NFTs (and crypto currency in general in a lot of ways) set out to solve problems that we don't experience in the real world because we don't live in libertarian fantasy. One of the strongest cases...
NFTs (and crypto currency in general in a lot of ways) set out to solve problems that we don't experience in the real world because we don't live in libertarian fantasy. One of the strongest cases for NFTs and the blockchain are for contracts and, well, the problem is that we already solved contracts like 2000 years ago and we did it without polluting tons of carbon too.
I don’t love everything crypto but we are FAR from having solved contracts. The solution being “the truth is whenever has more lawyers and more money” in so many cases is far from ideal. I can see...
I don’t love everything crypto but we are FAR from having solved contracts. The solution being “the truth is whenever has more lawyers and more money” in so many cases is far from ideal.
I can see the appeal of “we both agreed to this simple digital contact with BINARY style logic of “if A then B” “ which could then auto execute on completion of A.
The problem THEN is you move from high priced lawyers being the problem to best data practices and it’s about a wash.
I do think there are some serious advantages to auto executing style logic, even in more complex form, but I’m also not sure if needs to be decentralized
Just as a point of clarification, NFTs offer no functionality with regard to auto executing code, from what I understand. It’s more that a distributed system has concurred that the contract exists?
Just as a point of clarification, NFTs offer no functionality with regard to auto executing code, from what I understand. It’s more that a distributed system has concurred that the contract exists?
I guess I was unclear but yeah that's a part of automating a contract. It both has to execute, and has to exist as a record. The recorded execution is I guess, in theory, where an NFT could help,...
I guess I was unclear but yeah that's a part of automating a contract. It both has to execute, and has to exist as a record. The recorded execution is I guess, in theory, where an NFT could help, but even then I feel like it's another layer on the already awkward crypto foundation. A public ledger basically already solves this.
I think that they're compelling ideas but they have very limited real-world use. Being able to say that you own something and it's on a public ledger to prove it sounds really useful. It's tying...
I think that they're compelling ideas but they have very limited real-world use. Being able to say that you own something and it's on a public ledger to prove it sounds really useful. It's tying that into real-world property that is tenuous. And without any strong links between the token and the thing, functionally, it's a net negative IMHO.
Imagine an artist and a buyer. The buyer commissions the art and pays for it and the legal rights of ownership. But this kind of thing is done by contract - you have to have some sort of real-world proof that you own an IP, right? Usually in the form of a contract or a bill of sale. But where does an NFT come into play? As far as I'm aware, few if any legal bodies will accept an NFT as proof of ownership, so you need your document instead. So the NFT is only really selling you the idea of ownership but not the actual ownership. But real money - or cryptocurrency - is still spent on creating the NFT, so the buyer ends up paying extra money for something that doesn't have any legal value, and any financial value it has is subject to extremely volatile market whims.
I think the bigger problem is that the public ledger has no guaranteed right of existence. Your deed and home sale are recorded in the recorders office, and if it’s no longer around you have...
I think the bigger problem is that the public ledger has no guaranteed right of existence.
Your deed and home sale are recorded in the recorders office, and if it’s no longer around you have bigger problems.
The weird nature of crypto means that anything could vanish and it begs the question “why is this better than a normal database with proper records and audits?”
The only thing that is better about it is, that it is cryptographically ensured that a record cannot be forged after the fact. So you can basically replace or supplement trust in the organization...
The weird nature of crypto means that anything could vanish and it begs the question “why is this better than a normal database with proper records and audits?”
The only thing that is better about it is, that it is cryptographically ensured that a record cannot be forged after the fact. So you can basically replace or supplement trust in the organization that stores and manages the contracts with trust in the cryptographic mechanisms underlying the NFTs. While an insider a the recorders office could easily manipulate a particular record, they couldn't selectively manipulate a particular record in a blockchain.
