Why blockchain is not yet working (2018)
as1ndu.xyz
Why blockchain is not yet working (2018)
1–10 of 330 posts
Re: Why blockchain is not yet working (2018)
#24. Consensus Algorithms Are Wack
The popular blockchains only don’t offer transaction finality i.e you will never be 100% sure that you have gotten funds.
You can only be probabilistically to a high degree (but with zero guarantee) that you have received your funds. This is because in major consensus algorithm called Nakamoto Consensus, the miners elect the chain with the most computational work that are with in the same protocol i.e consensus rules.
For instance imagine your bank tells you that you that, “there is a 95% chance that we might have your life savings, however you may also not poses it in 5 years in case we discovered that the chain with the most work does not have your transaction in it”
Yes, I don’t think anyone wants to keep their life savings in such a bank.
Sure, and the probability that all of your atoms will teleport one foot forward is not zero. But it's sufficiently zero that we say it's zero, and we're rarely proven wrong.
It took awhile to become comfortable with probabilistic algorithms, but some (small) familiarity with physics helped. Nature is probabilistic, and seems to work.
This isn't a great counterargument, but honestly none of these arguments seem very persuasive. I'm not a blockchain fan, but it's like... "if someone steals your wallet, they'll get all of your money!" Well, duh. Everyone knows this and seems fine with it.
1. Scalability Issues The current bandwidth, i.e the data that can be transmitted through a blockchain with in a given time is simply too bounded for main stream adoption.
Ethereum for instance does a maximum of 13 transactions per second. Visa on the other hand can do 24,000 Transaction per second and peak over 40,000 transactions per second.
No mention of Lightning network to solve this. I've been out of the crypto game for years and even I know "Lightning is supposedly the solution." Did that ever pan out?
6. Price Volatility Very few people simply don’t have the stomach to hold a financial instrument that drops by over 50% with in a few months.
Re: Why blockchain is not yet working (2018)
#34. Consensus Algorithms Are Wack The popular blockchains only don’t offer transaction finality i.e you will never be 100% sure that you have gotten funds. You can only be probabilistically to a high degree (but with zero guarantee) that you have received your funds. This is because in major consensus algorithm called Nakamoto Consensus, the miners elect the chain with the most computational work that are with in the…
> Nature is probabilistic, and seems to work.
I think what you mean is that the currently most successful model of the physical world is probabilistic, but this doesn't mean that nature is.
Re: Why blockchain is not yet working (2018)
#44. Consensus Algorithms Are Wack The popular blockchains only don’t offer transaction finality i.e you will never be 100% sure that you have gotten funds. You can only be probabilistically to a high degree (but with zero guarantee) that you have received your funds. This is because in major consensus algorithm called Nakamoto Consensus, the miners elect the chain with the most computational work that are with in the…
> Everyone knows this and seems fine with it.
Are you sure? I believe a lot of people interact with blockchains through services like coinbase, without managing their own wallets.
Re: Why blockchain is not yet working (2018)
#5To the scalability issues section I'd add the environmental aspect. It has already been reported that by some estimates Bitcoin today consumes more energy than Argentina, and that's with an estimated 1 million active wallets per day. Argentina is home to about 50 million people.
This means that in order to have Bitcoin serve as the daily currency for all of Argentina's 50 million people would at least 50-tuple its energy consumption. This is comically absurd. It's thankfully not achievable or economical with the energy resources we currently have available, because pursuing a 4,900% increase in energy consumption would be disastrous.
If crypto sees any widespread usage I think it'll be as the backbone of financial systems, for big transactions between institutions. In that space I don't see much incentive for people to adopt it— there's already a good level of trust amongst the parties, various levels of transaction verification, and it's mostly all digital already.
[1]: https://www.bbc.com/news/technology-56012952
[2]: https://www.bitcoinmarketjournal.com/how-many-people-use-bit...
Re: Why blockchain is not yet working (2018)
#6The "lack of intuitive private key management" is worth expanding on.
Would it not be nice if you could, maybe for a small fee, have some trusted entity store your private key for you? One could perhaps imagine storing a printout of the private key in some kind of physical "safe deposit" box, to be maintained by such an entity? But maybe that is to cumbersome and we can do so entirely digitally? In that case I guess that said entity would need to demonstrate a proven track record of safely keeping a record of your net worth, as well as the various in- and out-going transactions between its different customers.
