By that time I had sufficient interest in crypto to take the time to read and understand what it was about, and how it was different.
I think you’re underestimating the amount of insider knowledge you’ve gained and the cost of attaining that insider knowledge. Eliezer is certainly surrounded by smart people and MIRI received half of its donations in crypto. Yet, they still did not invest in crypto. I think this is because they lacked deep insider knowledge about cryptocurrencies.
I think what usually happens is ― a rationalist hears about shiny thing X, investigates it thoroughly, concludes that in fact X is shiny, then takes action. There are a lot of things worth investigating at any given point ― VR, homotopy type theory, blockchain based technologies, genetics, deep learning, functional programming and so on. Unfortunately, you have to invest a lot of time to gain insider knowledge in any of these things.
I think people get stuck in the investigation process and don’t proceed further due to the temporal cost of attaining that knowledge.
In aggregate, take their ideas seriously even when they might not take their own ideas seriously.
To a crypto outsider like myself, Tezos still feels like another ICO scam. I googled Tezos for a while and still do not have any idea why this has a strong “signal”. I believe you’re advocating investing some money in Tezos since smart people think it’s cool and skip the investigation step.
I’m still confused about whether it is rational to skip the investigation step and blindly invest in Tezos without understanding how it is different.
A single bug in an Etherum dapp brought down The DAO, bug’s in dapps seem to be very costly.
Tezos uses a programming language for dapps that support proves of correctness. For normal programming few people care about proofs of correctness. Toyota started caring when their vehicles accelerated and some military contractors care but otherwise most people aren’t serious enough to care about bug freeness to program with proves of correctness.
A second issue with crypto-currency is transaction costs.
As the time of this writing the media bitcoin transaction costs $2.56 while the medium fee for Etherum is $0.27. Both are too high for a lot of the promised ways to use the blockchain.
The Bitcoin protocol’s solution is the Lightning Network that can transfer money but won’t run dapps.
Ethereum goes for sharding.
Tezos bets on the fact that mathematical advances in recently discovered STARKs will allow STARKs to be calculated faster and that this will allow scalling without sharding. I don’t enough enough details to speak about the merits of that decision.
The cost is actually not that high. I spent may be 5-10 hours researching Bitcoin (about 5 hours before I invested; 10 hours total). There aren’t that many things one can invest 5-10 hours into and instantly make money. In fact, none of your examples come even close. Those are huge fields that you need to sink a ton of hours into before you can start reaping rewards. And none of them have direct monetary rewards like crypto.
I googled Tezos for a while and still do not have any idea why this has a strong “signal”.
Ok, then may be I’ve overestimating how easy it is to look at crypto coins and have a rough guess at how good they are. It’s also possible I’m overestimating my own skill at it; it’s not like I have that many data points yet. (Not knowing you very well, I won’t make hypotheses about you.)
I believe you’re advocating investing some money in Tezos since smart people think it’s cool and skip the investigation step.
Absolutely not. I’m advocating reading the thinking and research that smart people did and following their pointers, but then also checking for sanity. Sanity checks are usually pretty easy to do, but if you can’t do them, then this strategy just won’t work.
Sanity checks are usually pretty easy to do, but if you can’t do them, then this strategy just won’t work.
I concede that Bitcoin is pretty easy to understand and sanity check (merkle trees aren’t that hard to wrap your head around ― I would have invested in Bitcoin in 2012 when I heard about it, but I was in high school and had no disposable income). But sanity checking Tezos is much harder:
It turns out that a silver bullet for chain validation is right on the horizon and under active research: recursive SNARKs. SNARKs, which stands for succinct non-interactive zero-knowledge proofs of knowledge are the technology used in Zcash for protecting the privacy of transactions (if you’re already objecting that SNARKs require a trusted setup, please bear with us, we have good news for you).
… This very counter intuitive possibility is a consequence of the PCP theorem. Rather than try to engage in economic “bets” that the transaction has been properly validated we can obtain true cryptographic assurance.
I don’t know enough to sanity check their scaling strategy or what makes Tezos unique in this respect. Even Ethereum itself is experimenting with zk-SNARKS and Vitalik Buterin wrote a series of articles explaning SNARKS on his medium blog which starts off with an article that says:
You’re not expected to understand everything here the first time you read it, or even the tenth time; this stuff is genuinely hard. But hopefully this article will give you at least a bit of an idea as to what is going on under the hood.
So far Tezos’ Unique Selling Point seems to be that they’ve used OCaml to implement some way of doing formal verification on smart contracts. But this alone seems like insufficient evidence to conclude that Tezos is shiny. To make matters worse, there seems to be some internal conflict between the co-founders of Tezos.
I don’t know how I could possibly successfully sanity check Tezos without understanding SNARKS and reading their whitepaper (both of which require non-trivial prerequisites).
Perhaps you have the background or access to people who have the background to evaluate Tezos properly. But I certainly do not and I would argue the cost is high enough.
To clarify, I’m not arguing that Tezos is shiny right now. It was shiny during ICO, and now it’s possible I wouldn’t buy it.
