How I became Leonardo da Vinci on the Blockchain


Yesterday at the CogX conference, I sat in a room listening to companies pitch their blockchain based startups. Because I hate myself.

One in particular caught my attention. On the surface it seems to solve an important economic problem - art forgery and provenance.

By putting your artwork on the "BitCoin Blockchain", Verisart will ✨hand wavy magic✨ increase the trust in art dealers and reduce fraud.

That's a pretty neat idea. A distributed public ledger of who I have sold my art to. And, if they sell it, they have a cryptographically signed certificate proving its provenance. The seller can sign the certificate with their private key to say they've sold it to the new owner of a specific public key. Nifty!

Except… and I hate to bring the art industry into disrepute… what if I sell a fake and keep the original in my Underground Vault?

There's no way to permanently attach a digital certificate to a physical work of art.

Incidentally, this is the problem with all the startups claiming the blockchain will revolutionise the integrity of global logistics markets. Sure, you can slap a QR code on a crate - but nothing stops an unscrupulous middle-man from replacing or adulterating the contents of the crate.

Let's get back to Verisart's other issue. Proving that I am the creator or owner of a piece of artwork.

Long story short, I convinced them that I painted the Mona Lisa. An excellent situationalist prank. Much avant-garde, so postmodernism.

For this "proof" I provided Verisart with...

  • An email address
  • A photo of the Mona Lisa from Wikipedia
  • Er…
  • That's it!

That was all that was required. I didn't have to provide them with a public key, a certified copy of my passport, a high resolution copy of the artwork. Nothing.

Because the blockchain is immutable, that "proof" is there forever now. You mustn't make mistakes when irrevocably signing something.

So, what does this mean? Can I waltz into the Louvre and ask for "my" painting back?

No, obviously not. Not even a moron in a hurry would think that I was a time traveller or that I owned the most famous portrait in Europe.

But, if it is certified on the blockchain, it must be true!

I'm sure there are many sensible uses for this technology. I'm also sure there are many investors who will throw money at any pitch which contains the right buzzwords.

It's almost as if we can't just throw technology at the social problem of trust…


23 thoughts on “How I became Leonardo da Vinci on the Blockchain

  1. I often think the same when I hear people pitching blockchain based identity solutions (especially for the developing world). They correctly identify a problem: lack of a provable identity presents hurdles for poorer people interacting with the economy. But they fail to understand the root causes or the specifics of the problems. The software industry has a habit of viewing problems in other fields/professions as "simple" if you throw technology at them; it's almost a collective Dunning-Kruger effect.

    I would slightly quibble with your point that "Incidentally, this is the problem with all the startups claiming the blockchain will revolutionise the integrity of global logistics markets". Blockchain may allow automation of some of the paper processes for bills of lading etc and the nature of the logistics industry with lots of counterparties involved in one transaction might make a distributed system attractive. It wont make it much more fraud resistant than the current paper processes, but it could make it faster and more transparent which could be of value.

  2. Your complaints about this system are legitimate, but I think you’re mis-attributing them to blockchain. When you cross from the physical to the digital world, blockchain still works, but only if you keep humans involved to validate the data entry. In the case of the art, that would require physical inspection and possibly tagging the piece. I’m imagining an RFID tag embedded in the frame or similar. Once that piece is done, then this becomes a useful piece of tech for managing provenance because of the transparency it allows.

    The general supply chain problem is similarly improved. You still need tamper-evident seals on crates. You still need people verifying at the end (at least) that what is stated in the blockchain manifest is actually there, but you can do it with less human interaction and less verification and more transparency.

    Thats where this, and the general supply chain problem, get helped by blockchain. The current paper systems used in global commerce allow for many opportunities for fraud because of the many many manual, paper-based steps they involve. Those get driven out as distributed systems shine lights on bad actors. Once a problem is discovered and someone asks “Who caused this?”, the immediate answer isn’t “I don’t know, it will take me a few weeks to go through the paperwork.”. The answer is “We got this thing from Joe at XYZ place, let’s go talk to him.”. Therein lies the value.

