Hacker News new | past | comments | ask | show | jobs | submit login
IBM has cut its blockchain team down to almost nothing – sources (coindesk.com)
292 points by walterbell 7 months ago | hide | past | favorite | 331 comments

In my experience - I have done a lot of pre-sales for an IT service company - the use case for blockchain can usually be solved with classical, cheaper and simpler solutions by having a central trusted authority for storage (like a third-party service provider), and appending using digital signatures.

For me personally, blockchain is a solution looking for a problem in most cases...

In other news, water is wet. Obviously!!!! Blockchain has earned a bad name because of sales people trying to sell it for something it can't do. Blockchain is a very very very inefficient database that removes the need for a central trust authority. That's it's only selling point. You don't need it if you are storing your customer data in a central database.

This. How many real use cases are there where you really need to distribute trust? For currency, there's some sense in it.

But when i hear things like using blockchain for contracts or for keeping track of supply chain... it always seems like what people really want is just a well organized and reasonably trustworthy third party who can handle this stuff. Does anyone have any specific examples where blockchain would be a better solution?

Real estate deed ownership database - immutable distributed database spread around regional government offices.

Speaking from my experience, the records are being kept horribly over here, there were many projects to cenralize it but the original data is in such sorry state that it is basically impossible. Not to mention that it is very profitable for some that it is in such sorry state.

They key idea is that you are unable to modify who owned what in the past, only change ownership in the current time, and the records, as per current system, are spread all over the local government entities.

Same thing for financial ledgers - immutable record of transactions to which only banks can write to.

I think that stocks are too volatile, but it is a such shock to me that you can only find out who actually owns a specific stock at T-2 ..

In both cases there is no competition for who solves the next block faster, so costs of computation costs can be kept low. I'm more attached to immutability of such databases, than distributed trust, but one cannot exist without other in blockchain.

>the original data is in such sorry state

How does Blockchain solve that problem?

It's the same issue with using it for supply chain verification, it's dependent upon the point of initial data entry. The Central American farmer enters the tomatoes that he just received via container ship from China into the Blockchain application as "organically grown domestic produce" and from that point forward consumers can rest assured about the tomatoes' authenticity and provenance. Or not.

Right, authority-less blockchain is only intrinsically trustworthy for purely digital goods. If you want to track software licenses it might work (although 99% of the time a centralized solution will be simpler and vastly more efficient).

And this immutability can arguably be a detriment. Laws and rules are messy and can be exploited, not having any recourse if you get scammed is not necessarily a huge pro.

The whole idea that you could import all the rules, laws and regulations of the real world into a computer program without any bug or unforeseen issue is probably one of the most ridiculous aspect of that whole blockchain industry. Especially since we already have a fairly consequential list of "smart contracts" that have been exploited in the wild despite being fairly basic compared to IRL regulations.

"Oh you lost your house because somebody found a weird edge case between two smart contracts that let them buy it for $1? Ah man, that sucks, but then THE CODE IS LAW so... Tough luck. Maybe try buying DOGE coins? I hear it's gonna moon soon. They used rocket emojis, so you know it's serious."

> Laws and rules are messy and can be exploited, not having any recourse if you get scammed is not necessarily a huge pro.

Yes, many times things like T+2 settlements on trades or taking three days for a transfer to fully clear are a feature, not a bug.

You start with Bitcoin being like cash, where once it's gone it's gone, unless a court can find the person and get it back. Then you don't want grandma to get scammed, so you build bank-like entities on top of the blockchain who re-create the three day window for transfers. In theory you could have different entities that have different settlement times as a tradeoff, but the modern system could do that now. The reason they don't is you need both parties to have the same rules or else your bank will get bogged down trying to figure out the rules of every other bank. Maybe you add an instant, non-revocable transfer like a SWIFT equivalent to go along with your 3-day ACH equivalent and, congrats, you've recreated the modern banking system.

Immutability of past transactions is just an artifact of an append-only database. Blockchain gives you nothing that an append-only database couldn't provide for that guarantee. Combine that with some server sharding, and there's not really a point to having blockchain do anything at all. The wholly designed inefficiency of "mining coins" is not necessary to having a distributed ledger.

You should have heard IBM's blockchain executive (I forget his name) giving his pitch at an investor conference in Palo Alto back in 2017. Standing ovation after an amazing presentation about disrupting banking, healthcare, and global logistics. I swear this industry runs on bullshit.

In other situations, the initial data entry depends on a trusted authority rather than a trusted other party. For example, real estate ownership deeds depend on qualified legals who write them correctly: you need to trust them to do their job, without mistakes, collusions, etc. at which point there's little practical difference between buyers and sellers trusting them to keep a public, official record with paper and electronic documents in their archives and trusting them to enter your transfer into a public blockchain system.

I don't think it would work for physically settled commodity futures because contract specifications require a particular quality of product, and that's part of the motivation to trade via an exchange.

But it would work for securities that have no distinguishable attributes from one another, especially if the extant problem with trading those securities is market access.

It doesn't work. But, I clearly remember this being something IBM advertised extensively.

It actually solves it quite well.

It enforces you to input it properly, and not do half assed job due to immutability, because whatever you input in there will stay there, attached with credentials/metadata of whoever entered it.

Currently the main problem is that docs are in sorry state, it was a rush job - and that state is very profitable for some people making dubious claims to old property.

It would be easy to notice fraud, and all real estate deals need to go through proper notary anyways - would they risk their license if their failures would be visible - for ever?

Mind you I'm not trying to propose a global solution, I'm proposing blockchain use for government entities and entities that work closely with them.

It is just that governments aren't centralized on modern world, and you can leverage decentralized nature of blockchain to have a common dataset and standard available to all of those entities.

It doesn't enforce anything. People can just transcribe wrong things into blockchains, then have to append patches to the information. Unless you can prove malfeasance nobody is going to lose a license over a mistake. I think you're vastly overestimating how much anyone would care just because blockchains are involved. Notary records are already "forever," no computers required, and there's nothing inherent to blockchains that causes people to type more carefully.

Plus there are lots of situations in real estate law (like ground rent where the freeholder is not identifiable) which don't fit into cute little data structures. This isn't a problem with documentation being a rush job, it's a problem with some of the records of ownership being hundreds of years old. Seeing as how I would have to hire a special archivist just to access data on an 8" floppy disk, I'm not confident that the blockchain is a long-term (i.e. centuries) solution to any of these problems.

> It enforces you to input it properly, and not do half assed job due to immutability, because whatever you input in there will stay there, attached with credentials/metadata of whoever entered it.

This is an unrealistic level of infallibility to ask of people. Like taking the backspace key off a keyboard to enforce people not making typos.

>all real estate deals need to go through proper notary anyways

I don't know if you are American or familiar with how real estate works here. It is decentralized in a way that most countries are not, and there is insurance that people buy against mistakes and problems with transferring property. In a way, you could argue that a single blockchain to track real property would be more centralized than the current system.

"the vast majority of U.S. states have opted for a system of document recording in which no governmental official makes any determination of who owns the title or whether the instruments transferring it are valid"


"A recording system combined with title insurance decentralizes records, creating redundancy. For example, when many records were destroyed in San Francisco's 1906 earthquake, out-of-town title companies maintained records that allowed landowners to prove ownership of their property."

As i said - over here, not in USA. we use completely different system.

There is a central record for each separate local government(located at the regional judicial office), with copies being sent to up the chain.

But the thing is - over here whatever is in that local deed book is the current legal state of the property.

Licensed notary is obliged by law to check the current legal state of the property, verify identity of the ones making the change, and do the change itself - but due to sorry state of the books first part isn't always reliable.

I am also not proposing a solution for USA problems, but for my country.

I'm not sure which country you're describing, but Andhra Pradesh in India had signed an MoU with the Etherium Foundation for property records. I still don't see the Blockchain could have resolved the mess that is property records in India. How would it have worked? Would each property owner hold the private keys verifying their ownership, and possession of the keys proves ownership? My friend in Andhra Pradesh discovered that a portion of the family's land had been sold by a relative, and each party had "legal" paperwork to show for it. All you need to do is pay off the right guy in the local office to smooth things along. I can see the value of a distributed ledger for something like this, but it's an all or nothing proposition.

Yes, but also the US system is not necessarily a problem if you like decentralization.

People say a blockchain is decentralized, but maybe it's more accurate to say it's a centralized system without anyone being in control. Which might in fact not be preferred to an actual decentralized system.

And doesn't allow for changes to the data model, right?

Quite frankly yes.

Because in most cases, speaking from experience, after model change they will be using a new model without doing proper conversion of old data. This happened to cadastral data over here - new schema required some fields explicitly, while old data did not have them digitized, and country-wide system was made with such assumptions. It does not work to this day properly.

Archive the old chain, deploy the new one with migrated latest state from the old chain transformed to new model. keep working.

And deployment requires a central authority to define.