While this sounds like a nice idea conceptually, the way it was used is absolutely insane:
Firstly simply storing a URL in the blockchain defeats the entire purpose, as you can simply manipulate the resource it points to without making changes to the blockchain itself. Instead of cryptographically integrating a unique, unforgable identifier of the actual thing to the blockchain you only get an unforgable address that points to where the actual thing is supposed to be found. Directly integrating bigger chunks of data into the blockchain (e.g. entire images) is infeasible, but whoever invented the concept of (mutable) NFTs that simply use URLs instead of cryptographic hashes as a replacement for the actual data fundamentally misunderstood the one advantage NFTs have over trusted organizations keeping the record instead.
Secondly people started to expect that being the first to mint an NFT gave them some kind of unique right to ownership to that thing. That's a bit like setting up a fancy contract claiming ownership of your neighbours' house, signing it yourself, and being able to prove that you did so at a particular time. If you go to your neighbours and to kick them out based on that thing, they'll just call the police and no court will accept your claim. In other words NFTs are a tool to proovably claim something that's utterly worthless if you don't have any real-world mechanisms/laws agreed upon by society to back up your claim's legitimacy.
So if we disregard the whole energy efficiency problem of blockchains for the sake of the argument, the only scenario where it could make sense to use NFTs would be to supplement a preexisting real-world process to store legitimate contracts in order to make their existence provable. Imagine the recorders office stores a digital copy of each signed contract (e.g. as a signed pdf file). They could then compute a cryptographic hash of any file (i.e. a short identifier that given a file can be easily computed by everyone, but not the other way around) and store it in the blockchain (i.e. mint an NFT for it). The blockchain they used could then be redundantly stored in a publicly accessible way. This way anyone in posession of the PDF containing the contract and access to a copy of the blockchain could verify that the recorder's office claimed it was authentic at the point in time the NFT was minted, and it would be computationally infeasible for anyone (even the recorders office) to manipulate that record of authenticity after the fact. I'm not saying it's necessarily a good idea to implement this mechanism, but at least its an example of using NFTs in a useful way that's not inherently batshit insane.
Could you elaborate why you say that? Simply put a blockchain is a chain of blocks that are linked with cryptographic hashes. If you change one block all subsequent hashes change and have to be...
I don't think that's true though, because you can do that in a database?
Could you elaborate why you say that?
Simply put a blockchain is a chain of blocks that are linked with cryptographic hashes. If you change one block all subsequent hashes change and have to be recomputed, and due to the cryptographic properties of the hash function used you can't take any shortcuts here. So to change one earlier block you have to invest at least as much computational power as everyone else has invested since the creation of that block to forge the chain. So the idea is that as long as you have a large number of actors contributing computations of hashes with a vested interest in an unforgable chain no onewillbe able to invest more computational power than everyone else combined.
While the specifics are a bit different, this underlying principle of NFTs and cryptocurrencies is the same. So forging an NFT record is equally hard as forging a bitcoin transaction.
What do you mean by 'you can do that in a database'?
To start with, lets use your example: No, they couldn't. Even in very basic database design they'd have to make a logged entry tombstoning the previous record and entering the new record to...
To start with, lets use your example:
. While an insider a the recorders office could easily manipulate a particular record, they couldn't selectively manipulate a particular record in a blockchain.
No, they couldn't. Even in very basic database design they'd have to make a logged entry tombstoning the previous record and entering the new record to replace it. Actual deletes and edits are rarely allowed in anything that requires any level of transparency or history.
You would have, at minimum, the actual records of the table, seeing that record 1 was tombstoned on date X by user A, and record 2 was made live on date X by user A. You would almost certainly also have the actual DB logs that track all this on the backend anyways, because again if you're building a database for the express purpose of this kind of thing, you'd absolutely use that functionality.
With just these techniques plus some basic rights management and best practices you have functionality good enough for the vast majority of circumstances.
From there you can salt/hash/encode/checksum records and tables themselves in a whole variety of ways, involving outside sources or trusted entities for audits on these, or literally giving them parts of the private/public keys or whatever to ensure that mathematically it was impossible for those older records to be modified. Given that one of the core tenants of most database design is a unique identifier for every record, that + the contents of the record + one of many encryption options will give you a unique and verifiable number. You force that data to be replicated to say, 5 different entities, and then if there's ever doubt you have all 5 produce what they expect the value of that record to be. If it's still that value, congrats, it's not changed.
