I got removed from a project because I was raining on everyone's blockchain fantasy. Later they begrudgingly admitted that blockchain wasn't going to solve the hard interoperability problem we were facing (bog standard need to pull & analyze data from multiple incompatible data stores held by multiple independent parties).

I can forgive all the business guys for that thinking given articles like these, but the tech guys who were fanning the fires of that fantasy and telling me I didn't understand blockchain....

I got the distinct impression that there is a growing cottage industry of tech consultants and outsource companies that realize that if they can get their clients to "do it on the blockchain" they will be able to bill 10x as much for a project that has a current very good solution (and relatively fast to implement solution). And for which blockchain won't be a benefit. But the client may very well be able to get more VC money, so it's starting to turn blockchain 'skeptics' into martyrs.

And the most frustrating thing is that the innovation in Satoshi's paper (bringing together in a unique way prior known solutions) is really exciting and there are very good uses for blockchain. The international shipping companies using blockchain to track shipping container provenance strikes me as really, really perfect.

The bad uses of blockchain is going to kill the tech before it's had time to mature naturally and find its place.

Thank you for being a voice of reason in a domain that has increasingly become overrun by frauds and zealots.

As someone who has been involved in blockchain research since 2011, it’s sad to see a promising technology (in limited use cases) turn into a buzzword thrown around by the many speculators and scam artists peddling vaporware and pumping crypto prices.

Yup: it's the Big Data hype train of 2013/2014 re-liveried and stopping soon at a station near you.

According to Gartner, blockchain has just passed the "Peak of Inflated Expectations". Next phase is "Trough of Disillusionnment".

http://www.gartner.com/smarterwithgartner/top-trends-in-the-...

This one is funny. While cryptocurrency gets most of the brickbats around here on HN, I think Deep Learning is also in the same boat.

Executives want to apply Deep Learning to everything and anything they can get their hands on. The results most of the time are underwhelming and could be had with a much simpler set of tools. Just wondering how long before Deep Learning goes into trough of disillusionment.

On the job I have to constantly shoot down big data / data mining / intelligent nn analysis crapware showeled to my bosses, and 99% of the time the answer is “you got NO data to speak of”

One day a consultant or another yesman will be able to check in and syphon precious resources with their pipe dream and I’ll watch six figures fly out of the window while my dev team struggles debugging simulated ipads with mac minis.

> One day a consultant or another yesman will be able to check in and syphon precious resources with their pipe dream and I’ll watch six figures fly out of the window while my dev team struggles debugging simulated ipads with mac minis.

Six figures? Oh sweet summer child. :)

First you'll need to burn seven figures to have a Big 3 advise your leadership that you need this tech. Then you burn the six figures implementing it with a lower tier firm.

We're in an industry ten years behind so we're just getting into Big Data. Obviously the issue that we can't write working normal software hasn't been addressed but we'll see how it goes. I'm actually relatively optimistic.

Take a step back and look at your work conditions. Maybe you should become the consultant in that story.

No thanks will not be able to sleep at night doing that line of work (not consultancy in general, just the overselling and running with the money bit)

It feels like lately bots and AI are the biggest hype-train. Everybody wants bots. Bots for this, and bots for that. With natural language processing bolted on somehow.

Got a simple, defined flow that could just be a decision tree? Nope, we need natural language processing in a chatbot, we can't possibly present a menu.

I feel like I'm fighting the Infocom parser again.

brick·bat ˈbrikˌbat noun plural noun: brickbats a piece of brick, typically when used as a weapon. a remark or comment which is highly critical and typically insulting. "the plaudits were beginning to outnumber the brickbats"

My WotD

Mine too: I'd never heard it before and had to look it up. Now I'm just lying in wait for an opportunity to use it.

Very interesting chart.

Indeed. Google "Gartner Hype Cycle" followed by the year to get a sense of the evolution of the different hype cycles.

I still remember my manager getting very excited about XBRL in 2007.

Taking over the VR hype train on the way, passing by the wreck of the 3D TV hype train, and competing with the AI hype train.

>it’s sad to see a promising technology (in limited use cases) turn into a buzzword

The same thing happened to REST. Every "web interface" is supposedly REST now, yet the client and server are even more coupled than they would have been if the interface were implemented in SOAP!

And here I thought I was the lone voice in the wilderness. From Roy Fielding's original paper:

The client-stateless-server style derives from client-server with the additional constraint that no session state is allowed on the server component. Each request from client to server must contain all of the information necessary to understand the request, and cannot take advantage of any stored context on the server. Session state is kept entirely on the client

I've heard enough verbal gymnastics to fill a book trying to excuse how modern web apps aren't violating this condition.

Not alone. I find the problem most people have is no suitable example they can think of. The examples we have online are "starbucks as REST", etc. A better example is the greatest REST client ever made: the web browser. When Facebook updates their "app", I don't even need to restart my "client" to get the updated features. These days I start all "REST: you're doing it wrong" conversations with this.

Yes REST tends to mean "web API you don't hate" or "web API written less than 5 years ago" and is heading towards just meaning "web API".

