5 ms·

"Blockchains are computers that can make commitments. Traditional computers are ultimately controlled by people, either directly or indirectly; blockchains inve

"Blockchains are computers that can make commitments. Traditional computers are ultimately controlled by people, either directly or indirectly; blockchains invert this power relationship, putting the code in charge." Great way to explain the concept. You could nit-pick all morning if it was strictly "true" in every which way, but this kind of selective simplification is what drives understanding. I'm sure many things which are well understood today, if you look at the way they are understood now, the mental models involved: many wouldn't have been generally be accepted in an initial pre-understanding phase by people with insight. Can you easily accept this explanation of blockchain?

bildung 6y ago
That is just meaningless marketingspeak, and obviously wrong. Blockchains are just data, so at most "spreadsheets that can make commitments". And the latter part is also wrong, as computers are controlled by silicon logic in exactly the same way as blockchains are controlled by their code: They aren't. Someone owns the computers, and someone owns the blockchains/ their data. And if a dispute happens, that will get resolved by courts. I'm well aware of the promise of this being different with blockchains, but so far that never happened.
- TheColorYellow 6y ago
  Technically the blockchain that would be referred to in the GP is more than just a data type. It would also include the algorithm for deciding the next block of valid transactions. I think the reference is supposed to imply that the consensus mechanism built into the system is responsible for dictating the validity of a transaction as opposed to a human.
- Taek 6y ago
  > Someone owns the computers, and someone owns the blockchains/ their data. That's the beautiful part. That someone is the end user, who is running a full node. The reason that blockchains require every node to have all the data is so that every end user is in full control of the commitments made by the blockchain. Efficient? No. But does it invert the typical power relationship? Yes! Blockchains put the user in control. (fine print: this is not necessarily true for many Proof of Stake chains, also not true many lite client models (including most Bitcoin lite clients), also not true if you have on-chain governance, also not true if the blockchain makes other compromising design decisions such as auto-updating, most cryptocurrencies fall short in at least half a dozen ways)
  - verdverm 6y ago
    Wrong, crypto is a scam that makes you think you have some semblance of power in a system no one uses in real life
  - bildung 6y ago
    If that were the case, why can a16z profit from that?
- febeling 6y ago
  Who do you think the people trust when they buy bitcoin? Is it the trust in your basement miner and your smartphone wallet, or is it perhaps the control through bitcoin core code? This trust amounts to 160bn USD per today.
  - bildung 6y ago
    They trust the exchanges, which is where most people buy their coins - and they don't actually own coins, they own entries in the database of the exchange. Actual OTC purchases of bitcoin are negligible in comparison and don't play a significant role in price discovery.
- gjs278 6y ago
  > And if a dispute happens, that will get resolved by courts. good luck getting your money back in a blockchain with a judges order
boomlinde 6y ago
> You could nit-pick all morning if it was strictly "true" in every which way, but this kind of selective simplification is what drives understanding. That it fundamentally isn't true is the basis of the 51% attack which is hardly a "nit". Overall I find this explanation too vague and abstract to be useful. A blockchain is a computer? Sure, you could perhaps squint and say that it is so in an abstract sense, but it doesn't really tell me anything about it. More concretely, you could say that it's a distributed append-only data structure where conflicts in additions are resolved using a consensus algorithm. You could then go on to explain how this data structure can be used to implement something like a public ledger or a contract protocol. I realize that the full article somewhat touches on these subjects between talk about "commitment computers" and "idea mazes". By going for a more concrete explanation, the weaknesses of the network would be evident and such simple falsehoods as the idea that a blockchain "removes the need to trust anything other than code" would not be propagated.
munchbunny 6y ago
I work with cryptography daily and I had to stare at that statement for several minutes to figure out what they were getting at. There's a kernel of truth in there, but in the end I think this statement is an example of being "not even wrong". It's not clear what they're trying to say. When you talk about people actually using blockchains, the last mile happens through "traditional computers", so blockchains vs. traditional computers isn't a meaningful contrast. Now if you said blockchains vs. traditional uses of cryptography, then you'd have a point. If they're trying to selectively simplify, then simplifying "traditional uses of cryptography" to "traditional computers" isn't a useful simplification. Blockchain is a way for software to decentralize decisionmaking. People are still controlling the computers. If you take away the "decentralize" part, there's plenty of non-blockchain software that does decisionmaking. If you take away the "decisionmaking" part, there's plenty of non-blockchain software that does decentralized stuff. If you take away the software part, email threads, internet forums, and even conference calls accomplish decentralized decisionmaking all the time. The following sentence is why I think the point is fundamentally wrong: For the first time, a computer system can be truly autonomous: self-governed, by its own code, instead of by people. That's just not true. People control the computers that participate in a blockchain. Those computers express the will of their operators, which in this case is participation in a decentralized decision engine. There's nothing keeping those operators from making their computers do anything else, they just won't be able to hijack the blockchain. But that's also not new, your computer acts as your agent all the time in systems where you can't hijack the rules of the system (aka most websites).
- febeling 6y ago
  Beats me how I write a comment about how a text helps understanding at the expense precision, and everyone attacks me how it wasn't even precise.
  - munchbunny 6y ago
    If that came across as an attack, I apologize. I was trying to express disagreement specifically that the text makes simplifications that don't make sense to either consumers using blockchain-derived products or developers thinking about using blockchains.

