I'm really confused by this post. It seems to me like: a) the worst place to mine anything is in a general purpose cloud with resource sharing out of your control and b) burning as many resources of a specific type as available is (I'd expect) the worst kind of customer behaviour for AWS since underutilization and overprovisioning is free money. So why does this official post exist?
I think the ability to mine is incidental here. This is more likely to be targeted at dapps developers by offering them an easy setup for both test and production environment.
Azure has this for a some time with Ethereum. Blockchain as a Service they call it. It's mostly for being able to test and develop things I assume, not for mining. But I'm still not fully understanding the value of these offerings. Ethereum does have multiple test networks already. One can develop in private, that seems the main advantage.
Has anyone here benefited from deploying a private Ethereum network?
I'm struggling to brainstorm practical use cases for my projects, other than perhaps for smart contract confidentiality reasons. It feels like the cost of maintaining more infrastructure and the potential waste of computing resources through mining outweigh any benefits but I am likely missing some great use cases.
> The use of blockchain technology will hopefully avoid the all too common problem that occurs when a claim comes through from the provider and it mismatches with the information the insurance company has in its provider directory.
People keep coming back with this use case but it always baffles me. Certainly you need to have some amount of trust in these parties not to falsify data? If so, why do you need a Blockchain instead of basically anything else that would be cheaper and simpler?
I hoped to find answers in this Forbes articles but all I get are vague claims like:
>The use of blockchain technology will hopefully avoid the all too common problem that occurs when a claim comes through from the provider and it mismatches with the information the insurance company has in its provider directory.
>CMS found 52% of Medicare Advantage plan provider “directory locations had at least one inaccuracy.”
How will the blockchain magically make sure that everybody updates all information? And if people suddenly somehow make sure to publish and update the info then why can't they do the same thing with boring old technology?
>“With increasing state and federal requirements relating to provider data maintenance and quality, tackling the high cost and redundancy in this space is a logical starting point"
Backups and redundancy existed before the Blockchain, also I'm sure nobody wants to publish private user data publicly for all to see so it's not like random users around the world are going to mirror your database.
So you end up with a "Blockchain" where only trusted people can access and contribute. What's the advantage over spinning up a PostgreSQL intance? I mean when you read the Quorum wiki[1] it reads more like a database setup guide than spinning a bitcoin node:
>The P2P layer has been modified to only allow connections to/from permissioned nodes.
And just like that the entire point of the Blockchain vanishes.
>‘Proof of work’ consensus algorithm has been replaced with ‘QuorumChain’ consensus, a vote-based consensus mechanism.
Surely I hope that medicare providers obey to some top-down rules to operate, not a consensus on a Blockchain. And given that all nodes are whitelisted anyway I'm not sure why such a voting is even required.
It's pretty crazy to me that when I read articles like this one on Forbes or the Quorum website I see pages after pages of hype and praise and yet so far nobody is capable of giving me one simple use case demonstrating the practical usefulness of the technology outside of cryptocurrencies.
The more I read about it the more I become convinced that the only real feature of the Blockchain is being a magic word that makes suits throw money at you. Maybe I'm just too dumb to understand it.
I had been thinking that I'm just too dumb to understand it was well, which still may be true. The conclusion I have come to is that blockchain is a solution looking for a problem, but every problem that people have come up with there are already solutions for that are far easier to implement and cost less to run.
Note that a private blockchain doesn't have to be expensive. The "cost" comes from the proof-of-work problem (the Concensus Algorithm) being made more difficult to solve over time, which occurs because there is an arms race between the miners to solve the block first.
If you have a private blockchain with just a handful of miners on it who are not racing to win, they can use much simpler computing power and keep the difficulty of the proof of work problem low.
In fact, in a private blockchain you don't even need to use proof of work. You can come up with some other concensus algorithm which is even cheaper when it comes to CPU (proof of stake is one popular algorithm, one can think up of many more)
Perhaps in digital advertising? In some cases you have multiple actors (say 10) with data simultaneously created by 5. Ensuring these agree today is hard and knowing who to trust is hard. Consensus algorithms and Blockchain’s can help the 5 + 5 here.
Distributed databases with a consensus algorithm in the backend existed for quite a while. What does this provide over (search for: raft log database, and pick one example).
The press release is vague for the same reason why I can't directly answer some of your questions/challenges.