Inherently, this cuts most use cases off at the knees. Some small country your route your shipment through changes requirements? Broken.

This wasn't about shipments, but about either transaction log, or property deed ownership system controlled by government.

> you are unable to modify who owned what in the past

So we throw out the entire body of law around fraudulent conveyance?

Yes. Many proponents of blockchain speak with glee at how doing so will "cut lawyers out of the process". Lol. In every generation there is someone who thinks they can do away with the legal profession. Lawyers have walked the earth for thousands of years and shall continue to do so for many more.

This very much like the back and forth we have on microservices vs monolith. Every generation likes the cool thing only to revert back to mean.

People will go decentralized, then centralized, then decentralized, etc.

We built centralized government bodies for a reason.

The aspect of blockchain + bitcoin that's pretty appealing is limited supply & formulated inflation & cryptographic guarantees.

Applying this to all other domains is stretching it a bit.

In local laws over here only the current state of deed book matters.

Whatever gets there, input by trusted and certified notary, IS the current legal standard of law. No exceptions.

If legal proceeding happens that wants to restore ownership, a new state is written into that book - it does not return to the previous state.

if fraud happens you can clearly see who, when did it, and can do another change restoring it to proper current state.

While also recording who fixed that mistake.

> In local laws over here

Pet peeve: posts in a global forum that mention "here". Which one did you mean?

Eastern european country, as i'm proposing a solution for problems that happen over here.

We, thankfully, do not operate under Common Law.

>Real estate deed ownership database

Real estate broker and former software developer here... I recall having this conversation with an investor type about three years ago. We could find no scenario where Blockchain was a meaningful improvement over a generic database. Same goes for every asset management scenario I've discussed with similar people over the years. This is not to say there aren't tons of opportunities to fix many of these real-world problems, just that Blockchain doesn't appear to add anything but does add to the complexity of the solution.

It’s only immutable if you get the whole world burning energy 24/7 to avoid a 51% attack. That solution is a little bit unreasonable, isn’t it?

Apparently there has already been 51% attacks: https://btcmanager.com/privacy-coin-firo-51-attack/

If you read closely i'm proposing government use only, where each governing body participating in the system has access to it.

You can keep compute systems down - i'm more interested in immutability and distributed nature than the trust issue of blockchain.

Then you don't need blockchain. Once you have trust among the participants, you don't need blockchain.

as i said - I'm more interested in a immutability of the whole chain, without any backdoors to modify old data because - speaking form experience - that leads to way more problems when dealing with local governments.

And the point myself and others are making is that you can get that without using blockchain. Blockchain is a specific technology that requires proof-of-work among distrusted actors for modifying a cryptographically secure distributed ledger. It sounds like you just want a cryptographically secure distributed ledger.

Isn't that why some coins are moving to proof of stake instead of proof of work? Solves the energy crisis. Idk if proof of stake prevents 51% attacks though.

Proof of stake does not offer any resistance to Cybil attack or (part time) collusion of stake holders.

In the case of such attacks the victims will find themselves in a situation that can be described as "shadow ban" - their transactions are not accepted or accepted too late, etc.

Proof of stake doesn't prevent 51% attacks. It eliminates the incentive. If an attacker controls 51% of coins, why would they attack the network? That would be attacking their own wealth.

Not if they can attack the network without anybody knowing. If an attacker secretly controls 51% of the coins (by, say, having three separate entities that each control 15-20%) and can manipulate the market in their favor by rejecting transactions that would hurt them, that's a big incentive.

seems like with state actors owning more and more of it, as well as the mining hardware that they could pull this off at some point in the future. Namely China right now.

There's no valid reason to claim that stocks in general are too volatile. Or you could equally say that stocks aren't volatile enough. What's the optimal level of volatility?

Reducing stock settlement times below 2 days wouldn't impact volatility. Those are orthogonal issues. And current blockchain technology isn't scalable enough to handle even a fraction of stock transactions.

I'm talking about how fast the ownership of stocks change, not the volatility of a specific stock - hence it was about scalability only.

that's not a blockchain though. That's just a linked list or (low tx/sec) timeseries database with a hash.

A block-chain deed database would be completely useless.

The deeds that matter are the ones stored by the people who will send around men with guns to kick you out if your name isn't on the deed that they have.

If you were rebuilding government from scratch as a decentralized entity, then sure. But all the living spaces on earth already have a jurisdiction, and they're not looking to disrupt themselves.

> a well organized and reasonably trustworthy third party who can handle this stuff

You hit the nail on the head. In the US we can trust our third parties. Europe too. But that's not always the case elsewhere.

I am reminded of a friend who told me he wanted to sell some stock held in a bank in his country in Latin America, but the bank refused to process his transaction for a few weeks. By the time he was finally able to sell, the price had fallen greatly. It later turned out that the owner of the bank was selling his own shares during that time.

Blockchain is good for systems where there are few transactions and energy is cheap, but there is no one single trustworthy actor. That's often the case in the third world.

Third world equities exchanges (not places like Brazil and Chile but other countries with very undeveloped financial markets) would benefit from blockchain. There's not much liquidity there and the individual investor ends up getting pushed to the side when the large players want to trade. Even if you have a central clearinghouse and a liquid exchange, corruption is a factor.

I don't pretend to know how best to implement blockchain for electronic exchanges, but I suppose the starting point would be to build a settlement system that has the imprimatur of a coalition of different governments (like Mercosur or Comunidad Andina), hold the shares in trust somewhere, and use the settlement system to process exchange-traded transactions from the constituent countries. Eventually, exchanges don't have to be the origination points of transactions, and you could presumably just get bids and offers directly from other participants over the internet. No guarantees that you are getting the best ones, but that doesn't matter as much for third world markets because the prices are wide and slow-moving anyway.

If you can't trust the government then what are the chances they'll buy into using a blockchain that takes away their corruption powers? The only way you could adopt it is if it's developed in a trusted country, but then you'd be pitching your corrupt government to use a (possibly) hostile countries system to manage property. Or that they will actually use the blockchain to resolve disputes rather than just telling you to piss off?

> If you can't trust the government then what are the chances they'll buy into using a blockchain that takes away their corruption powers?

OK well I can't speak for Asia or Africa but I am familiar with the Americas...

Aside from the fact that governments always look for PR campaigns that make them look fair, they also know they aren't going to be in power forever, and they suffer (get fired, thrown in jail, you name it) when power changes hands. If you can insulate the financial system from those shocks then it's beneficial for both parties because they each protect their financial downside. Also, the country as a whole grows when markets get healthier and can attract more investment. That's good for everyone.

It's not that you can't get the government to set up and use a particular system. It's that individual corporations within the country affiliate with a particular faction/party and try to bend the system to take advantage of other participants. The corruption happens under the skin. You can establish a competent central actor but you can't trust him not to bend the rules, like a horny professor when an attractive student with a B asks for a low A. The third world is one big hackathon for the misuse of institutions. Look at Morales' successful efforts to wriggle out from the grasp of term limits; he went to the Supreme Court and asked his buddies there to declare him eligible due to "human rights." He didn't say "I don't care about the Constitution," even though he wanted to violate its rules. They don't want the institution to look like it's crumbling, even when they're the ones chipping away at it.

If you elevate the exchange to a multinational blockchain then you have competing factions, lots of ebb and flow, and the potential for outside participants. And you get the "i don't have to trust the others" aspect of blockchain.

The whole point of developing robust systems is to make them failure tolerant in the face of perturbations, so we should apply them to the perturbed applications instead of building blockchain for rideshare/dildoes/whatever that helps nobody with an actual problem.

Even if you need to distribute trust, you may just need a system where half a dozen or so players maintain replicated databases with duplicate transaction logs and a way to compare them.

exactly. A database with a running hash is probably good for like 99% of "blockchain" use cases.

If you had a few traders who sit underneath a buttonwood tree and don't trade with anyone else (e.g. the founders of the New York Stock Exchange) it might be OK.

Say you have a consortium of five banks: 5 copies of the database is not a crazy price to pay for a high level of resilience.

Make that 5000 banks and you are paying 1000x as much but not getting 1000x the resilience.

So many people are obsessed with peer-to-peer solutions, I could also see a blockchain as an answer to (say) having a book lending database for a few friends.

You might be interested in https://trustlines.network

Sometimes I don't get the hackernews community. Internet moving from protocols to monopolies is bad. But when we have a cool protocol for ledgers, then no, we want a monopoly.

The problem is that the cool protocol for ledgers doesn't actually address many of the business needs while simultaneously being pitched as if it does.

Very few people are arguing that blockchain has no benefit. But it also has costs and in many cases the costs outweigh the benefits. Tracking the provenance of your tomatoes on a blockchain has real costs over putting the data in a database, and you still need to have someone you trust who can vouch that the tomatoes were grown where they said they were. If you're already trusting someone to say "these tomatoes are authentic, Nebraska-grown tomatoes" then you might as well have them manage the whole supply chain.

