Good, realistic assessment by @McKinsey consultants of how blockchain-based innovations “have started to reshape business processes,” but are still at an early development stage. Firms must focus on business value. Consistent with what I’ve seen.
Until someone can actually prove that the massive overhaul (operationally and culturally) that adopting blockchain will require is worth it, it will never happen. At the moment the juice is simply not worth the squeeze.
Blockchain is to asset transfer what the internet was to information transfer.
These arguments were all made 20 years ago. Replace block chain with internet and this could easily be an article from 1992.
I work internal strategy on the industry side now, and we've looked at blockchain for custody tracking (both for our own supply chain and as a SaaS product to sell-in to other enterprises).
Virtually all of the challenges are organizational, not technical. So making the case for a blockchain solution being a net value-add has been very difficult.
It seems even there, a trusted third party with almost any other type of centrally-controlled database, providing rights-based access to stakeholders, could accomplish the same thing with drastically better performance, for less cost, without any of the headache of blockchain.
You're right, but blockchain is a bloated, append-only database that has major performance problems, requiring drastically more complexity, with a significantly larger attack surface than conventional database. It also requires a dramatic change in the way we think of risk and liability. Those aren't the temporary downsides - they're by design. Its real benefit? It might save us the trouble and cost of *some*, but not all audits.
Blockchain is basically a digital Rube Goldberg machine for almost every application.
It's early stage technology, I'll give you that, but have you looked into Dfinity (threshold relay consensus), Kadena (proof of work at scale), or the work of Emin Gun Sirer for Avalanche (gossip) protocols?
You still have the trust the developer. You still have to trust the code base. You still have to trust the data center and the hardware. And in the case of private blockchain, you still need to trust your business partners and whoever is provisioning access.
Blockchain is "trustless" in the same way zero is the lowest number, or that friction doesn't exist in freshman physics class. It's a useful tool to bound the discussion so you can more easily describe the concept. But real world blockchain implementations don't exist in a conceptual vacuum and still require a hell of a lot of trust.
Edit: I forgot how English works
You have to trust the developer and code so much that you still have most of the traditional governance problems as you do with a database. You know that updates are going to come and that no code base is 100% perfect the first time around, so you have an upgrade path issue and scaling issue, and we've seen this happen in the cryptocurrency space already with the Bitcoin and Etherium forks. Because of these trust issues, it's hard to argue another major benefit that evangelists push - decentralization.
> You have to trust the developer and code so much that you still have most of the traditional governance problems as you do with a database.
You don't, though, because the code is public. You can just, yanno, audit it.
The DAO fork was a major success with the consensus deciding to reverse a decision on what was always intended to be an experimental network. Not sure what your problem is here, but "math" didn't fail. The developers did. Since then there's been work in smart contract auditing and security tools such as Mythril and Quantstamp as well as a plethora of new security experts chomping at the bit to audit your code before it goes on a public mainnet.
Wait what? It was a major success? Da Fuck.
It made it so Ethereum went back on its word. The fork was only successful because a small group of people overturned the contract meaning it’s not decentralized at all....
Yes your trusting the math written by the developers which sometimes is wrong.
> It made it so Ethereum went back on its word. The fork was only successful because a small group of people overturned the contract meaning it’s not decentralized at all....
It was a success because the developers proposed a change and the majority of nodes agreed with the change. That's called democracy, bruv.
> Yes your trusting the math written by the developers which sometimes is wrong.
The DAO fork had nothing to do with math and the code is public, so I suggest you be a little less trusting and do your own audit work... or hire an insured third party you DO trust to do it for you.
What? Ya it had to do with math, there was an exploit in the code. Coding is a set of logic based rules which is similar to math. I’m abstracting for reddit comments.
Lol ok so it’s democracy based on who owns enough for proof of stake now? because I’m pretty sure the Byzantine fault tolerance of the proof of stake concept did not take into account mass coordinated actions. In short, if a group all acts in one way to override for good, they can use it for bad too.
> What? Ya it had to do with math, there was an exploit in the code. Coding is a set of logic based rules which is similar to math. I’m abstracting for reddit comments.
Alrighty, you're outclassed here son. This conversation's over.
> Lol ok so it’s democracy based on who owns enough for proof of stake now? because I’m pretty sure the Byzantine fault tolerance of the proof of stake concept did not take into account mass coordinated actions. In short, if a group all acts in one way to override for good, they can use it for bad too.
You really shouldn't make assertions on things you do not understand. Perhaps try an inquisitive approach next time? You might learn something!
This is very true, and the reason that legally binding smart contracts are required and a next step of enterprise blockchain. As someone who works in the space, it's brilliant for a few use cases. Supply chain and cross-border settlement being the two best for now.
Can you explain how it works in supply chain and why it would be better than conventional methods? What does it offer that the current method doesn't? For example, wouldn't you still need physical verification of a product?
In my scenario, the supply chain needs to be audited at every step according to different criteria. Blockchain allows the audits, or the validation that these audits occured, to be shared with the consortium, with permissions set in advance according to business and governance rules (very efficient), in a manner that does not require repetition or repeat audits. Smart contracts can also be implemented along a complex chain, in a way where if conditions are met, then actions occur. Further, tokens can be created, representing assets, via smart contracts, adding more possibilities.
Like Equifax? There's paradigm shift in how you can design software that trustless and decentralized systems provide. It takes the control out of large monolithic and often-hacked central authorities and into the hands of a person. Self-sovereign identity is a major use-case, and with it the ability to grant visibility to your personal data. Granted, there's a tremendous amount of work to be done to get it into the hands of the general public, but to ignore the potential there is ... well it's worth the hedge at the very least.
the inefficiency is by design. the originators of blockchain are mostly libertarian types who were more interested in decentralized ways to store and carry money that would be outside the reach of governments and secure by consensus.
You're right that they're inefficient, especially bitcoin. But blockchain solutions, despite their inefficiencies, [do have value.](https://www.economist.com/the-world-if/2017/07/15/disrupting-the-trust-business). No one knows the extent of such value they will have though. I personally don't think they're going to replace traditional forms of currency for example. They're typically far too volatile in terms of value, too easy to lose, too easy to steal, etc.