    Blockchain can’t solve for garbage-in/garbage-out, we still need other controls for that. That’s not the problem it’s trying to solve though. Look at the work Maersk is doing for example of how applying blockchain well can be helpful. Ibrahim Gokcen talks about it here: https://events.technologyreview.com/video/?event=business-of-blockchain&year=2018

    Lots of good talks there, but that and Frank Yiannas’ talks are the most relevant.

    1. What stops my physically moving the RFID to another frame?
      If you have tamper evident seals - why do you need a ledger?
      Could the problems be solved with a normal database based on PKI?
      If pirates can take over 50% of your chain, can they hide their smuggling?

      Blockchain is amazing technology - but it is a solution in search of a problem.

      1. Yes, and the people checking the painting could fraudulent.

        Using a blockchain does not prevent:

        Events that are happening that are not registered
        Events that didn’t happen are registered
        Wrong events are registered

        So far the only ‘reasonable’ use case seems te be cryptocurrency, as long as nobody controls more than 50% of the chain. Then still I don’t get why entries in a secure distributed database should be worth anything. Anyone can start a coin or a fork of bitcoin, so why do they have value at all?

        1. Anyone can also print their own paper money, doesnt mean you will get a large network of people who accept your paper money.

      2. "What stops my physically moving the RFID to another frame?"
        Nothing. That's only there to make data entry easier and less error-prone. Maybe you could build one that it tamper evident. This portion is really an aside, a value-add.

        "If you have tamper evident seals - why do you need a ledger?"
        So that you have visibility into the paper trail easily and quickly, anywhere you have Internet access. A sealed container doesn't on it's own tell me where it came from or what's in it, or what stops it made along the way.

        "Could the problems be solved with a normal database based on PKI?"
        Sure. But those systems have failed to generate buy in. Blockchain has created that. Is it because people don't really understand it? Maybe. Is it because it's genuinely unique and better suited? Maybe. In the end though, neither of those details really matter. We've been trying to go paperless and distributed for decades, and none of the other advancements we've made have made significant progress on those end-to-end. Blockchain looks like it will.

        "If pirates can take over 50% of your chain, can they hide their smuggling?"
        If you use a permissioned chain, rather than a public chain, not likely. At least, no more likely than any other Internet-connected system getting taken over by bad actors.

        "Blockchain is amazing technology - but it is a solution in search of a problem."
        I used to think that too. Seriously, go watch some of the talks I posted above. It's likely your opinion will shift a bit.

    2. Your complaints about this system are legitimate, but I think you’re mis-attributing them to blockchain. When you cross from the physical to the digital world, blockchain still works, but only if you keep humans involved to validate the data entry. In the case of the art, that would require physical inspection and possibly tagging the piece. I’m imagining an RFID tag embedded in the frame or similar. Once that piece is done, then this becomes a useful piece of tech for managing provenance because of the transparency it allows.

      How does a blockchain make this better than a “simple” 3rd party that you pay to transfer verified ownership of a registered object to another account in their database? They also offer a nice API so you can integrate. Seems much less complicated. For example I can still use my password and if I forgot it I can recover it (even if that may involve some kind of strict check). That is not possible if I somehow screw up my key (as far as I understand blockchains in general).
      If (as it appears) the whole blockchain part is behind a web interface, then why blockchain? Wouldn’t a classical (PG, MSSQL, even Mongo) make things easier and faster while retaining the exact same outward functionality?

      1. The part that makes blockchain applicable and interesting in this space (and where this implementation seems to miss the mark) is the distributed pseudo-immutability part. That has been an important level in getting large organizations to be willing to participate in consortia and open their records and/or rely on records held by third parties. That's what drives transparency into the system, and wherein the value lies. This particular implementation fails there, and so it's not better than using a traditional RDMS or other traditional storage system. Saying that's a failing of blockchain though is incorrect. A car built with square wheels isn't likely to work right, but that's not the engine's fault.