This is all still "naive" stuff for the industry, but perhaps more importantly you can run blockchain style encryption locally to the db itself. Nothing about crypto is so unique you couldn't just replicate it on the server itself, and both Azure and AWS offer this service (ledger and quantum tables/databases respectively I believe). You can very easily spin up a ledger database in MS SQL and then it is fundamentally append only.
And all of this is without getting into the weirder details of things like Elliptic Curves and verifying that your "transactions" are on the same curve which can only be done if you have part of the valid equation.
The "unique" part of crypto is the idea that your verification is 100% outsourced and decentralized (in theory) because you have strangers from all over the world agreeing on the actual computed value and processing the transactions. Obviously the larger the % of ownership of the computation, the more likely the value could have still been tampered with, but you don't need millions of strangers to get more than enough proof that something hasn't been modified.
If you decide "well we don't need millions of strangers" and just run your own hardware and say have 2-20 outside auditors following best practices and standards, you are more than fine with your data, and arguably better because you're not stuck overkilling the problem, dealing with a complicated ecosystem, worrying about ownership attacks, and all the other overhead and outside issues crypto brings.
The reality is all the bullshit lies aside, the real reason for NFTs was to help cash out Crypto. Your $10 billion in crypto is worthless if you cannot buy anything with it. So you convince a...
The reality is all the bullshit lies aside, the real reason for NFTs was to help cash out Crypto. Your $10 billion in crypto is worthless if you cannot buy anything with it. So you convince a bunch of suckers to buy NFTs, which brings in more real money, which helps you cash out your crypto for real money.
Most practical hypothetical use a friend suggested (though maybe for block chain in general?) was to track ownership of digital games. That way you could sell or trade it. I guess the other one...
Most practical hypothetical use a friend suggested (though maybe for block chain in general?) was to track ownership of digital games. That way you could sell or trade it.
I guess the other one would be what I think was the original intent, which is to track who originally created an image. Basically you'd make the image and register it, and could then use that to challenge art theft and demand its removal from various sites.
It just kinda feels like extra steps, if you have all of those things you'd still need to appeal to a central authority right? I feel like both of those things are kinda block chain in general...
It just kinda feels like extra steps, if you have all of those things you'd still need to appeal to a central authority right? I feel like both of those things are kinda block chain in general anyways, but even then it feels a bit over complicated.
Yeah, they're not the best options, just the potentially practical uses for it. At bare minimum, I think having a token to prove ownership of an image could help issue takedowns to unauthorized...
Yeah, they're not the best options, just the potentially practical uses for it. At bare minimum, I think having a token to prove ownership of an image could help issue takedowns to unauthorized usages, or just add some extra weight if it reaches court. I feel like there's got to be a better system though which doesn't involve such environmentally destructive processes though...
The issue with this problem, and all other problems of this class (i.e. digital payments) is that it can already solved with a conventional database using a minuscule fraction fraction of the...
Most practical hypothetical use a friend suggested (though maybe for block chain in general?) was to track ownership of digital games. That way you could sell or trade it.
The issue with this problem, and all other problems of this class (i.e. digital payments) is that it can already solved with a conventional database using a minuscule fraction fraction of the compute resources. The reason you can't trade or resell Steam games is because Valve, publishers, and developers don't want that. Putting Steam libraries on a blockchain wouldn't change this.
Definitely. Which I think just highlights how pointless NFTs are overall: the hypothetical practical uses I've heard so far can be implemented via other, more efficient methods. The technological...
Definitely. Which I think just highlights how pointless NFTs are overall: the hypothetical practical uses I've heard so far can be implemented via other, more efficient methods. The technological capability isn't the primary roadblock to those methods.
I feel like this is technically inaccurate since the token itself is what matters, precisely because the image is, by nature of being a digital artifact, infinitely replicable. If the owner of the...
I feel like this is technically inaccurate since the token itself is what matters, precisely because the image is, by nature of being a digital artifact, infinitely replicable. If the owner of the image didn't backup the image they own in some sort of lossless format however, that's kinda on them.