It's an industry cycle. 15 years ago it was "XML all the things!"

The tech isn't going to die, it's just not going to change the world in any sexy way.

Maersk is using it for something as boring as secure claims for containers to fight corruption in shipping.

In government we're looking into running things like the land ownership registry and other public records on the tech. Not because it's hip, but because it'll save employee resources when we have to perform less audits, and because it'll speed response times.

And to the horror of most crypto-currencies enthusiasts, blockchain will likely end up being used as a tool just like SQL servers and distributed clusters, by large companies and governments instead of being the instrument of challenge to these established entities.

This progression already seems to be occurring. Recently the CEO of JP Morgan even derided Bitcoin as a scam, while JP Morgan promotes its own internally-developed blockchain Quorum.

Now why would a major financial entity want you to think external decentralized technologies are scams, and their in-house solution is fine and legitimate and secure.

Property title, availability of airspace, really any complex system that requires mutual cooperation (or at least awareness) could benefit from uniform distributed ledger technology.

How would blockchains help with property title management?

During the subprime crisis there were reports of loans which had been flipped so many times that the chain of title could not be determined.

A blockchain is needed to solve this problem though.

For a comprehensive answer see Avi Spielman's thesis, in "Blockchain: Digitally Rebuilding the Real Estate Industry"

"The research to date leads to the following conclusions: A blockchain title recording system is the future of title record keeping and would provide immediate benefits over the current title recording system, with additional benefits accruing in the future as blockchain technology grows in acceptance. However, at the moment, these benefits do not yet outweigh the costs and challenges associated with implementing a prototype blockchain title registry system in Davidson County, or elsewhere in the country. That being said, steps can, and should be taken now to lay the foundation for a blockchain system."

dci.mit.edu/assets/papers/spielman_thesis.pdfdci.mit.edu/assets/papers/spielman_thesis.pdf

I can't find an explanation in your excerpt nor in the paper that gives specific details as to how the blockchain is uniquely beneficial for property title management. I see some examples of how certain qualities of the blockchain could be beneficial, (e.g. property owners being identified by a private key), but you don't need a blockchain to get any of those benefits.

Aren't these already solved though?

can you link any property title stuff? do you mean commercial residential or both?

Sorry just seeing this now. Both, actually.

Avi Spielman, in "Blockchain: Digitally Rebuilding the Real Estate Industry" does a terrific job addressing and exploring "aspect(s)- recording property titles- by comparing the benefits and limitations of a blockchain with those of the current record keeping system."

dci.mit.edu/assets/papers/spielman_thesis.pdfdci.mit.edu/assets/papers/spielman_thesis.pdf

Well done for sticking up for reason. I first got involved in Bitcoin in early 2011, but left the whole scene in 2015 because of the dwindling number of serious programmers and innovators and the increasing number of fraudsters and "marketers".

Like you say, there are some incredibly innovative uses for the blockchain (another excellent one is claiming and tracking digital works of art like visual arts and music), but using the blockchain as a simple database replacement is ludicrous.

The tech media (and tech startups) has this bad habit of hyping up technologies before they're ready for prime time.

They did it with 3D printing, they did it with VR, and now they're doing it with blockchain.

They did appending 'Cloud' to everything too.

To be fair, most things are moving to cloud. There is just no hype around it anymore because it is considered standard practice now.

As well as deep learning and AI.

The sceptics in the shipping industry are plenty as well, though. And it has been awfully quite around that pilot. Many wonder if the real challenge is really trust or not just incompatible data formats and plenty of non-digitized data...

> incompatible data formats and plenty of non-digitized data

This is exactly what many with a cursory knowledge seem to think a blockchain will solve. Unfortunately, these are generally challenges people have to talk about and resolve via mutual agreement and compromise; you can't simply throw code at the issue.

Back in 2000 I was on a team of eight, digitising a freight system. (Handling about 50% IIRC of all foreign air freight to and from Sweden.)

For a "parcel" of goods, I counted 17 paper copies of the same shipping form, being sent to and fro. Back then IATA also used a custom text field format instead of a standard XML format I think they use now.

So while this was a long time ago, I too expect shipping to be fraught with lots of weird edge cases.

Nah, those bad uses won't kill the tech, but the bubble will burst. But yes, I also notice a lot of companies that are riding the blockchain train in scenarios where a central server would be fine, and all they'll end up with is a needlessly complicated solution to their problem.

If you don't mind expanding a bit, could you give some examples of areas where you think it has potential - I'm not really clued in enough at the moment to distinguish between real possibilities and snake-oil.

In the long run, countries which have fast tech adoption and strong logistic infrastructure will probably be the ones that will advance the field the fastest. It's exciting technology, but I so far don't see a lot of people using it in a practical way. We may be seeing China as the forerunner of blockchain technology, since there's immediate business advantages for their (enormous) manufacturing and logistics setup.

A lie has short legs: if you dive into private blockchain projects you will realize they are mostly proof of concepts.

This is not to say that ideas behind consensus, security, and smart contracts are not useful but you need to start with a concrete problem instead of adapting a fixed solution.