With the obvious possibilities for abuse, it is clear that such an industry would see a fair amount of regulations, and one can foresee never-ending tension between government, the established corporations, and potentially disruptive newcomers....
Re: Why blockchain is not yet working (2018)
#74. Consensus Algorithms Are Wack The popular blockchains only don’t offer transaction finality i.e you will never be 100% sure that you have gotten funds. You can only be probabilistically to a high degree (but with zero guarantee) that you have received your funds. This is because in major consensus algorithm called Nakamoto Consensus, the miners elect the chain with the most computational work that are with in the…
> Nature is probabilistic, and seems to work. I think what you mean is that the currently most successful model of the physical world is probabilistic, but this doesn't mean that nature is.
Mm. That's an interesting way of looking at it. But the assumptions are laid out by Feynman: https://www.youtube.com/watch?v=QRE0GxT6Zbw&list=PLgRI7D_FXE...
Although I'd love to urge everyone to sit through all six hours, I recognize this isn't too practical. So -- with apologies to Feynman -- I'll have to do a poor job of translating:
- The scientific method means that we are never certain we're right; we're only certain we're wrong. But (as with Newton's laws) it's pretty incredible we can last so long before we're proven incorrect!
- The world seems to obey mathematics. But math != physics. One helps the other; discoveries in math have applied to physics, and vice-versa.
- The heart of nature seems to be a purely mathematical construct, from which there is no escape. Many people have tried to model gravity as something other than the gravitational equation, e.g. suppose at every point in time there are particles shooting every direction simultaneously. Then gravity is merely us blocking these particles -- earth with respect to the sun blocks a certain fraction of particles towards the sun, and in fact inversely as the square of the distance. No math necessary! Unfortunately, doesn't quite work -- if it were true, the earth would have slowed in its orbit due to blocking the particles, and it wouldn't have lasted ~billions of years. So that's the end of that theory.
- The double slit experiment seems to prove that the world has no other model other than probabilistic (see lecture 6). There is some fundamental limitation preventing us from narrowing down nature further than this threshold.
- Every test designed to break this assumption has failed.
Therefore, I say that the physical world is probabilistic, and I will wager any sum of money you'd like to bet that it will still be true within, say, 200 years.
There is one interesting alternative: if everything in the universe is pre-destined. The complete absence of any kind of free will whatsoever. Every atom, every electron, every single subatomic particle, all predestined from the moment of the big bang til now. That seems to be the only way that it can't be probabilistic.
Re: Why blockchain is not yet working (2018)
#8To the scalability issues section I'd add the environmental aspect. It has already been reported that by some estimates Bitcoin today consumes more energy than Argentina, and that's with an estimated 1 million active wallets per day. Argentina is home to about 50 million people. This means that in order to have Bitcoin serve as the daily currency for all of Argentina's 50 million people would at least 50-tuple its en…
> This means that in order to have Bitcoin serve as the daily currency for all of Argentina's 50 million people would at least 50-tuple its energy consumption.
How do you justify this linear extrapolation?
Re: Why blockchain is not yet working (2018)
#9The "lack of intuitive private key management" is worth expanding on. Would it not be nice if you could, maybe for a small fee, have some trusted entity store your private key for you? One could perhaps imagine storing a printout of the private key in some kind of physical "safe deposit" box, to be maintained by such an entity? But maybe that is to cumbersome and we can do so entirely digitally? In that case I guess…
You just recreated consumer banking.
Re: Why blockchain is not yet working (2018)
#104. Consensus Algorithms Are Wack The popular blockchains only don’t offer transaction finality i.e you will never be 100% sure that you have gotten funds. You can only be probabilistically to a high degree (but with zero guarantee) that you have received your funds. This is because in major consensus algorithm called Nakamoto Consensus, the miners elect the chain with the most computational work that are with in the…
> Nature is probabilistic, and seems to work. I think what you mean is that the currently most successful model of the physical world is probabilistic, but this doesn't mean that nature is.
Yes, I completely agree. I think however that the point was that whatever we build out of our best available physical theories ultimately relies on probabilistic grounds.
Now on the other hand, we don't use quantum physics for every single human project out there, and for many human scale phenomena, we can act without concious probabilistic representations.