A new report sticks a pin into the huge hype bubble surrounding blockchain.
The report from research firm GlobalData, “Blockchain – Thematic Research” [fee required], argues that the new technology’s bubble “will burst in the next two years [and] will have lost much of its gloss by 2025.” Summarizing blockchain as “awash with hype, but with a powerful core value proposition” of distributed ledger tech, the report was researched over the last 18 months and involved interviews with about a hundred organizations.
Blockchain protocol certainly has some value, author, chief analyst and former software engineer Gary Barnett told me.
But, he cautioned, “it’s not magic.”
In “19 out of 20 cases” where blockchain’s use is touted, he said, “you can easily replace it with a database” or other kinds of technology.
From the report:
Blockchain technology, often referred to as “distributed ledger technology” (or DLT) is a means by which the participants in a distributed network can each maintain a copy of an immutable ledger of transactions, and where transactions can be executed without the need for a single, central, coordinating authority.
The ability to effectively execute transactions without the presence of a central authority is regarded by many as the chief benefit of blockchain technology; as it creates the promise that organisations will be able to transact business with one another without being subject to third-party control. However, this freedom from central control comes at a very significant cost, which renders many of the claims of cost reduction, and improved efficiency, open to debate at best, and simply false at worst.
GlobalData adds that DLT is “complex to implement, extremely resource intensive to operate, and will not replace modern database technologies on a wholesale basis.”
New forms of DLT are emerging that address these issues, but Barnett pointed out to me that many of the purported benefits of blockchain could be accomplished by other, more conventional means.
‘Will always be slower’
For instance, he said, many of the radical reductions in costs that blockchain advocates claim are actually because old business processes are being dramatically streamlined, such as eliminating all the middlemen commonly required to record transactions, and not because of blockchain itself.
In fact, blockchain tech is often extremely computer- and energy-intensive and processes transactions very slowly. A single bitcoin transaction — the original use for blockchain tech — requires about 5,000 times more energy than a single Visa transaction, the report says, because of the required consensus approvals and the duplication of data.
“Any transaction that requires the coordination of more than one coordinator will always be slower” than other solutions, Barnett noted.
Blockchain’s difficult-to-change recording of data is usually seen as a benefit, but that data is often only transactional. The “meat” of the data — such as medical histories or intellectual property — is often stored in an encrypted form in conventional databases or other storage repositories, with the blockchain itself acting as a gate and a pointer.
One key reason for that, Barnett pointed out, is that the entire blockchain is typically replicated in each node held by each participant, which would require enormous storage and processing power if it had to contain content as well as transactions or pointers.
It would be “insane [to] make 50,000 copies of [a] song on 50,000 machines,” he noted, instead of using a secure database.
Another reason content is often stored on conventional databases is because of the newly implemented General Data Protection Regulation (GDPR), which stipulates that vendors handling personal data for EU citizens must allow the permanent deletion of personal data if so requested.
‘A Holy Grail fantasy of technology’
Then there’s the issue of trust. Advocates define blockchain as a decentralized system that guarantees the validity of entries because they are typically verified by a consensus of all participants, a kind of trust without a central authority.
This can be useful, for instance, in tracking the provenance of products along a supply chain. A marketer, for instance, could audit the route taken by rare cocoa beans, from their harvesting in the jungles of Peru to their passages via train, truck and cargo ship, and eventually into a chocolate bar offered for sale in Chicago.
But, as Barnett points out, there’s the obvious fact that the blockchain consensus only verifies the transactional details of these cocoa beans — this entry was made on July 16 by such-and-such trader, for instance. The blockchain consensus can’t verify that these particular cocoa beans were actually found growing wild in the Amazon. And, he noted, there are other effective ways to verify the provenance of products without employing blockchain.
One of the most commonly cited values of blockchain tech is its use of smart contracts, first implemented with the second generation of blockchain, Ethereum. Smart contracts are self-fulfilling agreements embodied in software code, such as the release of payment once an ad has been displayed on an agreed-upon website.
But Barnett dismissed smart contracts as “a Holy Grail fantasy of technology” stretching back many years.
First, he noted, encoded business processes and automated settlements do not need blockchains. If you want to see business processes embodied in code, Barnett said, check out enterprise platform SAP.
Besides, he said, smart contracts add “developers to a room that already has lawyers in it, the two kinds of people most likely to [screw] something up.” As the complexity of the agreement increases, he said, the smart contract’s key attribute — triggering action when machine-verifiable terms are completed — loses its appeal.
‘Peak snake oil’
Smart contracts are not very smart, he said, since they are based entirely on agreements about whether a certain event is triggered. If someone delivers a pound of prime steak and the smart contract uses that trigger to deliver payment, he said, there is no provision to verify that this is actually steak and not horsemeat. The smart contract needs to verify the satisfied condition through digital means to be automatic, but digital verification has its limitations in this analog world.
“Almost every single technical problem has more than one technical solution,” he added, and the use cases for blockchain are no exceptions. The only possible exception where blockchain is the only feasible solution, he said, is cryptocurrency such as bitcoin. This was where blockchain was born, as a solution providing automatic consensus to make sure that a bitcoin can be spent only once.
But, aside from that and the specific use cases where distributed ledgers offer additional value, Barnett doesn’t see many “only blockchain” use cases emerging over the long run.
One key use case category for distributed ledgers, the report says, is cross-border trading, where a distributed ledger that does not require a central authority can be useful. These situations include financial instrument trading, land registries, settlement for gold bullion trading, supply chain audits or international trade settlements.
Blockchain’s value “is very narrowly scoped to a set of scenarios where it is impossible to agree on a central point of trust,” he said.
As for the more grandiose claims, Barnett told me, blockchain has now “reached peak snake oil, and we’re about to plummet into the caverns of common sense.”