Having said this, it is interesting to explore a pluggable way to make agreements between parties and follow defined workflows for interoperability. The main barrier with implementing this idea is not technical but political since it is difficult to create agreements about this architecture between companies (e.g. banks).

Thank you for your comment. What you said is the best balance I’ve found of how bitcoin (and in my opinion also AI) is a really really really cool technology with a lot of potential, but also a complete scam.

There's an easy way to not be frustrated: just buy Bitcoin independently from what you're working on. If you understand what that means to own BTC 10 years from now, you know that malinvestments won't get you frustrated.

Most of the future btc value is already inflating the current currency value.

And there’s still the risk of a crash and switch in that long of a timeframe.

I bet there’s more money to be had at cryptocurrency arbitrage

there sure is. but the marketplaces are so slow and fragile it's essentially impossible (aka too indeterministic/risky)

Shipping is a really interesting application I had never really considered! That seems like a perfect fit, like you say.

Is provenance in general a good usecase? e.g. chain of custody for evidence / drug testing / food labelling?

Isn't that main goal of Chain Link?

https://link.smartcontract.com

Still off of that project? Or did they actually admit it to you?

Blockchains are the future, aka merkel trees are the future, aka git is the future. Put the world in source control. Distributed, trustless, decentralised source control.

Facilitating inter-bank transfers is one promising application that's already operational: https://ripple.com/

Are there any actual benefits to this over current methods of interbank transfers?

Have you ever submitted a payment or received a payment and seen it marked as "pending". That's likely the result of a delay due to interbank transfers. Real time settlement is something theoretically possible with central banking using something like Fedwire, but in practice most banks use a third party clearinghouse like CHIPS. Additionally, if you have funds in USD and you need to make a payment in EUR, you'll probably need to work through a third party bank that deals with forex settlement. This all adds delays and costs.

With Ripple, realtime clearing and settlement happens automatically even between currencies. If my bank and the vendors bank both use ripple, I can use USD to pay the vendor EUR in real time, and the transaction completes in seconds.

Yes, but is there something that Ripple is doing here that could not be done by any non-blockchain distributed system?

The chief value in blockchain seems to be in getting people to agree to do something at all.

> Yes, but is there something that Ripple is doing here that could not be done by any non-blockchain distributed system?

Yes... as far as I'm aware, most (all?) other non-blockchain distributed ledgers require complete trust between parties. Its hard enough for two banks to trust each other completely, let alone all banks to trust all other banks. The old way of doing things is to either trust a central bank completely, which is easier to do, but far from distributed, or to trust a third party with a days worth of transfers and settling at the end of the day. Still not distributed, but requires less trust.

> The chief value in blockchain seems to be in getting people to agree to do something at all.

Exactly!

Blockchain technology is essentially nothing more than a cryptographically secure distributed ledger. There's probably other ways of having a secured distributed ledger without technically being a blockchain, and I'd be interested in hearing about them, but I'm not aware of their existence at the moment. Either way, I'm not sure why there's so much pushback on blockchains?

> require complete trust between parties

Doesn't it just require trust of a single party for settlements (or any number of independent settlements parties)? Which you still sort of require in the ripple model because ripple owns a crap ton of it's own coinage (60% of the pot).

> Doesn't it just require trust of a single party for settlements

If you're talking about two banks... There are thousands of banks out there all with varying levels of trust, and those banks all need to transfer funds between each other on a daily basis. Most banks only have direct relationships with a handful of other banks.

> Which you still sort of require in the ripple model because ripple owns a crap ton of it's own coinage (60% of the pot).

Yes, they own 60% of XRP available, but if you're settling a transfer between USD/EUR then you only need to trust ripple for the length of time for that settlement to happen.

> Yes, they own 60% of XRP available, but if you're settling a transfer between USD/EUR then you only need to trust ripple for the length of time for that settlement to happen.

Which for a bank would be all the damn day because they wouldn't just be doing one transaction every 2 months.

I participated in writing a money transfer service. There's a number of steps, and the actual transfer is just one of them. There can be multiple identity verification checks, and bank authorizations. Each bank has their own internal API, etc. Sometimes their API is just randomly down, if it is some terrible foreign bank.

I know for a fact that Bitcoin's transaction time is not "real time," so why would Ripple's be?

Because bitcoin relies on proof of work in order to determine consensus. Usually vendors require X confirmations, and confirmations happen every 10 minutes of so (every time a block is mined).

Ripple doesn't use proof of work, it relies on a web of trust. So, in order to transfer money from Bank A to Bank B, it relies on Bank A to have some path to Bank B that is trusted. If Bank A and Bank B don't trust each other, but both trust Bank C, then the money is routed through Bank C.

Not sure what you mean by identity verification, if you're talking about for the customers, then they'd still need to have their identity verified at the bank (for purposes of Ripple, a bank is a ripple Gateway, but not all gateways are banks). Honestly, I'm an armchair expert in money transfer systems -- pretty familiar with Ripple, but less familiar with other systems.

I'm always amazed that in this day and age, it takes three business days for my paycheck to direct deposit into my checking account...