Also, I don't think many people would complain about a monopoly that had open data access and open protocols and allowed interoperability. If your tomato-supply-chain-management company told you that you couldn't export the data about your tomatoes or you could only access it via their terrible app then THAT would be the argument for moving to a blockchain. If at any time they could inflate the supply of Nebraska-grown tomatoes by clicking a button then that would be an argument for blockchain.

The need for trust is a necessary but not sufficient requirement for putting something on a blockchain. Too often it's presented as sufficient.

There are already technologies that work similarly to blockchain (such as Git) without needing blockchain to work. You can have distributed, immutable databases without blockchain.

Blockchain's only real advantage is that it eliminates the need for trust. If your system doesn't need to be trustless, there are far more efficient (energy and otherwise) tools that already exist and accomplish everything else that blockchain does.

The supply chain stuff always sounds to me like a scam, and every time I had someone pitch me a blockchain supply chain startup it was indeed a scam (and I told them so). Now maybe there’s an element I’m missing, but there’s still a fundamental problem in linking a real world object to something on the chain. Digital currency works because it is only on the chain, and we manage to agree that it has value, or we use trusted third parties to move that value to the real world in something we can tell is real USD (although the banking system is so opaque that...)

You're not missing anything. I've developed a prototype blockchain for a supply chain before at the behest of a non-technical manager at some point. You hit the nail on the head with the problem of crossing into the real world. You can't validate the real world object mathematically, so if you lie to the blockchain, there's no way to tell.

So far the only use case that I can really find for this are digital identity systems where the blockchain itself serves a double purpose as both a highly redundant system, and a byzantine fault tolerant distributed storage system. I can see the value in being able to maintain a strong ledger of ID uses. The EU looked into this from what I understand, and I think they stood up a system for this kind of thing.

My big TLDR from working on chaincode and stuff like that is this: if your goal is to have a high security digital asset (eg: tokens and ID certs), then a blockchain might be useful. If your goal is to interface with the real world, then you're just going to be chaining yourself to a bulky inelegant solution that won't even address your core problem.

The first time I felt like there was real potential for a blockchain, outside of digital currencies, was with DNS + PKI systems. Pretty much identity for websites. I thought it was a much sexier alternative to the current way DNS works and how certificate authorities work. Now does it scale as well? Does it work well for light clients? And is it the right system when you know people will make mistakes and lose keys?

> it always seems like what people really want is just a well organized and reasonably trustworthy third party who can handle this stuff.

And also a third party to sue when things go south, or to correct errors (at scale, humans always make errors!), or to be subpoenaed to provide information, etc.

These are all advantages in some contexts, and disadvantages in others.

But those specific examples of when blockchain is better should probably also be better because there's no one to sue, no changes to make ever, and an information-hiding advantage.

Property deeds. It is amazing how bad counties are at such a simple charge, I know three people (in three different counties) who had to deal with the local government losing records - one of them only learning this after a neighbor tried to annex a chunk of his land.

> one of them only learning this after a neighbor tried to annex a chunk of his land

Property rights are messy.

How do you create a new category of mineral rights a hundred years into the blockchain?

Also: Rule change! City is changing parcel sizes by narrowing sidewalks. Another rule change! Single-family home neighborhood rezoned to multi-family; two plots are mostly (but not completely) combining to create a multi-family plot with air rights bought from the church next door. Rule change! Everything east of this line of longitude in those two towns (but not unincorported county land as of the 2009 geological survey) is now federal property per eminent domain.

Generally I think the blockchain is dumb, but I think recording property ownership would be a good usecase. Don't think about it as a highly structured database, think about it like a filing cabinet that can be accessed from the internet and anyone can contribute to the maintenance of.

> How do you create a new category of mineral rights a hundred years into the blockchain?

Same way you add a category of mineral rights now - you pass a law and you add a category name. Then you start recording documents that track those rights.

I actually think blockchain would be an interesting model for media ownership. Buying movies or other things online is sketchy because your 'property' basically gets tied up in the fortunes of a particular company. 'Buy' a movie from StreamIt.com and they go under? No more movie. I'd love to see a consortium that agrees to use a blockchain system to issue movie rights - the media companies agrees that anyone who has control of a wallet with a 'moviecoin' minted by them has the right to watch that movie. Even if all the original companies go out of buisness, new companies could keep running the network and charge smaller fees for bandwidth to stream you the movie you already own.

That might be a concern if you try to cram every possible property of the legal code into the blockchain. So maybe don't do that. But if you do, the whole thing is extensible - so it is possible. But don't.

> maybe don't do that

A partial log of property rights is no longer a property rights register. It's reference material. Annotations.

Pack it up Identrust, digitally signed documents and timestamping are is just annotations.

This is actually a problem I have a hard time to understand. In Germany, these records are kept by a government maintained registry. Every patch of land has a number and a record, and an owner. Changes need certification by a Notary. Records go back to, what feels, eternity. I never heard of any problems with the records themselves, problems usually arise regarding the agreements between parties to change said documents. No idea how blockchain would beat simple, court maintained paper.

The US generally uses a different system than a single registry.


That doesn't necessarily mean that a blockchain would be preferable. If one thought the US system is actually better than a registry, then one might consider a blockchain is essentially a fancy central registry.

Is it really that amazing when getting funding for anything takes pulling teeth? Voters have been convinced that investing in government operations is theft.

> Voters

Just the two? Because it has been a while since I heard anyone ask "Who is John Galt?"

OK, but you don’t need blockchain for that, any sort of Merkle tree will work.

You don't need a blockchain but a blockchain can potentially give the public guarantees that an opaque government solution usually does not.

How about a transparent government solution? https://www.registers.service.gov.uk/

The question wasn't "What is the ideal data structure?"

A former coworker has been using blockchain for seafood catch reporting in the Pacific. It’s a complicated legal situation (EEZs, treaties, illegal fishing, monitoring for habitat collapse) made more difficult by expensive/slow internet connectivity.

Storing contracts on the blockchain does seem a bit pointless. Smart contracts, OTOH, would actually enable new kinds of business if they can beade workable. That's a pretty big if, IMO, but in theory at least they seem like a valid use of blockchain.

Govt records, registries of people born, death certificates, deeds on land, swift for banking, any sort of info of public interest like procurement agreements, notary documents, licenses (govt level, like drivers license etc)

The very first thing in your list is literally a central authority.

It certainly should be, and this is no problem in developed democracies. Not so much in the developing world - leaves room for corruption.

and btw look at Estonia - implemented blockchain for most of its govt stuff - including judiciary, and they are among the fastest in the world in terms of time to resolve matters once they reach the courts

ref: https://e-estonia.com/solutions/security-and-safety/e-justic...

Imagine how impossible it becomes fix a wrongful conviction. You could never fully clear that persons name because the “crime” would forever be on the books publicly

True, it’s not something you can apply everywhere. But there are use cases where the downside is virtually nil.

Well yeah, blockchains are inherently doing what governments historically have done but in a safer, more scalable way. Anything that is just a public record that needs to be kept is a slam dunk for blockchains.

What is "safer, more scalable way" about blockchain?

Safe? There are many solutions to maintaining reliable records. Blockchains are one, but no better than many others

Scalable? I cannot understand how it is possible to think that scalability is a attribute of blockchains. A centralised blockchain is no less scalable than many other cryptographic solutions, but a lot more bother. Distributed blockchains are not scalable at all, are they?

On Safe: Smart contracts ensure only the right people have read and write access to the data with no way to cheat. The records are also safe in case of hardware failures so there's no way the data can be lost in an accident.

Scalable: There are a few aspects of scalability. The one you are referring to and what most people think of is number of transactions which is indeed limited right now but for most data is good enough, especially with L2s like Optimism and zero knowledge rollups. The other aspect is social scalability. If you want a system of records to be available to more people then the organization running it is the ceiling for how widespread adoption will be. A birth records system for Peru will never be used by Brazil, the US, or anyone else. For a system to scale larger than a single country then it needs to be credibly neutral where no person or organization controls it and everyone can trust that the system is fair. That's the kind of scalability that smart contracts offer. Worldwide systems that scale to anyone who needs to use them.

As another pointed out: supply chain does make sense.

There are plenty of cases where it's better to not have to trust on any particular entity. Examples:

- Marketing and advertisement campaigns. How much ad fraud exist and how much money is spent by advertisers without knowing if they are actually reaching their customers. Do you trust Google to actually mitigate these problems?

- Health care: remember Google Health and Microsoft Health Vault? Nowadays Apple/Google/Samsung wants to collect that data. Wouldn't you rather have it in a way that only doctors had access to it?

I don't understand the supply chain use case. You can't trust the third party to enter "shipping container #123 has 10,000 bananas" into a centralized database, but you can trust them to enter "shipping container #123 has 10,000 bananas" into a blockchain? They can still lie about what they put into the blockchain, it isn't magic.

