"I haven't seen anything as hyped in such a short period of time as blockchain ... The answer to every question seems to be 'blockchain'," said Peter Williams, chief edge officer at Deloitte's Centre for the Edge.
"I expect that at the next Miss World, 'What do you want to do?' 'I want to solve world peace by using a blockchain, and end poverty and world hunger as well," he told the APIdays conference in Melbourne on Wednesday.
Williams said we have an "irrational exuberance" for blockchain, the distributed ledger at the heart of Bitcoin and other cryptocurrencies. He's right. Just follow the Twitter account @bitcoin_txt for some of the more ludicrous comments from its fans. But as I've written previously, Bitcoin is an ideology, and it's unlikely to ever be workable for everyday transactions.
Williams' real message this week wasn't about Bitcoin itself, though, but about all the blockchain-for-world-peace idiocy.
It's worth remembering that the word "blockchain" doesn't even appear on Satoshi Nakamoto's original paper from 2008, Bitcoin: A Peer-to-Peer Electronic Cash System [PDF].
So what is blockchain? Williams offered three definitions.
One, blockchain is an "ongoing chain of hash-based proof-of-work, forming a record that cannot be changed without redoing the proof of work". That's the wording in Satoshi-san's paper.
Two, blockchain is "the consensus process for a specific distributed ledger application, Bitcoin".
Or three, blockchain is a marketing term for distributed-ledger systems in general.
The blockchain-for-world-peace enthusiasts seem to be using the word "blockchain" in that third sense. It's just like five years ago, when every problem would be solved by "the cloud". Indeed, blockchain has been proposed as the solution for real estate contracts, birth certificates, rights management in the music industry, and even real-time ride-sharing.
All that hype means that blockchain platforms are being developed. Take Ethereum, for example. It describes itself as a "decentralized platform that runs smart contracts: applications that run exactly as programmed without any possibility of downtime, censorship, fraud or third party interference".
Williams said that before we get carried away by all the blockchain magic, though, we need to look at how each distributed-ledger system is put together, and look at its performance.
What, specifically, does the ledger record? How smart or flexible is the ledger? Who is involved? Who do we have to trust? What is the basis of consensus? And how is the ledger distributed?
Spend a little time thinking through the trust, privacy, and security aspects of each of those components, and I suspect the magic will start to wear off.
As for performance, well, Williams points out that Bitcoin transactions do have a cost. His back-of-the-envelope calculations suggest that once you add in the cost of the hardware needed to process the blockchain, and the energy it consumes, it's around $6 per transaction.
And it takes quite some time for individual Bitcoin transactions to be confirmed as having consensus approval: around 10 minutes on average.
Newer algorithms are faster. Ethereum currently claims to take 17 seconds to process a transaction, for instance. That's better than Bitcoin, and even newer blockchain systems have the potential for further improvements.
But that's never going to scale up to something like the internation payment clearing-house SWIFT, and handle tens or hundreds of thousands of transactions a second.
In short, blockchain is a lousy database, and it doesn't scale.
"What these early iterations are showing us [is] there's some pretty significant problems to solve," Williams said.
"I still don't think the solutions that are being proposed are at the point where they could scale to those limits, but that's where I think the experimentation comes into it."
This isn't like the early days of the web, or like the current information security landscape, where there's a clear consensus of where we need to go, and it's just a matter of convincing people to make the move.
"Guys, this is not the nirvana that everybody thinks it is, because there are some real limits" Williams said.
Williams isn't alone in his criticism. Far from it. Turing Award-winning Israeli cryptographer Adi Shamir -- he's the "S" in RSA -- said a few sharp words at the RSA Conference in San Francisco this week
Shamir reportedly said he's yet to see a use case for blockchain that can't be solved with an existing, simpler technology.
Disclosure: Stilgherrian travelled to APIdays Melbourne as a guest of Sixtree.