I Don’t Believe in Blockchain
tbray.org
tbray.org
Having studied the BC stack, I would say that the biggest obstacle will be to get engineers to dare try something new. To many SW-engineers on the web-side today, don't really have sufficient skills to understand BC-technology. I don't mean to offend, just state what I see. We saw the same thing with p2p-tech before that.
Solidity as a programming language is an interesting abstraction and perhaps lowers the barrier somewhat, but it will take many years before the skills available match that for what is required to gain momentum on a broad scale.
If you're an engineer and wonder why mgmt keeps bringing it up, its because this will potentially help them build a platform (two-sided) market. If that occurs and you constructed it correctly, your company may well be able to skip on taxes as well, as a side-effect.
I find those sentences make an interesting pair: they're both huge assertions at odds with the general consensus and you didn't support either of them. The usual complaint of web developers is that they're endlessly drawn to new things, so it seems like some evidence would be in order to claim they're suddenly extremely conservative in this one area. Similarly, since Bitcoin proponents like to talk about about simple and easy to use it is, it seems implausible to blame everything on a lack of skill.
The comparison to P2P is interesting: the technology was rapidly adopted in a few areas where it offered a clear win but, similarly, had various people pitching it as the solution to all problems and claiming that skeptics were ignorant luddites. Since those products all failed, perhaps it would be better to work on answering those questions rather than dismissing them?
> If that occurs and you constructed it correctly, your company may well be able to skip on taxes as well, as a side-effect.
This kind of argument is not going to change the reputation Bitcoin has for being most of interest to crooks.
I'm referring to a new way of constructing software, using a transactional database at the core and defining every event as a transaction between two parties. Today engineers usually think of the platform as a distribution medium for data, aka. client-server. In the BC world refining this so that any party can take on any role, aka. decentralized/distributed is challenging. I'm trying to explain that this go far beyond a new JS framework or (no-)sql database, in the BC environment the engineer must be able to define and code business rules.
> Since those [p2p] products all failed
I don't think p2p failed because of technology, but rather because mgmt/industry wanted to keep control. See e.g Spotify or Skype as examples, there are many other interests at play than pure technological.
> tax evasion (value added tax and in some case corp. tax)
I agree with your view, but if you examine today's platform economies this is exactly what they have done to a large extent. They give away "free" services only to cash in on your personal data from a near tax free country setup. Examples are FB, Google, Apple, Amazon..
1. There is a lot of hype, sure, but there was around that whole Internet thing too.
2. The idea that geeks aren't working on it is patently absurd, and
3. I'm not sure how OP missed the jobs-to-be-done of financial transactions, proof of identity, proof of ownership, etc. Those are all pretty important jobs in trillion dollar industries.
How does blockchain perform better at sub critical mass?
In contrast, BitCoin offers nothing other that risk and uncertainty vs. using a credit card to buy online unless you're hoping to avoid law enforcement.
Frictionless (well, one day) payments.
Similarly, what does “frictionless” even mean? Consumers have very little friction currently and the situation is decent for most vendors. How much better can a new system be, especially one with significant operating costs?
Agree that many usecases would be better served by a centralised system, but many don't seem to grasp how useful an immutable database is.
Even with the advent of quantum computing and the havoc that Shor's algorithm will cause, there still won't be enough computing power to undo the Bitcoin blockchain.
(there's no need to do proof of work to track and publish a hash tree)
- Making darknet drug markets possible
- Allowed permissionless, highly convenient storage of wealth
- Allowed large, rapid, low friction money transfers anywhere in the world
Now, in 2017, Ethereum has a market cap of 8 billion, yet I can't find a single use case for this technology other than enabling other scammy ponzi-like coin offerings.
Distributed computing rewards - maybe for a subgroup off applications (eg. illegal stuff eg. piracy), but in general terms a centralised system is easier, cheaper and quicker to develop
Replacing national currencies - this is already bitcoin's killer app. Ethereum adds almost no utility to this space.
Looking forward to Mimblewimble.
cf. "the web", "ecommerce", "p2p", "social media", et c.
it's a tool, and a damn useful one, that enables some things no tools did before. that's it, and that's all of it.
I'm writing a book on this. (Nearly finished! Probably!) I have read thousands of pages of the worst PDF whitepapers you can imagine. I have looked all over to see if there's a "there" there.