Of course they can lie, but with blockchain it is much easier to detect the lie. If the banana producer has a receipt from the shipping company that says "10,000 bananas in the container for Jasper" and the shipping company only has 8,000 to put at your door, you can be sure it who was malicious.

Without a blockchain, if the shipping company and banana producers were colluding to skim you of 2,000 bananas, it's much easier for them to do.

If the shipping company and producer are colluding, blockchain does nothing to stop that.

No, but if the producer has a record saying they put 10 bananas and the shipping company only delivers 8, it will be undoubtedly on the shipping company to bear the costs of the difference.

OK. But what if the producer said they put 10 bananas, but they lied and actually put 8? When it finally gets to the recipient, nobody knows whether the shipping company stole them or whether the producer lied. The blockchain did not help.

Come on, it's not that hard...

I mention about the producer saying they have a receipt indicating 10 bananas. Whoever produced that receipt signed a message saying "I got 10 bananas from the producer" is now on record responsible for it.

So why wouldn't a central database work, with the producer entering how much they sent to the shipping company, and the shipping company entering how much they sent?

What in the world does the use of blockchain enable in this case?

Who controls the "central database"? What if the manager of the central database is colluding with the shipping company, or the producer?

What if the manager of the central database wants to block one of the producers to favor another one?

"What does the blockchain enable in this case?"

To get rid of the central database and all the potential sources of conflict?

Imagine like you trust edges but not middleman.

It doesn't eliminate that. But it gathers evidence. Where did they acquire the 10,000 bananas? Multiple parties must collude -- and then that collusion is on record.

I'm interested in if that data is useful to anyone?

Also, why not trust a central entity with that data?

>Also, why not trust a central entity with that data?

I.E. blockchain does nothing to solve this potential issue

> Wouldn't you rather have it in a way that only doctors had access to it?

How is a blockchain the answer here? The "distribute all data" and "keep history forever" attributes seem actively bad for healthcare use cases.

That is a good question. The answer is that you don't put the data in the blockchain, you just keep the log of who is interacting with the patient and who needs access to the records. ;)

Even which doctor someone visited is confidential data one ideally would not preserve when not needed anymore (and "ideally" might very well be law). Health data is messy, and I have trouble imagining a solution that puts it on a blockchain but allows only private access to everything and has a chance of fixing instances of leaked credentials while still getting value from the fact that there is a blockchain. (If there is prior work on this that actually goes into detail I'd be happy about pointers - sadly a lot of this stuff seems to end at the stage of whitepapers when it comes to public information)

>Even which doctor someone visited is confidential data one ideally would not preserve when not needed anymore.

Indeed, for example "John Smith went to see a HIV specialist on Monday" is protected information and would be a HIPAA violation if leaked. Its much simpler to throw it in a relational database and only let the HIV specialist and staff see the appointment.

https://brave.com/themis/ should be a good pointer about how to get an advertisement network that can avoid fraud and report views accurately while preserving privacy. The same principles could be applied to a permissioned blockchain that could be used by physicians and healthcare providers.

A public immutable and distributed log saying somebody sees various psychology specialists, or disease specialists over the course of a few years... this metadata is almost as personal as the raw data in many cases.

I am almost surprised that companies don't get into more trouble over location tracking. Your smartphone knows exactly when and how often you see a doctor.

All the doctors websites around here, including for specialists, use google fonts and google maps etc. I assume there is some metadata involved there too

I guess HN's pedantism will make me write a whole paper about what was supposed to be a simple example...

Anyway, both the example of advertisement and health care data need to also provide privacy. In both cases the data should be considered sensitive and not public. In both cases, the data can be protected through blind signatures.

[0]: https://en.wikipedia.org/wiki/Blind_signature

History of second hand cars. (Repairs, accidents, maintenance, original options, paint, etc)

Currently, there are various parties that try to collect this info, but it's still full of holes. Plus, it's not a single party, even for the same area (eg EU). It gets more complicated for import/export.

Let's say you can "solve" this by having a single worldwide party that handles it. Do you really want such a monopoly? How much will you pay then to get the data?

Why not solve it with a ledger protocol?

Since there currently is no such world-wide central party, why would a blockchain based solution be successful if the central party is (apparently) not?

In this specific example: are car dealers not participating because they don't trust the central party (case for blockchain) or because (lazy|not in their interest|inconvenient|...)? If the latter, then blockchain is not solving the actual problem.

You have a few parties now that collect this info and sell it. They are pretty successful. But as a buyer, you pay around $50, and the amount of info highly depends on the car and the party collecting it.

Right now, selling a car with an official maintenance book is always a benefit.

The real benefit is of course on the buyer side, who wants to verify what he is buying. So a car with an "official" log would be more valuable than without. When this is the case, it of course is better for the seller to have such a log.

I've been in this space for a short while, and I can definitely see it working.

But if the benefit is to the buyer, what incentive is there for anyone to enter information about the car's history into this blockchain? The person who owns the car isn't going to since it has a chance to reduce the value. And the company that's selling the info wouldn't start putting the info into the blockchain if it meant not being able to sell the data anymore.

An incentive to keep track of a car history is not really under discussion, since it's already happening right now. It's just really cumbersome right now and using very old tech (stamps in a books, etc). It's all made up of bits and pieces in every way.

When you buy a car right now, you expect it to come with at least a maintenance log. So that part is not an unknown.

Everything I mention already exists, it's just that it would be way easier and more complete when it would be handled by a 'standard protocol', ie a blockchain.

The same can be said for supply chain applications. But the second hand car industry has even more trust issues. Selling a car for $8000 and getting around $2000 extra because you can prove it has a proper history is not fantasy.


Money is a lot of things, but one of these things is it is a web of trust.

What a currency needs is a central bank. Historically we have tried a lot of things, and the most stable, reliable, and trustworthy turns out to be fiat currency with a central bank.

The "crypto currency", using blockchains, experiment has been a colossal failure. Trust matters, and it needs a source.

in what ways have crypto currencies been colossal failures?

Transaction capacity is far too low. Bitcoin is essentially unusable when transactions take over a hour to complete

Volatility. Currency must be vaguely stable. Bitcoin is all over the place, most of the others have very low liquidity.

The waste of energy.

Scams galore. What in this space is not a scam? There have been so many scams.


Seems likely there would be good uses for blockchains in decentralized systems. However, we have not built many decentralized systems so far.

Even beyond that, how often do you want to do business with / complete a transaction with someone who you trust THAT little?

I suspect there is a lot more trust involved in every day things than maybe Blockchain assumes.

>Even beyond that, how often do you want to do business with / complete a transaction with someone who you trust THAT little?

Not often precisely because of the risks but if those risks are minimized it can expand your range of potential business partners.

How does a blockchain ensure they ship the materials I requested? Did they harden the metal properly? Were the electronic parts purchased from a reliable supply chain so I don't get reclaimed ones? Did they use the right PSI steel? Did they test to known and documented quality standards, or do they ship me the ones that work, but fail quality checks that they reserve for their big customers?

I have manufactured a lot of electronic and physical products, and have had a lot of supply chain issues, but not one of them would have been solved with a blockchain.

Meanwhile the concept of escrow has exited for millenia and already solves that problem. 99% of the proposed usecases for blockchain can be done much more effectively with tech, infrastructure, and systems that have existed forever.

It's a lot like all the hype over NoSQL, when the principles underlying traditional relational databases have already been a solved problem since the 70's and work for 99% of use cases.

I'm not sure there would be a desire to do so. I think that often trust to some extent is a baseline requirement for most people to do business in the first place.

You are getting it backwards.

Do you know always who is on the other side of a purchase that you do online? I can bet you don't, yet you still do business because you trust the intermediary (credit card provider, Amazon, whoever) to ensure that both parties are getting what they want.

Blockchain and "trustless" systems are there for people who want to make business but do not have these intermediaries, or when the intermediaries themselves are not trustworthy (corrupt governments, inefficient companies, etc).

It is fine to buy a few things from an untrusted supplier off Amazon. It is a whole other issue to run production based on unreliable or untrusted suppliers. I have yet to ever see a company serious about production that would do this.

Are you in production? Or just think this is a good idea?

Until 2017, I haven't seen anyone that could raise capital from total strangers and start a company. In 2019 I was working at one of the many companies that did just that.

Until last year, I haven't seen anyone so investing in derivatives without a bunch of capital and access to a reliable brokerage. Today, you can do it without even having a bank account on your name.

> Are you in production? Or just think this is a good idea?

I am not in production, but the thought of creating business and opportunities that do not require trust between parties seems amazing to me.

It isn't. One of the reasons why e.g. the stock market is so huge is that participants rarely need to trust each other as the guarantees are provided elsewhere. Imagine if every buyer had to trust every seller (and vice versa) on every transaction.

Because they trust the exchange who does more than blockchain though. That example seems off.