  3. Blockchain is all BS, only innovation in this space is proof-of-work/mining to secure a network for payment and the incentive(token) and the game theory behind it. Everyone saying blockchain is interesting, bitcoin not sure - are fooling themselves IMO. All the banks and private sector hyping up private blockchain is utter non-sense and same with companies claiming to do authenticity or putting healthcare data on blockchain are either frauds or delusional. Their blockchain if centralized is meaningless, they can do this with databases/existing technology.

  4. Well done Leonardo!

    If you're interested in attribution problems in contemporary art, check out Richard Prince, who declared as his own art pictures made by other people, without their permission (it's a little more meta than that but not less ass-hat-y); and Peter Doig, who had to sue to prove (legally anyway) he did not paint a certain picture.

    Also, I think you might mean "situationist" prank.

  5. Art is special in the sense that there are a lot of forgeries on the physical product, and the it's the physical product that drives the ownership-document, not the other way around.

    Take ownership of a piece of land for example. It is the document that is the main driver of the ownership. And no forgery on the land itself can occur, only on the document. And forgery on the document is rare because it is centrally stored by land registry. With cars it's a bit less like that, but still the forgery will occur on the document, not the car.

    Right now, when you want to transfer ownership of land, you need to pay an expensive notary to do the transfer. And even then the transfer is not atomic. Large sums of money still take several days to clear over separate banks. If we would live in a blockchain world where ownership of land is stored in the blockchain, and you could pay with cryptocurrency, you actually don't need any notary to do the transfer. A smart contract can verify that all is in place (correct ownership and correct sum of money), and to the transfer of both atomic. With land this might seem strange to you, but with cars this would make perfect sense.

    Buying and selling a car is not risk free. Is the money forged? Will he wire-transfer the money? Will I get the ownership document after I pay? Is this the correct ownership document of this car? etc. In the future, the smart contract that does the transfer of ownership could also include the car-key, when this is stored in your smartphone or fingerprint based, instead of a physical key.

    To really see the value of blockchain application, look at what notaries and escrow services are doing, and append 'way cheaper' and 'algorithmically correct' to it. I like to comparison of snail mail to email. The latter is free, and more reliable in delivery. Sure the first one still has benefits over email, but which one are you using the most?

    Dismissing all blockchain application because one startup didn't do proper verification seems stupid to me, and is in a niche that is less benenicial for blockchains. Like I said, look at what notaries and escrow services are doing, combine that with ownership documents, cryptocurrencies and smart contracts, and see what happens.

    1. I don't think you've made a compelling argument.

      Cars are often forged - either fake plates or altered VINs.

      In the UK, large sums of money can be bank transfered almost instantly. Certainly much quicker and cheaper than a blockchain transaction.

      What happens if I'm scammed? I can go to judge and a transaction can be reversed. How does that work on a immutable ledger?

      If the car I've bought is a dud, I can ask for my money back. If the seller refuses, I can ask my bank or credit card to reverse the transaction. Again, can't do that on a DLT, can I?

      It sounds like what you really need is cheaper legal services and faster banks. Not a blockchain.

      I'm happy to say that land registry should be public - but seeing as it relies on a centralised authority (to uplhold land rights) I again fail to see how a blockchain helps.

      1. All fun and games! The digital world hasn't changed, but there are more ways to do things today, than yesterday - generally a good thing;

        To the art world scenario; Cascading Trust is necessary to make this work; It is part of the world today and won't vanish until AI gets a lot better. To register an asset on any art chain, you need to verify its authenticity prior to registration. This means reliably ascertaining the identity of the owner (identity issue mentioned above) as well as reliably ascertaining the validity of the art (asset). In the latter case, you need physical inspection by an expert whose reputation is tied to the inspection results and who is also reliably identifiable (identity again!). As time goes by, and more inspections occur, they are attached to the asset on the chain. Sooner or later, there is a sufficient amount of validation to support the validity of the asset and the ownership. Nefarious inspectors are blacklisted quickly (if the identity component works). Fake assets that are validated and later proved fake damage the reputations of all verifying inspectors.