As far as I can tell:
1. The only thing blockchain does in practice that anything else can't is decentralised cryptocurrencies.
2. The only thing decentralised cryptocurrencies do in practice that ActualMoney doesn't is markets for illicit goods and services.
note the "in practice" - it's had eight years of "possibilities", by now I think we can look at what it's done.
also, 3. the decentralisation isn't permanent - Proof of Work benefits from economies of scale, meaning mining recentralises, and this is exactly what happened from the FPGA era on.
The usual business proposition for Blockchain(tm) is supply chain management. Blockchain doesn't solve any of the actual problems (bad data - it turns out that crooks don't put careful documentation of violations into the data, and bribe local inspectors and adjudicators) but does offer a monopoly position to the company defining the blockchain in question.
Bitcoin and cryptotech are loaded with geeks, just not the same ones as the author associate with. And cryptotech do a lot of stuff way better than existing tech. You have to be totally blind not to see it, IMO. :D
It was the same with "noone will need a personal computer" laughed by older-school engineers, or "The Internet is a funny toy with its funny packet switching, but real business-critical communication needs a reliable networking like a connection-based phone system".
Cryptotech (not only blockchain-based) is disruptive, groundbreaking and as transformative as personal computers were. But it took like 30 years to get personal computing to the point where it was obvious. And it won't be much faster this time. There will be blockchains, technologies, standards, businesses going up and down, people getting rich and/or broke, with winners shaping a new world.
Gotta quibble with this one. 30 years after the PET shipped we were solidly in the middle of the Web 2.0 revolution.
Yes, ideas take time to mature. No, the PC in't a good reference. By 1981 it was clear where we were going.
Blockchains are useful for reducing the overhead due to lack of trust when you're working with other companies. You can't really make that happen without management on both sides getting involved. There's no way a back office geek working on his own can make complicated auditing and settlement procedures go away, and that's the sort of thing where blockchains help.
(And of course not every new technology gets adopted first by corporate geeks; iPhone apps are one example of consumers taking the lead.)
Do you see consumers commonly paying in BTC? Real offline shops accepting it? (Never mind tiny things like Ethereum.)
Can you expand on what you're defining this problem as and how Bitcoin reduces costs relative to the traditional approaches with trusted third parties, PKI, etc. I've seen the marketing pitches but they're always conspicuously lacking the kind of specific “we pay X to do Y but could pay Z” pitches which I've seen used to justify bringing new tech inside.
Interbank settlement is one example. Right now it involves auditing process that take days. A blockchain implementation could reduce it to seconds. Insurance and supply chains are a couple other things they mentioned.
Pretty much any time you're either doing a lot of mutual auditing, or paying someone a lot to maintain a trusted transactional database for a group of companies, there's an opportunity to reduce costs and/or speed things up with a blockchain.
For now, that probably doesn't mean a public blockchain, due to privacy and scalability issues. The companies in the EEA are focusing on private versions for now; if you have a known set of nodes and trust that no more than a third of them will commit fraud, you can get away with much faster consensus mechanisms. At the EEA conference they said if Ethereum succeeds in its scaling and privacy efforts, they hope to move at least some applications to the public chain, but that's a couple years off.
That's an "is" statement, not a "could" statement. Please show me, not a possibility, not a press release about a pilot scheme, but a present-day working example of this in practice as part of an industry.
1. There are not only brilliant people leading these projects (Vitalik on Ethereum in particular), but blockchain is a huge passionate discussion amongst the SWE communities I find myself in.
> for example the laughably slow transactions-per-second of most real-world blockchain implementations.
2. Is he/she just talking about Bitcoin? Apparently one means 'most'. There are plenty of plans around scaling in various tech: Raiden for Ethereum is an interesting example for smaller transactions.
> I could maybe get past the socio-political issues, the misguided notion that in civilized countries, you can route around the legal system with “smart contracts” > I still don’t think the world needs it.
3. Doesn't go to support their arguments at all. A simple and obvious use case here is: banks dragged down by capital obligations from Dodd-Frank regulation are looking everywhere for cost savings.
Blockchain is interesting to everyone because there are fairly easy to get involved in financial instruments associated that have been doing very well lately. With that, there are plenty of nerds investing a lot of time into technologies to make these networks truly useful.
Name a single bank who is using blockchain in an essential component of their operations (not just a proof of concept). Go on, I'll wait.
Sure, blockchain is "interesting" and the people involved are "brilliant" but where is it delivering a 10x improvement over the status quo?
Blockchain is a classic example of a solution looking for a problem.