..And here you can trust the blockchain instead of the other party.

I don't think Blockchain provides what an exchange provides.

How many friends do you have that you would be willing to pay a round of beers or cover their tab at a restaurant?

How many of those friends would you put to invest in a house with you?

Hell, how many of those friends do you trust to give you a backup key to your house for emergencies?

Trust is not binary.

I'm not sure I understand the point of the questions. Why wouldn't I do those things if someone is a friend?

I think the level of trust at the point you do business may in fact mostly be binary. The scale of 'so little trust I need to use blockchain' is pretty high, I'm not sure there are that many reasons to do so.

Really? Of all of the people you have some kind of friendly relationship, you wouldn't hesitate to put a more significant amount of resources at risk with any of them?

> The scale of 'so little trust I need to use blockchain' is pretty high.

Let's say a family member comes to you and say: "hey duxup, I have some medical bills to pay but I am short of a couple of grand. If I go on my credit card they will charge me 10% interest/year. So I was thinking of borrowing it from you and pay you 5% interest. By the way, here is the key to my car which is worth double the debt, the car is yours if I don't pay in time. It's a good deal for me, it's a good deal for you and you have no risk of losing your money"

You might say "Oh, it's a family member, so I trust you. You can have the money, pay back whenever you can". Which is fine. Or you can actually say "Alright, I will take the deal" and just hope that he gets to pay you back without any incident. In any case, no legal entities involved (every one is just going by the word) and the cost of settling a dispute is more of social capital (family troubles) than actual money.

Now, suppose that instead of a family member, it is a co-worker that you know for just a short while. You don't know if the person really had medical issues and you have no personal reason to be connected to the person. But you can get the keys for the car and you got some kind of legally-binding document regarding the conditions of the loan. It seems like a good deal, so why not take it? Not to mention the social capital, is a couple of hundred dollars "profit" worth the risk of having to deal with an someone you don't really know and all the potential troubles?

Now, supposed that instead of a co-worker, it is a total stranger online. You have no clue if the person needs the money for medical bills or if the money is for cosmetic surgery. In fact, it could be just for them to spend on drugs. Still, whoever is proposing the deal has placed a token that controls ownership of a car under a smart contract. If the loan is not paid, you get the token and you can easily sell the car on a distributed marketplace. And by the way, the token keeps a verifiable record of the car history, so any potential buyer can be sure of its value. All of this (giving the loan, checking the car value, signing on the deal, enforcing the rules and resolving potential conflict) can be done on a web3-enabled browser and costs maybe ~$10 in transaction fees. Why wouldn't you take such a deal?

I think when people thinking blockchain they are thinking in line of bitcoin, so it is confusing.

Better to think as "signing messages". also better to think as offline on some steps.

if the internet is to be decentralized this it should all be block chain but i digress not going to happen.

Filecoin appears to be an appropriate use case for distributed trust. Haven't really looked into it though.


> Does anyone have any specific examples where blockchain would be a better solution?

Currency, payment processing and banking are definitely the primary use cases - which is nothing short of revolutionary, in my opinion. In less stable/poorer societies, access to banking, trustless payments, and an immutable ledger are immensely valuable.

Even here in the west, I would kill for an alternative to PayPal for international payments. I was recently forced to receive payment via PayPal. Between their 'generous' exchange rates and fees, I lost out on hundreds of dollars. Shortly after, they locked my funds for 25 days while they reviewed my account. I had zero recourse and just had to wait them out. The same transaction on Cardano would've cost 7c plus the cost of exchanging it back to fiat.

Supply chain is still reasonable, because you can't really trust any part of it.

From my limited knowledge (basically 0) of supply chains, I disagree. (Though I am neutral on the statement "you can't really trust any part of it.")

The anecdote I heard was that say you have a blockchain to keep track of your goods shipping on trucks. When the trucks leave and enter a facility their payload is entered into the blockchain. The idea is that it will make sure that every party along the supply chain can say "well I got the next person what they needed" so you can easily put blame on and correct for missing/failing items.

Sounds good, right? Well, not so fast. It turns out that some of the items were stolen from a truck (whether by an employee or otherwise). The truck arrives at the next facility, scans in and certifies that they have the proper items and none are the wiser!

Hold on, can't they just physically check at each stop to actually make sure that they have the right cargo?

Bingo! At that point, what problem does the blockchain actually solve? Blockchain works okay-ish for digital goods (cryptocurrency), but completely falls apart for physical goods.

> Supply chain is still reasonable, because you can't really trust any part of it

But the part you can't trust is in the physical realm. The fanciest blockchain won't tell you if the package's contents have been tampered with or pilfered.

You can't trust it but somehow your entire business already depends on it?

You can very much trust it. Mostly, that is. The issues usually occur when data flows and physical flows get disconnected. No way blockchain will solve that.

Most companies aren't completely aligned with all their suppliers. Just look at Apple trying to root out child labour. A formally verified supply chain has some limited use cases because you can prove some pieces aren't lying to you (e.g. that you're actually not sourcing your parts from conflict materials, and they are coming from somebody reputable).

> Blockchain has earned a bad name because of sales people trying to sell it for something it can't do.

Sales people in this case meaning the entire community. Bitcoin’s design meant that anyone who buys in later will be making money for early adopters, and that gave everyone a strong conflict of interest clouding their technical judgement. Subsequent blockchains might not have copied the exact deflationary model but the get-rich-quick-without-contributing-anything mentality has become universal.

> Blockchain has earned a bad name because of sales people trying to sell it for something it can't do.

> Sales people in this case meaning the entire community.

Not the Bitcoin community, but certainly most of the others. The Bitcoin community has been pretty clear about Bitcoin's use case since the beginning - inflation-proof digital money - and as long as it remains constrained to that scope then that's actually something it can do.

Bitcoin's programmability was deliberately limited to prevent overloading the network and compromising or diluting its core use case. That also prevented the community from advertising it as a general purpose computing platform and panacea for all the world's ills. That's something all the other blockchain communities have been doing, not Bitcoin's.

Yes the Bitcoin community, more so than most others because it's built on a deflationary model which heavily favors early adopters. One of the major reasons it's failed on its goals to be a currency is that the deflationary model encourages passive speculation holding it rather than using it, but that only works as long as new money is being pumped into the system. Even cursorily following the topic will show many examples of advocates making unrealistic promises which make little sense from the perspective of building a healthy community but are entirely understandable when you realize that they are hoping to see a personal windfall from everyone they convince to use it.

I recently wrote about a disaster that happened to an online ticketing service which decided to use a blockchain as a database

There are lessons to be learnt who try to put everyth to a blockchain https://ruky.me/2020/12/23/somethings-are-never-meant-to-be-...

More importantly, you don’t need it if your domain effectively creates a centralized authority. All of the examples of tracking John Deere parts on the chain were stupid, because functionally there had to be a central party controlling the list of what serial numbers were valid and which weren’t; John Deere themselves.

Its even worse. In many cases blockchains still require a central authority that controls the write access to the system.

A blockchain / distributed ledger can be easily flooded with spam. The only way around that is to make access expensive. That works for the "actual" blockchain because the effort to put something in the chain is the money.

In applications where you don't have want expensive writes, a central gatekeeper (or at least a cabal of power-nodes) is needed to restrict access to prevent spamming.

In my experience, when people say "blockchain" what they actually want is something closer to a Merkle tree like what Git uses for its commit hashes.

It's great to be able to say "here is the list of changes, and I can re-run them and come up with the same hash, so nothing was tampered with"

The "here is a random key that took a lot of work to find and when you add it to the hash chain we get X number of zeroes" part is really not that useful in most cases.

> Blockchain is a very very very inefficient database that removes the need for a central trust authority.

When run at a large enough scale that 51% attacks are infeasible.

my knowledge is near none over the technicalities.. but I had this view that blockchains would be good for planet wide "unattended" data sources.

something like storing a blog text for thousand years, especially after the author passed away where the blog could not be maintained by the author.

no this is not helpful for the flow of the system, more like supply tracking, logistics, farming..

Except it has no power over the physical world. Just as one can lie on a ledger and say slave labor did not make these resources, so too can some one lie onto a block chain about where resources came from.

The chains can only really enforce protection on their own currencies, no?

It was not about the initial data but the overall coverage and self management. How many businesses still have their own ad-hoc half manual uninteroperable process and apps ?

The race to the bottom design of most blockchain tech also facilitates defacto central authorities.

There are details about blockchains that are often missed. That is, not only is trust distributed, but software execution is distributed. Smart contracts in bitcoin and eth are an interesting concept. Think AWS lambda but distributed.

So user centric programming abstractions where users can execute queries and directly put data in the database is interesting concept. Of course you can also do this without a blockchain too if you can deal with a trusted party.

Indeed - a blockchain only makes sense in decentralized trustless proof-of-X systems.

Everything else could be a Git repository or variation thereof (commit chain vs blockchain).