However, the fact that we're using a permissioned network does not automatically say that trust is absolutely granted. In this case it is more of a wall for a rather large and weird garden and at this stage of development it is helpful to know at least who is walking around in the garden. What they are up to is a different matter.
Also, the press release mentions the problem of many provider directories existing and not every directory contains the same information. Provider data is a tent pole in the American healthcare industry and it is used in just about every area of it, from claims processing to personal health record management.
You ask why not just have a central database for it? Well, the answer is simply "why do you think there are so many directories already?" The field is littered with directories that aimed to be "The single master directory for everyone!" There is a large amount of "not invented here syndrome" and an equal amount of "we have 20 years of data models that wont fit your commercial directory." So every large payer ends up building their own and then spends a lot of time and effort maintaining the data in it. Other consumers/producers of provider data may also maintain their own directories or subscribe to a commercial product.
The network we constructed is not simply yet another provider master database but-on-a-blockchain. It goes beyond that and really does use key features of what most consider to be a blockchain's major benefits.
I apologize for not being able to share details of what I consider to be really neat stuff but more news is on the horizon.
> You ask why not just have a central database for it? Well, the answer is simply "why do you think there are so many directories already?" The field is littered with directories that aimed to be "The single master directory for everyone!"
How does a block chain solve this better than any other solution? You still have to get all of the participants to adopt it, and that's an organizational problem that doesn't seem to me to be any easier than with any other solution.
Running code on the Ethereum network is pretty expensive, so I imagine a private network would be much cheaper. Of course, just running your apps in a more centralized fashion is significantly cheaper. There may be some edge cases where you don't need the same levels of trust as the main Ethereum network but also don't have complete trust over the participants.
I've always been a big fan of the technology behind cryptocurrencies, but have also always struggled with the use case for private blockchains.
Alternative (authority based) consensus mechanisms are available for private chains [1][2] so you don't need to mine. Both the Kovan and Rinkeby test nets use proof of authority based consensus mechanisms.
Private networks are useful for test environments and proof of concept development.
Apart from that a private blockchain is basically a glorified git repo, so in general it's pretty good that big companies are discovering that cryptographically linked immutable data structures are useful, but you are correct that the impact will be quite limited.
Note that a private blockchain doesn't have to be expensive. The "cost" comes from the proof-of-work problem (the Concensus Algorithm) being made more difficult to solve over time, which occurs because there is an arms race between the miners to solve the block first.
If you have a private blockchain with just a handful of miners on it who are not racing to win, they can use much simpler computing power and keep the difficulty of the proof of work problem low.
In fact, in a private blockchain you don't even need to use proof of work. You can come up with some other concensus algorithm which is even cheaper when it comes to CPU (proof of stake is one popular algorithm, one can think up of many more)
The benefits of a private blockchain is that it's a way for multiple companies/organization to share specific data with each other without them having to rely on a traditional database which can only be maintained by one of them.
Why would they not want to rely on a single database?
It's high risk: One one company maintains it, so the others are relying on that company to keep it up. This isn't easy. If that database goes down for some reason the other companies can only wait while the database is fixed. This waiting can cost millions of dollars a minute. With a blockchain based database all the companies have their own copy of the data so one company having an outage will not affect the others
> One one company maintains it, so the others are relying on that company to keep it up. This isn't easy. If that database goes down for some reason the other companies can only wait while the database is fixed.
Or you could use a system with a well designed replication mechanism and designed to work offline. For example http://couchdb.apache.org/ should provide the same functionality you're after. If you have multiple entities that don't trust each other 100%, you can structure the data to be essentially an event log and filter the replication to reject modification/deletion.
"Last year, Yiannas asked his team to trace a package of sliced mangoes back to their source. He says it took them six days, 18 hours, and 26 minutes. After Walmart used the software developed with IBM to track mangoes from a farm in Mexico to U.S. stores over a 30-day period, the same exercise took a mere 2.2 seconds."
If you don't think immutable distributed ledgers won't help improve food safety, among thousands of other use cases, you probably haven't been paying attention to the current E. Coli outbreak in the US where the FDA has no idea where the contaminated food is coming from:
https://www.washingtonpost.com/news/to-your-health/wp/2018/0...
Based on my observation we are reaching PeakBlockChain(tm):
1. Half of the posts in my LinkedIn use words block chain - they are not coming from usual P&D crowd. I'm seeing lawyers post "Where are all the #ladies of #blockchain. So inspired by #blockchain diversity efforts!"