What a ridiculous argument. With blockchain really just becoming interesting over the last 2 years, you think banks, many of which use systems 20-30 years old (bonds are still primarily traded over phone) are going to switch over to it immediately. Give me a break.
> Sure, blockchain is "interesting" and the people involved are "brilliant" but where is it delivering a 10x improvement over the status quo?
International bank transfers can take days, have large transaction fees, and have a non-trivial amount of book keeping error. Whenever those same transactions could take 30 minutes with no human oversight I'd say the speedup is much higher than 10x. With that, it seems like you're more interested in making claims without being properly educated on what you're talking about.
Whatever problems you may have with international bank transfers, they work: every business in the world relies on them. For all the supposed benefits of blockchains, no blockchain solution can make that claim, or any claim close to it.
We could make a million of such transactions per minute using $1000/month hosted server. We have the technology.
The problem is legacy infrastructure and the fact that banks benefit from high fees on international transfers.
There is at least one successful business (transferwise) that makes international transfers much cheaper and faster using terribly inefficient work-arounds.
If banks were at all motivated to fix the issue, they would, using Java and a database. They don't need blockchain (which, btw, is slower to settle a transaction than updating a record in a database).
It is not like the problem hasn't been attacked many times from different angles.
But I agree with you that "Blockchain is a classic example of a solution looking for a problem" assuming you're talking about the hijacked usage of "blockchain".
> Name a single bank who is using blockchain in an essential component of their operations (not just a proof of concept). Go on, I'll wait.
I think this is exactly the point. Banks trying to employ their own blockchain is the same as companies trying to employ "Intranet" when the Internet first came out. It won't work.
I'd say that bitcoin has already succeeded, just not for the use cases that the author cares about.
It's not a matter of caring. He points out that whatever uses blockchain has (or hasn't) proven successful for, don't amount to a significant economic (or social) contribution. He's talking about objective quantity, not subjective quality.
He's not disputing that smart people work on it, or that there aren't variations on Bitcoin, and so pointing those things out doesn't respond to his article.
You can disagree with him as you can disagree with anything, but if you want to support the argument that it's "awful", you'll need to do better.
What we actually see: 370 BTC a week on LocalBitcoins, which is negligible in the context of the country.
The closest to a verifiable claim: that miners who were stealing electricity were arrested. Of course, stealing the electricity has long been a favoured Bitcoin mining strategy.
Maybe the OP could have written a post in 1990, 17 years after the first handheld mobile phone prototype, about the tech "still having not reached mainstream consumers, therefore being a failure."
The OP seems unaware how fast Bitcoin is growing. Not mainstream yet, but it's growing. https://en.m.wikipedia.org/wiki/History_of_bitcoin#Growth
Bitcoin is no exception. It may survive, but the story about mobile phones (and any tech that succeeded) is irrelevant here. We can have this discussion when and if it is a consumer success.
Similarly, where is that short-term growth coming from and how would you distinguish it from a speculative bubble? Lots of companies reported amazing growth from 1995-2000, too, but anyone who looked at their fundamentals couldn't justify it.
Immutability causes bitcoin to have no counter party risk, just like gold. It has limited supply, similar to gold. Along with all the other money-like properties of gold.
Yet it is far more portable and accessible.
Their best properties and the banks' needs are not aligned. When you add in stuff the banks need like forcible retraction of transactions, and transaction fees, you have essentially destroyed blockchains. The moment a major bank releases production ready blockchain of their own it will be such that most will not get any real value out of it.
What they are doing (not what they are saying) is that they are trying to figure out how to hug the blockchain technology to death.
That's probably giving them too much credit for awareness of what they're trying to do. What I suspect they're doing is trying to get a handle on the hype about the glories of immutable, cryptographically secure storage technologies promised by blockchain, having conversations with blockchain startups consisting mostly of mutual misunderstanding and then hiring IBM or Accenture because they say "blockchain" a couple of times at the beginning of their presentation and then spend the rest of their time talking the bank's language and proposing to solve problems they already know banks actually have (not using a blockchain) in ways which don't create bigger theoretical problems.
[yeah, the people who definitely are doing the hugging to death are the mainstream IT consultancies that know that "blockchain" is the new "cloud" for sales hype to get people to rebuild a system that was only put in a couple of years ago. But I bet some of their middle management believes they're actually embracing the technology too]
S/he never said all implementations are slow, just that most are (which is true if you count all those toy implementations done by hackers who did it just for the learning experience; I alone contributed two).