Specifically, a Merkle tree[0], which is the hash tree root of both Git and most blockchains.

[0]: https://en.wikipedia.org/wiki/Merkle_tree

From an academic paper titled "Do you Need a Blockchain"[0] comes the following flowchart, which I think is a great explanation for business failures in this area: https://lh5.googleusercontent.com/DdXkPEPGlfox5LZLR94Bnxz8hq...


It lacks a box for "are my potential investors gullible enough to give a higher valuation/be more likely to say yes if I mention blockchain?"

It seems a fairly significant proportion of blockchain use comes down to that.

Corporate blockchain - as in a way IBM pitched it - yes of course. Most of the people from open blockchains were saying that for years now.

As for open blockchains, like Ethereum, there is really no financial system like it, and the amount of innovation there is crazy. In theory you could replace it with a centralised system, but that’s like saying that you could replace Internet with AWS and just keep everything there.

I get to evaluate a lot of blockchain related pitchdecks, and the first thing I look for is always whether they want to insert themselves as a central clearing house in some form, and almost all the startups I see that has put blockchain somewhere does that.

If they do, that's an instant huge red flag indicating it's probably there because they thought blockchain would get investors excited, not because they have a vision that need it.

A very, very few are using blockchain as a way to enable something larger, by e.g. trying to create an ecosystem where they themselves are just another untrusted player, where blockchain can make sense. And of those that do that, an even smaller number has done that in a way that seems like it's creating a bigger opportunity for them instead of shooting themselves in the foot.

Blockchain solves trust censorship and control. We dont see problems in the US that need those solutions because our institutions are trustworthy and strong. However we literally just had a month where the president was deplatformed, a social network was deplatformed, Robinhood blocked participation in the markets and as always the Fed / Treasury are printing massive amounts of money. I think Americans are going to start valuing trustless systems a lot more after all of that.

>the use case for blockchain can usually be solved with classical, cheaper and simpler solutions by having a central trusted authority for storage (like a third-party service provider), and appending using digital signatures.

Yes, Amazon QLDB is a good example of that, and is the right solution for most use cases where a central trusted authority is acceptable.


In other news a decentralized system can be replaced by a centralized system. I don’t think it’s a very useful point.

It's very "faster horse" syndrome. If people wanted to see the actual applicability of decentralized systems, they'd look into what's happening the early stages of the Ethereum platform building, specifically decentralized finance. But it's easier to just say "blockchain is a bad MySQL" and get 100 upvotes.

Exactly! Why don't we let Google manage all of our email in a single database. It's a way better system! ;)

Agree, but that's how lasers were for the first 2 decades (guessing) of their existence.

Yeah, this was pretty obvious even in the original bitcoin paper.

Pretty much this. Would be interesting to know how many tens of billions were wasted on blockchain projects that nobody needs.

I'm still looking for blockchain-pitches where a small MySQL server wouldn't do the same job better and with less complexity.

I have plenty of criticism of the blockchain hype, but IBM does this for all buzzwords regardless of how useful the underlying technology is. It is all marketing to get executives at large companies to cut large checks. Once those checks are cashed they drop everyone and move on to the next buzzword.

The current buzzword is hybrid cloud. Which is awesome for IBM because most large companies already have some on-prem and cloud resources. So IBM inserts a bunch of people into existing processes and resources and claims credit for creating a hybrid cloud for their customers.

Red Hat employee here. Contrary to your "inserts a bunch of people" characterization, IBM wanted to get into hybrid cloud so they acquired Red Hat in one of the largest tech acquisitions in history, and we create hybrid clouds for our customers. As we are part of IBM now, IBM is in fact providing hybrid cloud to its customers from start to finish in a substantive way. So there's really no comparison to be made to technologies for which IBM has not made substantive investment.

I'm working on many projects with RedHat, and have heavily dealt with IBM in the past.

I can definitely confirm that the IBM statement is true. Execs signed on many projects, and we were always stuck with a blue pile of unusable garbage at the end. For twice the price orginally agreed to...

Working with RedHat as of today is still, well, working with RedHat. Highly competent people building things that will run well for a long time, and (so far) still at a reasonable cost. I do start to see some changes on pricing (high increases are on the horizon...), and more red tape around things that don't fit in the standard boxes. So i'm trying to decouple some areas from being fully dependant to more standardized/vendor agnostic models to keep options opened (mainly in container space).

My Condolences. RedHat was an awesome company and helped make Linux and open source a success. I hope the bluewashing goes slow and you are able to have a good experience there for quite a while.

Different Red Hat employee here, that said, this opinion is my own and I have no real authority to say anything beyond my experience. I, in my job, have a lot of contact with IBM folks now for obvious reasons, but I think it's a slow Red washing rather than the other way around. The independence of Red Hat is fiercely defended and is largely respected, even by IBM. I think we all want this to work as it is, rather than make huge changes, especially culturally.

IBM employee here, can confirm (the Red-washing) after having developed a product that HAS to be running on OpenShift and HAS to go through RedHat's horrible scanning and catalog system. Of course this is in addition to IBM's existing horrible scanning and catalog registries. It's a red layer on top of existing blue layers of tape. It's multi-colored bureaucratic tape all the way down.

I'm sure RedHat can stave it off longer, but a fair amount of the Softlayer folks were optimistic at first too.

IBM never had a Softlayer president (and also paid an order of magnitude less than for Red Hats).

Red Hat employee here as well. Didn't feel any difference, it's still a great place to work, but I don't usually interact with IBM folks.

So it's becoming Big Red instead of Blue Hat? That's encouraging.

One could argue it’s already finished. I was hearing almost identical versions of “no, really, IBM is serious about cloud and that’s obvious from buying us” about a month after SoftLayer closed their deal. I’m starting to wonder if that messaging is in the acquisition welcome packet because the similarities to what my SoftLayer friends argued is remarkable (implying the next step, if history is predictive, is for half of Red Hat to leave and IBM to rename cloud offerings again).

I overheard that Rackspace exited the hybrid market entirely rather than accept IBM terms, which says a lot if true, but it’s watercooler chat and not worth much. It’s plausible because in an “architecture by strategic acquisition” endeavor one courts the cheap options first.

It always takes about 2 years for the corporate hegemony to take over a major acquisition, during which time they have been keeping the new staff under the impression that they bought them to change the acquirers culture not the other way around... then the real re-orgs start.

As an outsider interacting with folks from IBM and Red Hat, if I didn't know about the acquisition, I wouldn't have noticed any major change in their behaviours. So far IBM-in-the-large have been admirably hands off.

So far IBM-in-the-large have been admirably hands off.

Key words "so far". I'm sure people from Truven and all the other acquired companies that make (made?) up Watson Health would have said the same thing initially as well. Right up until they all got laid off...

Sure, but "yesterday's weather" is about a good a prediction as can be made.

I don't think the OP was claiming that IBM doesn't create hybrid clouds, because they clearly do. But "hybrid cloud" is definitely a buzzword right now.

Sure it's a "buzzword" but OP was claiming "Once those checks are cashed they drop everyone and move on to the next buzzword", as though IBM does not have substantive capabilities and investment in the hybrid cloud area. This is the opposite of reality. The phrase "inserts a bunch of people into existing processes and resources and claims credit " suggests IBM is not actually providing the service as well, also not the reality.

Some buzzwords have stronger underlying value than others. Hybrid Cloud is very critical for orgs that want to migrate or partly use cloud resources, say for moving capital expense to operating expense or for burst uses, but don’t want two different tool chains. IBM/RH’s offer is more flexible then Azure Stack or pieces of Google like Anthos, but the flexibility comes at a higher cost.

In any case, is buzzword but not the same variant of buzz as blockchain.

I somehow missed/forgot IBM bought Red Hat. Reading OP's comment, I certainly didn't had the impression we were talking about a 34 billion dollars acquisition.

Give it five years

Buddy of mine worked at a company who was tied to IBM closely. IBM was rolling out its first 'cloud' options and he was so excited to get some on demand type resources.

IBM guys walked him through the process. Then he got an email that it would be a few weeks while IBM installed some more servers in their data center.... he looked closer and realized all his actions on the website was just generating emails to someone else who went and did the thing and emailed them back if they were done or not or had to wait.

Later the process resembled an actual cloud solution, albeit a year or two behind everyone else in terms of features and workflow.

IaaS: Infrastructure as a ServiceNow ticket

Call it "lean startup MVP" and you should be able to raise a few million bucks.

We had a meeting when I was with IBM where MCMP (https://www.ibm.com/support/knowledgecenter/en/SSJP6K_191218...) was introduced as the target front-door for all cloud project requests.

The initial integration is literally just firing off ServiceNow tickets or emails to the appropriate work teams until further automation is created.

The object being to train the end-users to go to the portal instead of contacting people directly or to go directly to the cloud vendor MCMP may be front ending.