But OCaml was certainly one of the big factors that lept at me. It signalled that they were serious about writing code that could be proven to be correct. (This is in contrast to Ethereum’s Solidity.) Basically: 1) I looked at what they were promising to do, and it seemed to have enough additional good things beyond Ethereum, 2) nobody I saw was disputing their tech, so I didn’t have to understand it in detail, just enough to verify that it seemed innovative, and 3) they had a credible team (this was before the conflict). To me that was sufficient at the time, given that much much shittier ICOs were making money hand over first.
I think you’re underestimating the amount of insider knowledge you’ve gained and the cost of attaining that insider knowledge. Eliezer is certainly surrounded by smart people and MIRI received half of its donations in crypto. Yet, they still did not invest in crypto. I think this is because they lacked deep insider knowledge about cryptocurrencies.
I think what usually happens is ― a rationalist hears about shiny thing X, investigates it thoroughly, concludes that in fact X is shiny, then takes action. There are a lot of things worth investigating at any given point ― VR, homotopy type theory, blockchain based technologies, genetics, deep learning, functional programming and so on. Unfortunately, you have to invest a lot of time to gain insider knowledge in any of these things.
I think people get stuck in the investigation process and don’t proceed further due to the temporal cost of attaining that knowledge.
To a crypto outsider like myself, Tezos still feels like another ICO scam. I googled Tezos for a while and still do not have any idea why this has a strong “signal”. I believe you’re advocating investing some money in Tezos since smart people think it’s cool and skip the investigation step.
I’m still confused about whether it is rational to skip the investigation step and blindly invest in Tezos without understanding how it is different.
The core case for Tezos is the following:
A single bug in an Etherum dapp brought down The DAO, bug’s in dapps seem to be very costly.
Tezos uses a programming language for dapps that support proves of correctness. For normal programming few people care about proofs of correctness. Toyota started caring when their vehicles accelerated and some military contractors care but otherwise most people aren’t serious enough to care about bug freeness to program with proves of correctness.
A second issue with crypto-currency is transaction costs.
As the time of this writing the media bitcoin transaction costs $2.56 while the medium fee for Etherum is $0.27. Both are too high for a lot of the promised ways to use the blockchain.
The Bitcoin protocol’s solution is the Lightning Network that can transfer money but won’t run dapps.
Ethereum goes for sharding.
Tezos bets on the fact that mathematical advances in recently discovered STARKs will allow STARKs to be calculated faster and that this will allow scalling without sharding. I don’t enough enough details to speak about the merits of that decision.
Another issue with Tezos is governance: https://www.coindesk.com/tezos-new-board-mean-money/
The cost is actually not that high. I spent may be 5-10 hours researching Bitcoin (about 5 hours before I invested; 10 hours total). There aren’t that many things one can invest 5-10 hours into and instantly make money. In fact, none of your examples come even close. Those are huge fields that you need to sink a ton of hours into before you can start reaping rewards. And none of them have direct monetary rewards like crypto.
Ok, then may be I’ve overestimating how easy it is to look at crypto coins and have a rough guess at how good they are. It’s also possible I’m overestimating my own skill at it; it’s not like I have that many data points yet. (Not knowing you very well, I won’t make hypotheses about you.)
Absolutely not. I’m advocating reading the thinking and research that smart people did and following their pointers, but then also checking for sanity. Sanity checks are usually pretty easy to do, but if you can’t do them, then this strategy just won’t work.
I concede that Bitcoin is pretty easy to understand and sanity check (merkle trees aren’t that hard to wrap your head around ― I would have invested in Bitcoin in 2012 when I heard about it, but I was in high school and had no disposable income). But sanity checking Tezos is much harder:
― Scaling Tezos
I don’t know enough to sanity check their scaling strategy or what makes Tezos unique in this respect. Even Ethereum itself is experimenting with zk-SNARKS and Vitalik Buterin wrote a series of articles explaning SNARKS on his medium blog which starts off with an article that says:
― Exploring Elliptic Curve Pairings
So far Tezos’ Unique Selling Point seems to be that they’ve used OCaml to implement some way of doing formal verification on smart contracts. But this alone seems like insufficient evidence to conclude that Tezos is shiny. To make matters worse, there seems to be some internal conflict between the co-founders of Tezos.
I don’t know how I could possibly successfully sanity check Tezos without understanding SNARKS and reading their whitepaper (both of which require non-trivial prerequisites).
Perhaps you have the background or access to people who have the background to evaluate Tezos properly. But I certainly do not and I would argue the cost is high enough.
To clarify, I’m not arguing that Tezos is shiny right now. It was shiny during ICO, and now it’s possible I wouldn’t buy it.
But OCaml was certainly one of the big factors that lept at me. It signalled that they were serious about writing code that could be proven to be correct. (This is in contrast to Ethereum’s Solidity.) Basically: 1) I looked at what they were promising to do, and it seemed to have enough additional good things beyond Ethereum, 2) nobody I saw was disputing their tech, so I didn’t have to understand it in detail, just enough to verify that it seemed innovative, and 3) they had a credible team (this was before the conflict). To me that was sufficient at the time, given that much much shittier ICOs were making money hand over first.