Your "obvious use case" is not obvious to me. Nor are the foundations of your argument, that oversight and regulation drags viable banks down so much they need blockchain support to "save money". As you pointed out later in this thread, they are decades behind anyway.
OP's point is that s/he does not see the same traction other (failed) technologies had. I agree. You disagree. So we agree to disagree. But that does make the OP's opinion or your arguments any more/less valid.
This is precisely what I see happening with bitcoin. Bitcoin addresses keep popping up more frequently every day (with WannaCrypt being one recent and unfortunate example). Despite the scalability issues and regulatory uncertainty there isn't an easier way to send value around the globe. And this isn't simply because banks face a huge regulatory burden. The settlement systems in place are very complex and archaic and blockchains promise a fresh start for a global settlement layer.
---
I looked into blockchain technologies carefully and I’ve ended up thinking it’s an overpromoted niche sideshow.
I’ve seen wave after wave of landscape-shifting technology sweep through the IT space: Personal computers, Unix, C, the Internet and Web, Java, REST, mobile, public cloud. And without exception, I observed that they were initially loaded in the back door by geeks, without asking permission, because they got shit done and helped people with their jobs.
That’s not happening with blockchain. Not in the slightest. Which is why I don’t believe in it.
That said, I know a lot of people using the tech and they're making money with it, solving real technical problems. I suspect maybe Bray doesn't know about it because it's easier to assume that it isn't than to verify it so.
There are specific industries where block chain limitations don't matter, and its guarantees do.
Personally I think both the hype and the anti-hype exist because we finally gave this family of data structures a snappy name.
Also, I found the BG solution he was referring to and it's rather soft.
Bitcoin is growing because there are enough people on the world who are willing to risk it. They don't trust Bitcoin, they are gambling.
What is this referring to? In all of the ways that I've seen the Byzantine generals problem phrased, it's presented as impossible to "solve".
But I have no idea what he's talking about when he says "And without exception, I observed that they were initially loaded in the back door by geeks, without asking permission, because they got shit done and helped people with their jobs."
Is he saying Bitcoin is not being developed by geeks?
Is he saying Bitcoin is not "helping people get shit done and helping people with their jobs"?
There are valid criticisms, and there are really baseless rants like this. Just because you're "old" doesn't mean you know the future from your past experience. Tons of talented old people (and younger people too) with tons of experience never realized the potential of the Internet when it first came out.
Oh, and when Internet first came out, it never was meant to "get shit done and help people with their jobs", for a long time. It was a fringe culture. If he's as experienced as he claims to be, he should know better.
It's not like I'm working on some bitcoin business or anything but I am offended that you think Bitcoin is the same type of hype as "AI, IoT, etc.". Trust me I'm the biggest hater of the AI hype in 2016-2017 (driven by VCs and BigCos for their own interest). The reason I know these are hypes is because most technical people who actually have been working in these fields know exactly what's possible and what's not, but the non-technical VCs and marketers are driving the "market" (that doesn't exist) thinking that something groundbreaking has suddenly happened. The fact is: nothing groundbreaking has happened. The technology is getting better for sure, but compared to bitcoin these are nothing new.
Bitcoin is actually something new and a novel piece of technology, which is different from all the other marketing hypewords you compared it with. At this point I think the chances are 50/50 that it will succeed, but really nobody knows what will happen. Anyone claiming to know the future is being arrogant.
I don't follow coins closely anymore but the main one I saw was namecoin, which hasn't yet seen much use.
There are many companies now where you can get your salary paid in Bitcoin. In fact, it is so infinitely more convenient, I think that only the forces of mental inertia hold this approach from immediate universal adoption.
Blockchain is at place right now where managers and marketers are asking "What can blockchain do for us". But where are the examples of "Big company X implemented blockchain to solve problem Y?" maybe it's too soon for those.
But these "top down" approaches never work. That's how it played out with Internet vs Intranet. When the Internet was first gaining traction all the bigcos thought the money was in Intranet because that was more "monetizable". MS went down that route and never recovered.
I also used to be skeptical about all these VCs jumping into Bitcoin ecosystem, but now it's 2017 and if you look around, they kind of gave up on investing in "companies", and are instead trying to buy up the currencies themselves, which means everyone's on level playing ground, which is exactly what happened with the Internet.
I am more annoyed at these marketer types who jump into Bitcoin than anyone, but I don't think these people will end up winning because they have no substance.