but it is indeed mostly negative.
I see excitement over crypto constantly in my corporate Slack server (my company has over 50,000 employees). While it's not exactly outright promotion of cryptocurrencies/NFTs, a lot of my coworkers around the globe seem to be very excited over blockchain technology and crypto in that sense. Just today I found someone asking about people to follow on social media for information about web3, DAOs, and Ethereum Name Service.
I'm going to take a slightly controversial position here for Tildes and say that even if this is true, I still think it's important to actively take a stance against crypto even if it produces subpar discussions. I think the existence and popularity of crypto is a material threat to the climate, and I think it's really important to do whatever we can to protect the climate, even if that is just making sure no one touches crypto on Tildes. I live on the planet, and I don't like what crypto is doing to accelerate its destruction. Because of that, I think it's more important that we establish a community consensus against crypto than to ensure quality discussions around it. Of course, I would like both, and I think we have seen good discussions around crypto generally even if most people are on the same side of "crypto is bad and a scam".
Yes I disagree with this position. Creating a negative consensus around an issue because members dislike the issue is pretty much the textbook definition of an echo chamber; it's the same thing that happens on sites with lower quality discussion. I think the ability to have nuanced discussion even about topics you disagree with is an important part of creating a high quality discussion space.
It's not just a disagreement though. Cryptocurrencies are accelerating the destruction of the planet. I live on the planet, so I don't want that to happen. It's like a lesser version of the situation of hate speech. I don't just "disagree" with hate speech; it has a materially negative effect on the world. I think preventing that negative effect on the world is more important than preserving high quality discussion. This isn't to say that we shouldn't have both, just that one, to me, is far more important than the other. It's more than just disliking the issue or disagreeing with crypto and NFTs; it's a real and present danger to everyone.
On the other hand, discussing NFT’s isn’t the same as making or buying them, and the NFT hype (which seems to be dying down) has only an indirect effect on the price of Ethereum, and the price is what actually controls the blockchain’s burn rate. So we are two steps removed.
Also, Tildes is a tiny site, with tiny influence compared to the rest of the Internet. We don’t have any real influence over whether NFT’s are popular or Ethereum is popular, let alone climate change. They are enormously bigger than us.
That means we don’t need to treat Tildes moderation as being of terribly high-stakes, planet-shaking importance. The main thing at risk is our own enjoyment of this website.
Yeah, you're right on that. I think we should absolutely encourage better discussion around NFTs. I definitely think we can do better in provoking good, nuanced discussion around crypto while also ideologically opposing it.
Personally speaking, I wish that I had a set of peril-sensitive glasses set to sensor out all things crypto, blockchain, or any other speculative market.
IMHO these are all marketplaces full of moderately rich unscrupulous people taking money away from moderately rich people who simply don't know any better than to buy stupid things for the sake of status symbolism. But the sad thing about cryptocurrency in particular is that there are some decidedly not-rich people who will end up even worse after the bubble eventually pops.