We interact with IBM from time to time at my work and in general they don't seem like the most competent people in the world. But also it varies quite a bit depending on which sub-group within IBM you are interacting with.

The nice thing about IBM is that they have an army of lawyers and industry connections that a small company like mine can piggy-back off of now and again.

In terms of Hybrid cloud (launched in 2011 if memory serves) it's more of a 'connect the tech to the buzzword' process, one in which IBM assignes a buzzword to tech its had for decades. Same process with 'cyber' and 'cognitive computing'.

IBM is too slow to actually develop anything ahead of the market without acquisitions.

Think about it, IBM 360 was basically an AWS Outpost 50 years ago!

Is there an example of a blockchain implementation that is tried and true outside of cryptocurrency? I remember seeing when large tech firms were jumping on the bandwagon and proposing things that blockchain could do. Everyone was like "so, that problem has already been solved for years with X, blockchain provides no advantages".

There's some really cool stuff going on in the distributed storage space thanks to this - blockchains have provided a way to pay for, prove you've stored data and provide collateral should you fail to store it, establishing the whole contract without the need for a centralized party, allowing anyone to sell excess storage. Sia is probably the most mature implementation at the moment, but there's a number of competitors in the space.

But as you can see even this is pretty cryptocurrency adjacent. For anything that doesn't rely on distributed trust you're much better off making a database, perhaps with a hash chain glued to it if you need to prevent reversion.

Ironically, tying a cryptocurrency to a real-world use case like paying for storage is the fastest way to throw cold water on price speculation.

Once people can price a coin in terms of real-world assets like storage, it becomes easy to see when the coin is overvalued.

IMO, this is why Bitcoiners aren’t actually interested in making Bitcoin easy to use for payments. They only want two use cases: Buying and holding. Selling or spending will only drive the price down.

> this is why Bitcoiners aren’t actually interested in making Bitcoin easy to use for payments. They only want two use cases: Buying and holding.

TL; DR The political history of deflationary currencies, which until modern times meant any metal-based currency that wasn't being mined at a rate similar to or faster than economic growth.

The cryptosphere is trying to push a populist narrative for cryptocurrencies. Calling it the people’s money or democratized banking.

This seems doomed to backfire when people realize that the vast majority of crypto is held by lucky early adopters. A hypothetical Bitcoin dominated economy would be nearly the polar opposite of fair or democratized. I expect the populist angle to fail as people catch on to the incentives of early adopters, both private and institutional.

I don't quite see why the existence of very rich people make a currency unfair, or non-democratized. I don't think the selling points for populists included "...and everyone will have roughly the same amount of money, too!"

The best I can think of are the possible governance challenges if the major holders have interests that run counter to those of small spenders, say, when arguing over block size or other structural changes.

It's not about the existence of very rich people, it's that wealth is (was) allocated based on luck and being in the right place at the right time to speculate on cryptocurrency.

> I don't think the selling points for populists included "...and everyone will have roughly the same amount of money, too!"

That wasn't the point at all.

It would be one thing if we introduced a cryptocurrency alternative to existing wealth in a 1:1 ratio.

However, that's not possible with Bitcoin. The only way to get some Bitcoin is to trade your existing wealth for whatever amount of Bitcoin can be purchased on the open market. That value has changed by literally 30,000X over the course of a few years.

Migrating to a Bitcoin economy would be a massive wealth transfer from late adopters to early adopters. We're obviously not going to replace current money with Bitcoin, but if we did it would mint billionaires out of people who invested trivial amounts of money in 2010.

However, no new value is created in the process. So where does the billionaire-level wealth of early adopters come from? It comes from late adopters.

Moving bit by bit to a Bitcoin economy is a great way to transfer wealth to early adopters and make everyone else pay for it. Bitcoin people won't admit it, but that's a huge part of the draw of Bitcoin speculation and the reason why so many Bitcoin holders are pushing for a Bitcoin economy.

> So where does the billionaire-level wealth of early adopters come from? It comes from late adopters.

That's an interesting thought experiment. Suppose we convert to a bitcoin economy today. My bank account gets converted over, and my next paycheck will be in bitcoin. Did I lose any money today?

If you are bidding for goods and services against an early bitcoin holder, yes.

>Once people can price a coin in terms of real-world assets like storage, it becomes easy to see when the coin is overvalued.

Prices work in the opposite manner, it's the sum interactions of market participants that drive the price equilibrium. Something being related to a "real-world asset" doesn't make it any easier to price. See "Economic calculation problem" for further discussion.

>when the coin is overvalued

Or undervalued. If you are speculating on exchanges on a daily basis it doesn't matter but for long term holding, which is the easiest strategy for most retail investors it's not a bad thing.

People who have been through this cyclic process already are looking towards the end of the current market run and are suggesting something like a 60% btc 30% eth 10% altcoins that might spike a lot for holding until the next cycle. Why? Because btc/eth are "stable", which in crypto terms they are. This is an important change in sentiment.

Don't forget that gold has different sell and buy prices to discourage speculating on it's pricing( just enough so that right about anything will outperform it in the relevant timeframes). It wouldn't be unreasonable for something similar to happen to asset types of crypto.

I really want this concept to work, but I still cannot ever get someone involved in these projects to really give me a realistic (paying) customer base that scales.

I'm ready and willing to stand up a rack full of spinning rust for such a concept to be a provider, I just have yet to make the model work or trust there is a customer base willing to store their data in such a manner.

IPFS: https://ipfs.io/ Filecoin: https://filecoin.io/

blockchains are all about decentralization which will make more and more sense in the age of de-platforming (facebook, robinhood, twitter, AWS with Parler.. you name it), it makes sense why nobody would want to work on this technology for IBM, completely defeats the purpose.

There's so many other interesting applications for cryptocurrency too. In terms of gaming could you imagine in-game tokenized economies? What if Blizzard built a World of Warcraft token and had a real-time in game economy. What if Rockstar did it with Grand Theft Auto? All of a sudden these IP's have real world value tied to them in the form of a cryptocurrency - it really is interesting if you think it through.

In our system (has event-sourcing parallels) we can optionally use it by chaining subsequent event hashes to each other and letting blocks of such events be signed by an external signing server.

This does not require a bitcoin de-centralized signing process though and is better done via a normal external authority.

It is now harder to falsify server data, since you not only have to recreate all hashes from the moment of change in the event log, but also have them be authenticated by the external source. In addition to the underlying ZFS snapshots a lot needs to be compromised to change data without detection

IMO, Storj (storj.io) is the most mature in terms of viability. The project uses blockchain for metering and economics of the ecosystem but uses a more traditional and practical architecture that make's sense for apps and companies to use.

They seem to be taking a more performant and practical(read: usable) hybrid blockchain approach.

I do like what Filecoin is doing albeit Filecoin is more academic/provocative and a few years away from viability, IMO.

The main thing I like about Filecoin is how Filecoin is tied to IPFS, but IPFS is not tied to Filecoin. It provides useful functions to applications building on IPFS and uses proven tech, but if I just want to use IPFS, I don't have to touch any crypto.

from the article: "Big Blue’s major blockchain networks are FoodTrust, a farm-to-supermarket tracking system backed by Walmart; and TradeLens, a shipping container logistics blockchain backed by Maersk. IBM has also added the Trust Your Supplier network and previously had a go at payments via World Wire."

Is there any analysis out there about how well these systems work, and why a blockchain works better than a traditional inventory management system?

All I can find on these implementations are by blockchain startups who have a pretty obvious interest in making blockchain solution look good. But…how is this better than a good old database? They say “trust”, and I guess that’s true in the sense you can verify the changes to the ledger. But isn’t the lack of trust in the original producers? How is it more trustworthy if they catalog this with a blockchain versus a database?

But since IBM is canning the blockchain group, it's safe to say that those implementations are far from tried-and-true.

Not an "implementation", but for me cryptocurrencies seem like a nice way for replacing the slow international wire transfers that you get from the average bank.

I honestly haven't used crypto this way much, but it seems like it would be a decent way to transact across borders.

This is one of the markets some alt coins with low transaction fees and high number of transactions per second try to tackle.

Decentralized derivatives trading https://injectiveprotocol.com/


LARPing, Hobbiests, Art

All multi-billion $ industries.

Corda? Hyperledger? Fabric?

Agriculture Supply chain?

I keep a small collection of applications where I think blockchain applications actually make sense. So far, most involve things where transparency in a distributed network is very important:

1. Running a lottery over the Internet. There are a lot of people who would be happy to buy a share in a typical national lottery as long as they could be sure they were not getting cheated. I don't care if e.g. 10% of the money goes to fund schools, as long as it's fair and transparent.

2. Internet gambling. If I bet on something, I want the odds to accurately reflect the bets of all the players, and I want to be able to get my money out if I win.

3. A currency for use in online advertising. A lot of the money that advertisers pay mysteriously disappears into the hands of middlemen before it gets to the sites that host the ads. There is also a lot of fraud. Having a transparent and accountable system would make it more effective for people who are not parasites.