Well, the token only contains a URL to the image as a payload, which I always thought was dumb. I always thought the token should contain a hash of the NFT as well, that way one could prove...
Well, the token only contains a URL to the image as a payload, which I always thought was dumb. I always thought the token should contain a hash of the NFT as well, that way one could prove ownership even in the absence of a web host being online.
Can it still have value as proof of ownership of an image, when it no longer displays which image it claims ownership of? Surely this is like if a property deed didn't include the address of the...
Can it still have value as proof of ownership of an image, when it no longer displays which image it claims ownership of? Surely this is like if a property deed didn't include the address of the property. And resell value must be out of the question now too.
It's more like the address on the deed stopped being a real place. Which brings me to my sudden intense curiosity of the day: how would land ownership interact with events like "I dug a hole to...
It's more like the address on the deed stopped being a real place.
Which brings me to my sudden intense curiosity of the day: how would land ownership interact with events like "I dug a hole to the center of the earth and this area kind of doesn't exist anymore", or "a gravitational anomaly appeared and now spacetime is curved a bit differently along the formerly straight property line"?
Those questions are probably already being tested in coastal areas that are losing ground to rising sea levels. Or if they aren't yet they will be soon.
Those questions are probably already being tested in coastal areas that are losing ground to rising sea levels. Or if they aren't yet they will be soon.
Maybe I'm just dumb but, has there ever been a legit use case for NFT's yet? I haven't seen anything compelling for any NFT so far.
I mean, they're doing exactly what they were designed to do.
The question of if it was a good idea to begin with, though, is open to interpretation.
Cryptocurrency at least kinda made sense to me, maybe? But NFT's always just seemed really dumb even as an idea.
NFTs (and crypto currency in general in a lot of ways) set out to solve problems that we don't experience in the real world because we don't live in libertarian fantasy. One of the strongest cases for NFTs and the blockchain are for contracts and, well, the problem is that we already solved contracts like 2000 years ago and we did it without polluting tons of carbon too.
I don’t love everything crypto but we are FAR from having solved contracts. The solution being “the truth is whenever has more lawyers and more money” in so many cases is far from ideal.
I can see the appeal of “we both agreed to this simple digital contact with BINARY style logic of “if A then B” “ which could then auto execute on completion of A.
The problem THEN is you move from high priced lawyers being the problem to best data practices and it’s about a wash.
I do think there are some serious advantages to auto executing style logic, even in more complex form, but I’m also not sure if needs to be decentralized
Just as a point of clarification, NFTs offer no functionality with regard to auto executing code, from what I understand. It’s more that a distributed system has concurred that the contract exists?
I guess I was unclear but yeah that's a part of automating a contract. It both has to execute, and has to exist as a record. The recorded execution is I guess, in theory, where an NFT could help, but even then I feel like it's another layer on the already awkward crypto foundation. A public ledger basically already solves this.
I think that they're compelling ideas but they have very limited real-world use. Being able to say that you own something and it's on a public ledger to prove it sounds really useful. It's tying that into real-world property that is tenuous. And without any strong links between the token and the thing, functionally, it's a net negative IMHO.
Imagine an artist and a buyer. The buyer commissions the art and pays for it and the legal rights of ownership. But this kind of thing is done by contract - you have to have some sort of real-world proof that you own an IP, right? Usually in the form of a contract or a bill of sale. But where does an NFT come into play? As far as I'm aware, few if any legal bodies will accept an NFT as proof of ownership, so you need your document instead. So the NFT is only really selling you the idea of ownership but not the actual ownership. But real money - or cryptocurrency - is still spent on creating the NFT, so the buyer ends up paying extra money for something that doesn't have any legal value, and any financial value it has is subject to extremely volatile market whims.
I think the bigger problem is that the public ledger has no guaranteed right of existence.
Your deed and home sale are recorded in the recorders office, and if it’s no longer around you have bigger problems.
The weird nature of crypto means that anything could vanish and it begs the question “why is this better than a normal database with proper records and audits?”