2. I cannot go to a restaurant/bar/cafe/coffee shop in NYC without hearing someone next to me talking about blockchain and awesome stuff they are going to do to be riiiich riiich rich on #blockchain. Yesterday, I overheard two Mormon kids on a mission talk about blockchain over lunch.
3. IBM's ads are all about blockchain.
4. And now Amazon is jumping into Blockchain-in-a-box
5. Despite being 3 years old a compelling use case for smart-contracts has yet to emerge, nor has any use for blockchain become common place except as a clandestine currency.
CryptoKitties (Yes, cute little cartoon cats) that you can trade, auction, gift, breed and more with contract services on Ethereum might be the fun, fun, fun and compelling use case you're looking for, see https://github.com/cryptocopycats/awesome-cryptokitties/tree...
I sincerely hope I detect sarcasm. Reminds me of tulips. takes shot for cryptocurrency bingo How's that beanie baby investment working out for everyone? ;p
The best use case for blockchain that I've seen so far is Dada.nyc, a platform where you can buy art with Ethereum. What I love about it is that in this marketplace, every time an artwork is resold, the artist gets paid. Opposed to traditional art where you get paid once and never see anything again when your art appreciates wildly after you die
FYI: To get started with Blockchains from Scratch (Zero) coding / building your own in thirty lines of Ruby, Python, Go, JavaScript, ... I collect starter articles and free code samples at the Awesome Blockchains page @ https://github.com/openblockchains/awesome-blockchains
46 comments
[ 86.7 ms ] story [ 2031 ms ] threadIt lets Sales send this article to people who write in asking "can I build my blockchain on AWS", which lets Amazon make some money off the craze.
Literally the first use case mentioned.
I'm struggling to brainstorm practical use cases for my projects, other than perhaps for smart contract confidentiality reasons. It feels like the cost of maintaining more infrastructure and the potential waste of computing resources through mining outweigh any benefits but I am likely missing some great use cases.
We have done a lot of analysis and can see large benefits of using Ethereum based private network in this way.
https://www.forbes.com/sites/brucejapsen/2018/04/02/unitedhe...
How?!
I hoped to find answers in this Forbes articles but all I get are vague claims like:
>The use of blockchain technology will hopefully avoid the all too common problem that occurs when a claim comes through from the provider and it mismatches with the information the insurance company has in its provider directory.
>CMS found 52% of Medicare Advantage plan provider “directory locations had at least one inaccuracy.”
How will the blockchain magically make sure that everybody updates all information? And if people suddenly somehow make sure to publish and update the info then why can't they do the same thing with boring old technology?
>“With increasing state and federal requirements relating to provider data maintenance and quality, tackling the high cost and redundancy in this space is a logical starting point"
Backups and redundancy existed before the Blockchain, also I'm sure nobody wants to publish private user data publicly for all to see so it's not like random users around the world are going to mirror your database.
So you end up with a "Blockchain" where only trusted people can access and contribute. What's the advantage over spinning up a PostgreSQL intance? I mean when you read the Quorum wiki[1] it reads more like a database setup guide than spinning a bitcoin node:
>The P2P layer has been modified to only allow connections to/from permissioned nodes.
And just like that the entire point of the Blockchain vanishes.
>‘Proof of work’ consensus algorithm has been replaced with ‘QuorumChain’ consensus, a vote-based consensus mechanism.
Surely I hope that medicare providers obey to some top-down rules to operate, not a consensus on a Blockchain. And given that all nodes are whitelisted anyway I'm not sure why such a voting is even required.
It's pretty crazy to me that when I read articles like this one on Forbes or the Quorum website I see pages after pages of hype and praise and yet so far nobody is capable of giving me one simple use case demonstrating the practical usefulness of the technology outside of cryptocurrencies.
The more I read about it the more I become convinced that the only real feature of the Blockchain is being a magic word that makes suits throw money at you. Maybe I'm just too dumb to understand it.
[1] https://github.com/jpmorganchase/quorum/wiki
If you have a private blockchain with just a handful of miners on it who are not racing to win, they can use much simpler computing power and keep the difficulty of the proof of work problem low.