bildung 6y ago

That is just meaningless marketingspeak, and obviously wrong. Blockchains are just data, so at most "spreadsheets that can make commitments". And the latter part is also wrong, as computers are controlled by silicon logic in exactly the same way as blockchains are controlled by their code: They aren't. Someone owns the computers, and someone owns the blockchains/ their data. And if a dispute happens, that will get resolved by courts. I'm well aware of the promise of this being different with blockchains, but so far that never happened.

TheColorYellow 6y ago
Technically the blockchain that would be referred to in the GP is more than just a data type. It would also include the algorithm for deciding the next block of valid transactions. I think the reference is supposed to imply that the consensus mechanism built into the system is responsible for dictating the validity of a transaction as opposed to a human.
Taek 6y ago
> Someone owns the computers, and someone owns the blockchains/ their data. That's the beautiful part. That someone is the end user, who is running a full node. The reason that blockchains require every node to have all the data is so that every end user is in full control of the commitments made by the blockchain. Efficient? No. But does it invert the typical power relationship? Yes! Blockchains put the user in control. (fine print: this is not necessarily true for many Proof of Stake chains, also not true many lite client models (including most Bitcoin lite clients), also not true if you have on-chain governance, also not true if the blockchain makes other compromising design decisions such as auto-updating, most cryptocurrencies fall short in at least half a dozen ways)
- verdverm 6y ago
  Wrong, crypto is a scam that makes you think you have some semblance of power in a system no one uses in real life
- bildung 6y ago
  If that were the case, why can a16z profit from that?
febeling 6y ago
Who do you think the people trust when they buy bitcoin? Is it the trust in your basement miner and your smartphone wallet, or is it perhaps the control through bitcoin core code? This trust amounts to 160bn USD per today.
- bildung 6y ago
  They trust the exchanges, which is where most people buy their coins - and they don't actually own coins, they own entries in the database of the exchange. Actual OTC purchases of bitcoin are negligible in comparison and don't play a significant role in price discovery.
gjs278 6y ago
> And if a dispute happens, that will get resolved by courts. good luck getting your money back in a blockchain with a judges order

boomlinde 6y ago

> You could nit-pick all morning if it was strictly "true" in every which way, but this kind of selective simplification is what drives understanding. That it fundamentally isn't true is the basis of the 51% attack which is hardly a "nit". Overall I find this explanation too vague and abstract to be useful. A blockchain is a computer? Sure, you could perhaps squint and say that it is so in an abstract sense, but it doesn't really tell me anything about it. More concretely, you could say that it's a distributed append-only data structure where conflicts in additions are resolved using a consensus algorithm. You could then go on to explain how this data structure can be used to implement something like a public ledger or a contract protocol. I realize that the full article somewhat touches on these subjects between talk about "commitment computers" and "idea mazes". By going for a more concrete explanation, the weaknesses of the network would be evident and such simple falsehoods as the idea that a blockchain "removes the need to trust anything other than code" would not be propagated.

munchbunny 6y ago

I work with cryptography daily and I had to stare at that statement for several minutes to figure out what they were getting at. There's a kernel of truth in there, but in the end I think this statement is an example of being "not even wrong". It's not clear what they're trying to say. When you talk about people actually using blockchains, the last mile happens through "traditional computers", so blockchains vs. traditional computers isn't a meaningful contrast. Now if you said blockchains vs. traditional uses of cryptography, then you'd have a point. If they're trying to selectively simplify, then simplifying "traditional uses of cryptography" to "traditional computers" isn't a useful simplification. Blockchain is a way for software to decentralize decisionmaking. People are still controlling the computers. If you take away the "decentralize" part, there's plenty of non-blockchain software that does decisionmaking. If you take away the "decisionmaking" part, there's plenty of non-blockchain software that does decentralized stuff. If you take away the software part, email threads, internet forums, and even conference calls accomplish decentralized decisionmaking all the time. The following sentence is why I think the point is fundamentally wrong: For the first time, a computer system can be truly autonomous: self-governed, by its own code, instead of by people. That's just not true. People control the computers that participate in a blockchain. Those computers express the will of their operators, which in this case is participation in a decentralized decision engine. There's nothing keeping those operators from making their computers do anything else, they just won't be able to hijack the blockchain. But that's also not new, your computer acts as your agent all the time in systems where you can't hijack the rules of the system (aka most websites).

febeling 6y ago
Beats me how I write a comment about how a text helps understanding at the expense precision, and everyone attacks me how it wasn't even precise.
- munchbunny 6y ago
  If that came across as an attack, I apologize. I was trying to express disagreement specifically that the text makes simplifications that don't make sense to either consumers using blockchain-derived products or developers thinking about using blockchains.