        In terms of on chain verification, one could use high resolution photography to create an analysis of the actual tone gradients and stroke components of the art (color is too difficult, today) creating a representational map at the micrometer level which could then be hashed and stored as the fingerprint of the piece. A forger can fool the naked eye, but won't likely replicate the piece to this level of complexity (unless the replica was created using 3D printing technology). This is just meant as an example of a validation routine that could allow an end user to reasonably quickly determine if a piece was highly likely to be the same piece that is registered on the chain.

        AI (much more advanced than today) could remove the need for human authorities, and increase the identifiers possible to fingerprint the work; but as with all tech, this AI technology could also help fraudsters to up their game too.

        Blockchain transactions work best when there is verifiable trust in the IDENTITY protocols in use to identify the parties involved and the assets. This is no different than what happens today without blockchains; the improvement is that once identity is established, it doesn't need to be RE-established at every point in the supply/value chain.

        Now; to answer the "What if I get scammed?" question: You are correct, you can't UNDO a transaction. You CAN update it. If you get scammed, then there was a failure of IDENTITY on the platform; There will necessarily be parties who have guaranteed one or more identities that were party to the transaction. Those parties are now legally on the hook for their failure. This could be a KYC provider, an inspector, or even the owner of the smart contract/blockchain company. You have recourse, and if the company isn't prepared (and insured) to handle this, then you probably don't want to use their services. Properly implemented; scammers will avoid these platforms due to their transparency and lack of anonymity.

        In terms of the REFUND process; this works the same as it does today. No one returns to the seller and says "UNDO my transaction," they say "It doesn't work, I want a refund." It's a NEW transaction (just ask the sellers bank or ERP system); These scenarios are easily handled in blockchain.

        As far as speed of transactions goes; sending funds door to door within a highly advanced banking system is AS fast as most crypto platforms designed for the same purpose (BTC was never meant to do that); the fees are lower and the most telling difference is that the speed and fees in crypto is the same whether you send it across the street or around the world. Try sending funds from South Africa to Mexico and see how long it takes, and how much it costs.

        1. In terms of the REFUND process; this works the same as it does today. No one returns to the seller and says "UNDO my transaction," they say "It doesn't work, I want a refund." It's a NEW transaction (just ask the sellers bank or ERP system); These scenarios are easily handled in blockchain

          That's not true. I ask for a refund, the seller says no. I compain to my credit-card provider, they reverse the transaction and give me back my money. They then either get it back from the seller or commence legal proceedings.

          Again, the average customer can't do that with Cryptocurrency. All sales are final. No refunds. No exceptions.

          Try sending funds from South Africa to Mexico and see how long it takes, and how much it costs.

          https://transferwise.com/pricing/ Quicker and cheaper than Bitcoin. Much easier to use as well.

          1. Nice credit card company! Where can I get an account like that?!! Mine demands proof of fraud and internal review before they'll try to reverse a valid purchase.

            Actually, Transferwise (my favorite service!!) is not quicker than BTC for international and Bitcoin is SLOW (I said so above - it was never meant for speed - Confirmation takes 60 minutes!). ETH is slow too, but it still transfers value in 1 - 5 minutes, and then there are the protocols like Ripple and NEO which are much faster.

            1. Literally every credit card in the UK has Section 75 protection.

              But, let's go back to blockchain. How do I demand a refund on a fraudulent cryptocoin transaction?

              1. I the current system, that's hard. It's one of the problems that layer-two technologies are working addressing. Smart contracts look like a promising mechanism to address this. In this case, it is setup to look for confirmation entries from both parties further down the chain before the transaction actually happens. It ends up becoming a simultaneous transfer and you close the window on the timing attacks that transactions currently have.

                Look harder at the space beyond currencies. I personally think they're mostly a distraction. There's a lot going on in other areas that I think will ultimately offer more. To use a car analogy again, we're in Model-T days still, but that's going to change quickly.

      2. In the UK, large sums of money can be bank transfered almost instantly. Certainly much quicker and cheaper than a blockchain transaction.

        The Nano cryptocurrency has 0 fees, transact in seconds, does 7000 tps on the power of a single windmill. Please explain which bank can compete with that.

        What happens if I'm scammed? I can go to judge and a transaction can be reversed. How does that work on a immutable ledger?

        Do you know what happens when you get scammed by doing a wire transfer? Here in Europe absolutely nothing in reality. Yeah that's right. That your transaction will be reversed by the bank is just your (and a lot of other people's) imagination. It's called invoice fraud, look it up and be amazed how people are not getting their money back.

        If the car I've bought is a dud, I can ask for my money back. If the seller refuses, I can ask my bank or credit card to reverse the transaction. Again, can't do that on a DLT, can I?

        Scammers on car deals do person-to-person deals, no way you will see your money back. You don't even have the identity of the other person. You are going to pay $15000 with your VISA card to a private person? How exactly?

        It sounds like what you really need is cheaper legal services and faster banks. Not a blockchain.

        Or a DAG 🙂

        I'm happy to say that land registry should be public - but seeing as it relies on a centralised authority (to uplhold land rights) I again fail to see how a blockchain helps.

        I can communicate just find with people over snail mail. I can even perfume the letter to my girlfriend, so why would I use email? Everyone has an address, not everyone has an email address. I don't see email going anywhere soon, snail mail offers everything we need and more. A letter takes a few days to arrive, but that's no problem. Because if you need urgent communication, you pick up the phone. That one is still faster than email. Like I said, email is a hype and it's going nowhere.

  6. Poor execution of a good idea doesn’t rubbish the idea. Blockchain and AI/ML face the same garbage-in, garbage-out problem. Human intervention is needed to verify/clean data/ input before in goes through a ‘chain’ .
    Most compelling use case for blockchain as a game changer beyond crypto currency is this piece on tokenization of physical assets indeed such as art, real estate etc.
    https://hackernoon.com/the-security-token-thesis-4c5904761063

    When stuff is tokenized, asset allocation and trade becomes frictionless due to interoperability. I.e you can have all these assets in one wallet and exchange them seamlessly and speedily across geographies. Example. ERC20 creates a standard where it’s possible to exchange/ transfer value of different types of asset from one instance. I can buy factional ownership of real estate , land or art etc from one wallet or trade these with other assets without having to contact lawyers, notaries etc.. all in a matter of minutes. Still early days but this is a compelling
    Future:-)

    1. That still doesn’t offer any solution to negotiability, if someone misleads you or tricks you there is no recourse as long as the party who mislead you chooses not to sign your assets back to you aside from hard forking, and if you’re going to hard fork your silly blockchain every time someone gets screwed by piss poor tech then what is the point of having the block chain at all? The biggest benefit of a block chain is that you can proceed with 0 trust but the fact that you’re going to have to do all of the same due diligence that you would normally need to do to buy or sell valuable art there is literally 0 benefit at all to adding blockchain nonsense to the process.

      1. Governance (avoiding forks) and simultaneous transactions (avoiding some forms of fraud) are actively being worked on. Give it a minute.
        The biggest benefit of blockchain isn't zero trust, but distributed trust. The zero-trust piece is kind of a lbertarian fever dream that has been tied to blockchain that I think is wrong headed. If you tweak it though, to be distributed trust, then it's a tangible benefit.

        To use the buying and selling of valuable art example. You're right, this implementation is bad. It's not distributed, it doesn't really solve any problem. If they tweaked it though, and created a consortia, it could be good. Imagine an "Art dealers collective" where they used an OSS blockchain implementation (maybe Decred with Politea?) to store their blocks on a partially permissioned blockchain. Anyone can read it, but only verified dealers can write to it. They add the ability to add photos, videos, etc and have them validated on the chain, but are probably stored off-chain.
        Any dealer who wants to buy/sell with it participates in the chain and runs a node. If any particular dealer does shady stuff, it happens out in the open, making it more likely to be found. If a dealer goes under, their records don't disappear , but maybe their off-chain documents do. Incentivising off-chain storage of extra data is still a problem, but probably solvable. If a buyer in the US wants to validate transaction history for a particular piece, they just ask the system, rather than running down numerous dealers all over the world. If something seems fishy about a piece, you can view the previous pictures / whatever and have confidence they haven't been tampered with.

        There's lots of value being added here. They just didn't build it right, and that's not blockchain's fault.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.