In fact, in a private blockchain you don't even need to use proof of work. You can come up with some other concensus algorithm which is even cheaper when it comes to CPU (proof of stake is one popular algorithm, one can think up of many more)
However, the fact that we're using a permissioned network does not automatically say that trust is absolutely granted. In this case it is more of a wall for a rather large and weird garden and at this stage of development it is helpful to know at least who is walking around in the garden. What they are up to is a different matter.
Also, the press release mentions the problem of many provider directories existing and not every directory contains the same information. Provider data is a tent pole in the American healthcare industry and it is used in just about every area of it, from claims processing to personal health record management.
You ask why not just have a central database for it? Well, the answer is simply "why do you think there are so many directories already?" The field is littered with directories that aimed to be "The single master directory for everyone!" There is a large amount of "not invented here syndrome" and an equal amount of "we have 20 years of data models that wont fit your commercial directory." So every large payer ends up building their own and then spends a lot of time and effort maintaining the data in it. Other consumers/producers of provider data may also maintain their own directories or subscribe to a commercial product.
The network we constructed is not simply yet another provider master database but-on-a-blockchain. It goes beyond that and really does use key features of what most consider to be a blockchain's major benefits.
I apologize for not being able to share details of what I consider to be really neat stuff but more news is on the horizon.
How does a block chain solve this better than any other solution? You still have to get all of the participants to adopt it, and that's an organizational problem that doesn't seem to me to be any easier than with any other solution.
This sounds like you're looking to pull a https://xkcd.com/927/.
I've always been a big fan of the technology behind cryptocurrencies, but have also always struggled with the use case for private blockchains.
Private networks are useful for test environments and proof of concept development.
Apart from that a private blockchain is basically a glorified git repo, so in general it's pretty good that big companies are discovering that cryptographically linked immutable data structures are useful, but you are correct that the impact will be quite limited.
[1]: https://github.com/ethereum/EIPs/issues/225 [2]: https://wiki.parity.io/Aura.html
In fact, in a private blockchain you don't even need to use proof of work. You can come up with some other concensus algorithm which is even cheaper when it comes to CPU (proof of stake is one popular algorithm, one can think up of many more)
The benefits of a private blockchain is that it's a way for multiple companies/organization to share specific data with each other without them having to rely on a traditional database which can only be maintained by one of them.
Why would they not want to rely on a single database? It's high risk: One one company maintains it, so the others are relying on that company to keep it up. This isn't easy. If that database goes down for some reason the other companies can only wait while the database is fixed. This waiting can cost millions of dollars a minute. With a blockchain based database all the companies have their own copy of the data so one company having an outage will not affect the others
Or you could use a system with a well designed replication mechanism and designed to work offline. For example http://couchdb.apache.org/ should provide the same functionality you're after. If you have multiple entities that don't trust each other 100%, you can structure the data to be essentially an event log and filter the replication to reject modification/deletion.
If you don’t need the trust model part of the blockchain you do not need a blockchain.
"Last year, Yiannas asked his team to trace a package of sliced mangoes back to their source. He says it took them six days, 18 hours, and 26 minutes. After Walmart used the software developed with IBM to track mangoes from a farm in Mexico to U.S. stores over a 30-day period, the same exercise took a mere 2.2 seconds."
https://newfoodeconomy.org/blockchain-food-traceability-walm...
If you don't think immutable distributed ledgers won't help improve food safety, among thousands of other use cases, you probably haven't been paying attention to the current E. Coli outbreak in the US where the FDA has no idea where the contaminated food is coming from: https://www.washingtonpost.com/news/to-your-health/wp/2018/0...
Yes, absolutely. I liked it a lot as it added whimsy to a subject that could easily be written in a much drier style.
Appreciate it!
1. Half of the posts in my LinkedIn use words block chain - they are not coming from usual P&D crowd. I'm seeing lawyers post "Where are all the #ladies of #blockchain. So inspired by #blockchain diversity efforts!"
2. I cannot go to a restaurant/bar/cafe/coffee shop in NYC without hearing someone next to me talking about blockchain and awesome stuff they are going to do to be riiiich riiich rich on #blockchain. Yesterday, I overheard two Mormon kids on a mission talk about blockchain over lunch.
3. IBM's ads are all about blockchain.
4. And now Amazon is jumping into Blockchain-in-a-box
assuming you actually meant this as a criticism separate from the one after: payment channels rely on smart contracts to work
Make it easy to use and let people discover that it’s just not really needed for the cat majority of applications.
https://www.techspot.com/news/74246-amazon-granted-patent-tr...