The only thing that is better about it is, that it is cryptographically ensured that a record cannot be forged after the fact. So you can basically replace or supplement trust in the organization that stores and manages the contracts with trust in the cryptographic mechanisms underlying the NFTs. While an insider a the recorders office could easily manipulate a particular record, they couldn't selectively manipulate a particular record in a blockchain.
While this sounds like a nice idea conceptually, the way it was used is absolutely insane:
Firstly simply storing a URL in the blockchain defeats the entire purpose, as you can simply manipulate the resource it points to without making changes to the blockchain itself. Instead of cryptographically integrating a unique, unforgable identifier of the actual thing to the blockchain you only get an unforgable address that points to where the actual thing is supposed to be found. Directly integrating bigger chunks of data into the blockchain (e.g. entire images) is infeasible, but whoever invented the concept of (mutable) NFTs that simply use URLs instead of cryptographic hashes as a replacement for the actual data fundamentally misunderstood the one advantage NFTs have over trusted organizations keeping the record instead.
Secondly people started to expect that being the first to mint an NFT gave them some kind of unique right to ownership to that thing. That's a bit like setting up a fancy contract claiming ownership of your neighbours' house, signing it yourself, and being able to prove that you did so at a particular time. If you go to your neighbours and to kick them out based on that thing, they'll just call the police and no court will accept your claim. In other words NFTs are a tool to proovably claim something that's utterly worthless if you don't have any real-world mechanisms/laws agreed upon by society to back up your claim's legitimacy.
So if we disregard the whole energy efficiency problem of blockchains for the sake of the argument, the only scenario where it could make sense to use NFTs would be to supplement a preexisting real-world process to store legitimate contracts in order to make their existence provable. Imagine the recorders office stores a digital copy of each signed contract (e.g. as a signed pdf file). They could then compute a cryptographic hash of any file (i.e. a short identifier that given a file can be easily computed by everyone, but not the other way around) and store it in the blockchain (i.e. mint an NFT for it). The blockchain they used could then be redundantly stored in a publicly accessible way. This way anyone in posession of the PDF containing the contract and access to a copy of the blockchain could verify that the recorder's office claimed it was authentic at the point in time the NFT was minted, and it would be computationally infeasible for anyone (even the recorders office) to manipulate that record of authenticity after the fact. I'm not saying it's necessarily a good idea to implement this mechanism, but at least its an example of using NFTs in a useful way that's not inherently batshit insane.
I don't think that's true though, because you can do that in a database?
Could you elaborate why you say that?
Simply put a blockchain is a chain of blocks that are linked with cryptographic hashes. If you change one block all subsequent hashes change and have to be recomputed, and due to the cryptographic properties of the hash function used you can't take any shortcuts here. So to change one earlier block you have to invest at least as much computational power as everyone else has invested since the creation of that block to forge the chain. So the idea is that as long as you have a large number of actors contributing computations of hashes with a vested interest in an unforgable chain no onewillbe able to invest more computational power than everyone else combined.
While the specifics are a bit different, this underlying principle of NFTs and cryptocurrencies is the same. So forging an NFT record is equally hard as forging a bitcoin transaction.
What do you mean by 'you can do that in a database'?
To start with, lets use your example:
No, they couldn't. Even in very basic database design they'd have to make a logged entry tombstoning the previous record and entering the new record to replace it. Actual deletes and edits are rarely allowed in anything that requires any level of transparency or history.
You would have, at minimum, the actual records of the table, seeing that record 1 was tombstoned on date X by user A, and record 2 was made live on date X by user A. You would almost certainly also have the actual DB logs that track all this on the backend anyways, because again if you're building a database for the express purpose of this kind of thing, you'd absolutely use that functionality.
With just these techniques plus some basic rights management and best practices you have functionality good enough for the vast majority of circumstances.
From there you can salt/hash/encode/checksum records and tables themselves in a whole variety of ways, involving outside sources or trusted entities for audits on these, or literally giving them parts of the private/public keys or whatever to ensure that mathematically it was impossible for those older records to be modified. Given that one of the core tenants of most database design is a unique identifier for every record, that + the contents of the record + one of many encryption options will give you a unique and verifiable number. You force that data to be replicated to say, 5 different entities, and then if there's ever doubt you have all 5 produce what they expect the value of that record to be. If it's still that value, congrats, it's not changed.
This is all still "naive" stuff for the industry, but perhaps more importantly you can run blockchain style encryption locally to the db itself. Nothing about crypto is so unique you couldn't just replicate it on the server itself, and both Azure and AWS offer this service (ledger and quantum tables/databases respectively I believe). You can very easily spin up a ledger database in MS SQL and then it is fundamentally append only.
And all of this is without getting into the weirder details of things like Elliptic Curves and verifying that your "transactions" are on the same curve which can only be done if you have part of the valid equation.
The "unique" part of crypto is the idea that your verification is 100% outsourced and decentralized (in theory) because you have strangers from all over the world agreeing on the actual computed value and processing the transactions. Obviously the larger the % of ownership of the computation, the more likely the value could have still been tampered with, but you don't need millions of strangers to get more than enough proof that something hasn't been modified.
If you decide "well we don't need millions of strangers" and just run your own hardware and say have 2-20 outside auditors following best practices and standards, you are more than fine with your data, and arguably better because you're not stuck overkilling the problem, dealing with a complicated ecosystem, worrying about ownership attacks, and all the other overhead and outside issues crypto brings.
The reality is all the bullshit lies aside, the real reason for NFTs was to help cash out Crypto. Your $10 billion in crypto is worthless if you cannot buy anything with it. So you convince a bunch of suckers to buy NFTs, which brings in more real money, which helps you cash out your crypto for real money.
The same thing is happening now with meme coins.
Most practical hypothetical use a friend suggested (though maybe for block chain in general?) was to track ownership of digital games. That way you could sell or trade it.
I guess the other one would be what I think was the original intent, which is to track who originally created an image. Basically you'd make the image and register it, and could then use that to challenge art theft and demand its removal from various sites.
It just kinda feels like extra steps, if you have all of those things you'd still need to appeal to a central authority right? I feel like both of those things are kinda block chain in general anyways, but even then it feels a bit over complicated.
Yeah, they're not the best options, just the potentially practical uses for it. At bare minimum, I think having a token to prove ownership of an image could help issue takedowns to unauthorized usages, or just add some extra weight if it reaches court. I feel like there's got to be a better system though which doesn't involve such environmentally destructive processes though...
The issue with this problem, and all other problems of this class (i.e. digital payments) is that it can already solved with a conventional database using a minuscule fraction fraction of the compute resources. The reason you can't trade or resell Steam games is because Valve, publishers, and developers don't want that. Putting Steam libraries on a blockchain wouldn't change this.
Definitely. Which I think just highlights how pointless NFTs are overall: the hypothetical practical uses I've heard so far can be implemented via other, more efficient methods. The technological capability isn't the primary roadblock to those methods.
Like cryptocurrencies, they are still a solution desperately searching for a real problem to solve.
This exactly. I just feel that anytime people ask about NFTs, the ghost of Eduard Khil appears, singing his "trololol" song.
I feel like this is technically inaccurate since the token itself is what matters, precisely because the image is, by nature of being a digital artifact, infinitely replicable. If the owner of the image didn't backup the image they own in some sort of lossless format however, that's kinda on them.
Well, the token only contains a URL to the image as a payload, which I always thought was dumb. I always thought the token should contain a hash of the NFT as well, that way one could prove ownership even in the absence of a web host being online.
Oh well, dumb implementation for a dumb concept.
Can it still have value as proof of ownership of an image, when it no longer displays which image it claims ownership of? Surely this is like if a property deed didn't include the address of the property. And resell value must be out of the question now too.
It's more like the address on the deed stopped being a real place.
Which brings me to my sudden intense curiosity of the day: how would land ownership interact with events like "I dug a hole to the center of the earth and this area kind of doesn't exist anymore", or "a gravitational anomaly appeared and now spacetime is curved a bit differently along the formerly straight property line"?
Those questions are probably already being tested in coastal areas that are losing ground to rising sea levels. Or if they aren't yet they will be soon.
Alternate: https://archive.is/CL2QV
Reddit discussion: here or on old reddit