Anything else?

> where I think blockchain applications actually make sense.

Blockchain (may) make sense for these applications but did you stop to think if the applications themselves made sense ?

Lotteries, internet gambling and internet advertising are three cancers in my book, I have no interest in solving their problems.

Decentralized exchanges (e.g. uniswap) - users provide all the liquidity, get all the profit, has guarantees etc. Currently eth has high fees (this wasnt the case until recently and wont be in the future with layer 2) so it might look less great but it already has huge traction especially for smaller coins who you'd otherwise have to trade through tiny exchanges that are harder to trust.

Similar for non-currency exchanges (like synthetix) which provide a comparable ways to trade near-arbitrary assets, though admittedly that is less proven to be a clear winner so far.

Compliance and regulatory audits (e.g. ISO certification, SOC 2, the list goes on) seem like a candidate to me. There is typically a standard set by one organization, that has to be complied with by (multiple departments within) another organization, and verified by an auditor external to the first two organization (as well as to auditors within each of the organizations.) A database managed by any of the three players wouldn't be trusted by the others, so it seems like a candidate for multi-party distributed trust.

A central database with records signed by each of the organizations would solve the problem. Those records are independent of each other (my compliance status is independent of your compliance status) so do you no have the need for a consensus algorithm across all of the parties.

and you specifically need proof of work to solve this?

Does it make audits, faster, cheaper, more comprehensive, and more accessible by stakeholders? If so, thats the killer app.

It's orthogonal to all of the above

How would it?


Sorry, but none of these need or can use a blockchain.

1) Powerball seems fine and well. How do you make sure that the records on this blockchain were put there legitimately? (if you use crypto, you are saying crypto is the only use case) How do you generate a winner fairly?

2) Again, how do you make sure all the bets are on this chain? How do you know that all the bets are on the chain legitimately? How do you get the money in and out of this chain? (if you use crypto, you are saying crypto is the only use case) How do you fairly reflect the outcome so people get paid appropriately?

3) You are saying that you just want an ad network that is focused on making the minimum amount of money. The bigger problem with online ads is that most are never even viewed by real people. People don't need a blockchain to chase the $$. Either the advertisers sees results or not and then they stop paying. If a hoster doesn't make any $$, they try a different service to make more. They absolutely don't want to track money from the advertiser through X stops for each ad.

Internet gambling has the problem that you have to trust the "oracle" deciding the dice rolls. You give all your money to an anonymous address that promises you'll get your money back. You still need trust.

In code we trust, and code is law.

The good thing smart contracts are inspectable, so there is at least the open source aspect to help mitigate concerns.

How do you perform validatable RNG in a smart contract, without an untrustable oracle service? The problem with smart contracts is that they can't deal with outside interfaces.

The only gambling smart contract I'm aware of is the Mayweather/McGregor one, which broke because the API they used to report the win changed from reporting 'w' for win and 'l' for loss to 'W' for win and 'L' for loss, and nobody could update it.

Thankfully the author was trustable, after all, and put a refund method in it, otherwise it that money would have been gone forever. [0]

So, the use cases have been tried and they've all failed.

[0] https://www.removeddit.com/r/ethtrader/comments/6w5wcn/impor...

By having the users themselves provide the random input. For example there could be a two player game where each person provides a number and if the sum of those numbers is even player 1 wins and if it's odd player 2 wins. To prevent the second person who submits a number from cheating, both players instead commit to their number by first sending a hash of it. Once both people have submitted their hash, then they both reveal the number they committed to and those get added together. If you fail to reveal your number within a reasonable amount of time, then you lose.

> In code we trust, and code is law.

Which means any and all bugs or exploits are also law. And we know for certain that there will be bugs and exploits, which is why smart contracts are such an exceedingly daft idea.

This is why contracts implement functionality so they can be paused or even upgraded. There are other things you can do to mitigate risk like adding tests, using static analyzers, having the code be audited, following best practices, etc.

On the bright side it might help get industry to more responsible software engineering practices again, rather then move fast and break things and have the users be the guinea pigs.

Do people with gambling addictions really deserve a public record of all their behaviour to exist for people (including coworkers, family, and future employers) to read for all time, including your descendants?

I've seen proposals for Digital Governments platforms using blockchain-based democracy that sounded promising.. but because blockchain space is saturated with fintech people, these projects most of the time get run into the ground when colliding with crypto/ico/tokenization philosophies.

> Internet gambling. If I bet on something, I want the odds to accurately reflect the bets of all the players, and I want to be able to get my money out if I win.

I can't see why a blockchain would help here.

I don't like risk, but I do have recent experience with Internet "gambling" because the previous President of the United States of America persuaded his followers to bet that he'd actually won an election he lost, so I was able to make money answering trivia questions like "Did Donald Trump win the US presidential election in November?" (No).

Although there were some conventional bookies offering this bet, most of the money (as I understand it) went on exchanges, there were hundreds of millions of dollars in the exchange I used. On an exchange they're matching bids against offers, as you would on a stock market, you can either propose odds yourself or you can take the best existing odds somebody else has proposed on the other side of the bet. As with the stock market, when things are liquid enough it's possible to establish a small range of odds in which matches will occur right now, and over time that range changes.

It seemed to me that this offers everything you'd want, the assurance you'll get your money comes down to the institution offering the exchange, and its regulators, but it seems to me this would also be a problem for a hypothetical "blockchain" gambling system.

> but it seems to me this would also be a problem for a hypothetical "blockchain" gambling system.

It depends what form of gambling you're referring to.

A blockchain based roulette smart contract could provide absolute transparency to anyone placing a bet. The exact process which determines the outcome is known, and it cannot be changed. The assurance that you get your money is the smart contract.

A smart contract based prediction market, such as Augur, can offer similar transparency. You have the assurance that, should the market be resolved in your favour, you will get your money. The main concern is the governance structure which determines the outcomes.

And where does proof of work comes into play here?

A lot of the comments here are about trust. About how there is a technical solution to mistrust.

I am a bit confused: Trust really matters. It is one of the most important distinguishing features between societies. Societies with higher levels of trust tend to do better than others. SO they taught me when I studied development economics.

Trust is between people, it is a complex fabric. How can some technological solution replace that?

Put on your black hat (temporarily) and analyse every use case of technology to replace trust and think How could I subvert this? Will not take you long.

We all need to be kind to one another, be trustworthy ourselves, and verify where possible, and be confident in each other.

There is no technological replacement for that

The answer is transparency: in a cryptographic decentralized system you still have human interactions, but they are in the open (in most systems) and can be verified. This is better than relying on gossiping alone.

No, they're not? You can still have backroom deals with a blockchain, you just don't put the backroom deals on the blockchain.

It depends what you're doing, and what you're allowing in your system. You're talking about an omnibus account[1] or a trade that hasn't been settled (but will be settled at some point).

Depending on your system, there are ways to be less transparent. You also lose the transparency that you could benefit from the system yourself, and there's some limitation to it (the amount of the omnibus account is still visible, and money still moves between omnibus accounts).

On the other hand, wherever you are on the spectrum, it is still a more transparent system compared to a blackbox centralized system, or a multitude of blackbox systems trying to settle between one another all the time.

[1]: https://www.google.com/search?q=omnibus+account&oq=omnibus+a...

You do know that events can happen outside of the system, right? Adding more steps to the system won't help.

Not sure how you got that from my post.

I would not consider it a solution to mistrust, but rather a way to obviate it altogether.

Bitcoin is simply a way to achieve consensus about value that is secure and decentralized. Many people consider it a currency, and certainly that is true, but I think it best to consider it a network, or even better, a form of communication. You can 'speak' on the Bitcoin network in very powerful ways. You can broadcast transactions and transfer value with extraordinary security and robustness. Statements like this, without Bitcoin, require enormous effort: legal systems, governments, and ultimately armies.

This result is obviously not a technical one, but a social one. It is mediated by technology, but the rules and incentives of participation are social. You trust Bitcoin 'speech' because you understand what motivates network participants and conclude that fraud is phenomenally unlikely.

But a society that uses Bitcoin is not suddenly without trust, or even diminished in trust. It simply no longer requires trust for certain things.


> Societies with higher levels of trust tend to do better than others

Trust, in this sense, could be considered as a resource. More trust lets you do more things. But if you don't require trust for a function, that trust may be used for other purposes. It is in this sense that trust can be considered a liability: you have to 'spend' trust to achieve an end, and thus must maintain its supply.

A society that cooperates well can achieve more, but may also be more vulnerable if that trust should wane. To whatever extent trust and cooperation are self-reinforcing, I think history has plenty of examples of such systems collapsing to make this a worthwhile consideration.

The $GME situation is a perfect example of where trust broke down which a blockchain solution could solve.


Trading was halted by brokers at a critical time. If trading was done via the blockchain nobody could pull shenanigans like that.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact