Remember blockchain? Or to be precise: remember the get-on-the-bandwagon blockchain technology hype of the last ten years? The ‘big new thing’ that gave us crypto (bitcoin and friends), non-fungible tokens (NFTs) and lately the crash of FTX?
What is blockchain? In the most simplistic way, you could say blockchain is a way where parties that distrust each other can have a shared administration. It replaces trust in parties with trust in digital technology (this is probably the shortest non-tech-speak definition you’ll find…).
Given that the main use cases for blockchain seem to be dark money, tulip-bubble like speculation, and Ponzi schemes, of which only the first is a real — though questionable — use case, it is somehow weird that so many, especially financial, organisations decided they had to invest in this ‘next big thing’ roughly 5 years ago. Not many of those that did invest have actually something running on blockchain infrastructure by now. And some who actually managed to do it, moved away.
The Dutch startup Hyfen created a blockchain-based solution to move people’s assets from one Dutch collective pension fund to another. Last November they published a blog that they — after only 18 months or so — had moved away from blockchain, and explained why they did. What makes this example extra interesting is that Hyfen is not just a party that failed and moved away, they are a party that succeeded and then moved away.
A short while ago also, Dave Karpf wrote a blog post about the importance of making sure your ‘next big thing’ comes with an actual use case. He is right, but there is more. There is a hierarchy of three. Successively: (1) an actual use case, (2) the use case must be realistic/doable, and (3) the realistic/doable use case must have acceptable risks.
Arguments that sell hypes such as “don’t get left behind!” “don’t become a dinosaur!” are more effective in influencing us than we like to admit. We see ourselves as realistic actors but ‘over the top’ optimisim (opportunities) or pessimism (threats) trigger a response in us because they diverge from the ‘reality’ in our minds. Optimism and pessimism sell. Woe the realist.
Take this example: Dutch startup Hyfen actually did create a blockchain-based solution for transferring ‘future pensioner’ reserves from one pension fund to another.
It helps to understand a bit of the background why people thought it would be a good use case: The Dutch pension system is an interesting mix of state contributions (paid by taxing current workers), and collective and individual savings. These are called the first, second and third ‘pillar’ of the Dutch pension system. The state pension (‘AOW’) covers a minimum. A lot of the pension payments come from the ‘second pillar’ collective system, which today contains people’s savings at a size of roughly 1.5 times Dutch GDP. For details, see the aside.
Anyway, when you change jobs in The Netherlands, there is a mechanism in place to voluntarily take your savings in one collective fund with you to another collective fund, this is called ‘value transfer’ (Dutch: ‘waardeoverdracht’).
Dutch workers generally (though thanks to some negative developments like the gig-economy, less and less) save for their pension their entire working lives. The money goes into highly effective (about 6-7% per year safe long term return on investment) and efficient (asset management cost for large funds typically around 0.7%/year) collective pensions, sometimes organised per company, often organised per industry/segment, but in fact governed by the (future) pensioners and the employers, who together pay. E.g. most health care workers save for their pension by saving in the ‘Pension Fund Care and Health’. The Dutch have been saving (instead of simply taxing) for their pensions for more than a 100 years *).
Hyfen built a system for this value transfer, using blockchain. And they obviously did a decent job on being in control of the solution, as they are working in the financial world and they got the OK of the regulator (agreed, that doesn’t say all, but it also doesn’t say nothing). So, there they were, with a blockchain-based solution, OK’d by the regulator. And then they stopped using blockchain. Or, to be precise, after a pretty short period they moved away from their blockchain-based solution, and turned back to the role of a central trusted agent using their own private data. This they recently explained on their blog in a post titled Why we stepped away from Blockchain (Nov 2022):
The hype that made every organisation announce some form of a blockchain initiative has definitely receded. Over the years, many organisations have quietly written off their investments. If you feel this claim requires some substantiation, feel free to check the searches on Hyperledger, which are today down 75% from their 2018 peak (the GitHub contributors on this open source project is down 95% from its 2017 peak).Hyfen — Why we stepped away from Blockchain. Note: the link was added by me, it references the ‘ownership’ part of an earlier article about cloud computing.
Our story clearly belongs to the second category, so what happened?
our experience has been that if you want to make it work consistently with high industry standards, it will on average cost more, require more time, you’ll be facing critical issues more often [and] future possible benefits didn’t outweigh the costs
The truth is, central parties have a worse reputation than they deserve. Every organisation benefits the outsourcing of some operations to a trusted party while focusing on their own business. Even with untrusted parties, countless solutions exist to hold them accountable. In our experience, the legal frameworks used around the business world are mostly fit for purpose. When they are not, legal and governance innovation is an entirely valid way of making things happen without requiring trust-less technology such as blockchains.
To be frank, I think that “central parties have a worse reputation than they deserve” is an irritating statement if it comes from people who used to trash that reputation (and hype blockchain) in the first place. Bad mouthing critics (‘nay-sayers!’, ‘dinosaurs!’) is a rather irritating part of the demagoguery that is often part of hype (and we’ll get back to that below). It is of course nothing new: you can get people to believe almost anything.
It is also a bit hard to keep on bashing ‘central parties’ if — from the perspective of 2017–2018 unexpectedly — you have become one…
But, it must be said: Hyfen gives a good and open account of why they had to move away from blockchain, why their initial (naive?) convictions did not hold. As a result, they offer a good example why ‘the next big thing’ is not always the right way to invest. After all, they started going into blockchain in 2017, they state the system ran for about 18 months until September 2022. So they worked on blockchain for 4 years, got something live in the heavily regulated financial world (which is impressive), then shortly thereafter they decided to drop blockchain and 18 months after going live, the blockchain was turned off. And their story is all the more interesting as they obviously did a decent job on implementing it, so Hyfen is not just a party failing and moving away, they are a party that succeeded and then moved away.
Hyfen is not just a party failing and moving away, they are a party that succeeded and then moved away.
Which brings me to my actual subject.
It was IT founding father Edsger Dijkstra who already lamented that IT was the discipline with “the highest quack density in the world”.
Blockchain is an obvious example of an oversold hype, based on naive assumptions. True believers galore. At the time all these organisations invested in it, say 8 to 4 years ago, it seemed to be a sure bet to the people who were in charge of investing their organisation’s future. They were above all being told — in a very convincing manner by those true believers — that they could not risk to ignore this new technology. Ignore it and they would be or become a dinosaur. Substantial funds (billions of euros/dollars) went into it. Little came out of it.
I was prompted to write this piece in part by this post from Dave Karpf about the fundamental importance of ‘real use cases’ and especially the tweet by @tante he mentions in it about use cases (note to self: stop using tweets as source, especially Elon’s naive version of Twitter is about as morally OK as FIFA). @tante says: “Blockchain is about as old as the iPhone. Are smart phones still looking for a use case? Same for video-capable SLRs, flash-based storage or the Tesla roadster. All these things are Bitcoin’s age.”. Dave makes the point (actually, he made that point in February of 2022 already) that FTX was marketed without an actual use case, the most poignant example being the 2022 Superbowl Ad:
Watch the ad, really. Watch it. And ask yourself — with Dave — what is the argument here?
The argument literally at the end is: “don’t miss out on the next big thing”. Such a ‘fear/threat’ argument is very convincing from the perspective of human psychology, and the people selling it are not necessarily all crooks; many generally are ‘believers’ themselves. (Really, the importance of the subject ‘how do people get convinced of something’ cannot be overstated in this Information Revolution). So, how do you guard yourself against such hijacking of your mind? One solution is to force yourself to see if there is a realistic use case. This expands in a few different aspects, which form a hierarchy:
- Not having a use case at all — the point Dave Karpf is making — seems a very good way for non-technical people — like most general managers — to spot a suspect innovation when they are hyped. This is something you do not really need technical know how for. In the case of blockchain: why is a complex, distributed ledger better than a shared central one with legal guarantees? Really, why? E.g.: what problem are we solving with blockchain technology, exactly? Goldman Sachs not trusting Santander? Solid pension fund A not trusting solid pension fund B? Really?
- But suppose a convincing use case is presented? There are convincing use cases for digital artificial (general) intelligence and quantum computing. But while the use case can be convincing, that doesn’t mean the technology is actually able to deliver or that the use case. While you might be able to imagine a use case, you might not have a realistic (doable) use case. The lack of quantum computing (not of transport — quantum key exchanges or even a full quantum internet are realistic) algorithms that can help in the digital domain might be a good example. We may have some quantum hardware now, but it is probably only good for simulations of quantum problems. Nice for physicists and chemists, not so much in the digital domain. And doable also includes: is it maintainable, manageable, safe, in-control, etc.? What effort does it take to create and keep it running in a good way (remember Hyfen)?
- And finally, the use case may be there, it may be realistic — as in: doable — but what are the risks? The IT models of the quants that led to the subprime mortgage disaster which led to the financial crises of 2008 and beyond might be a good example. They had a compelling use case, they were doable (a lot was simply spreadsheets), but they came with a rather high level of — totally ignored — risk. You may have a realistic (doable) use case, but it comes with unacceptable risk.
Blockchain suffers from all three. Dark money remains the main real use case for blockchain, because yes, crooks have trouble trusting each other. Putting the technology in place to solve other problems turns out to be — see Hyfen’s blog — overly complex and inefficient, a shared central solution is technologically far simpler, and legal measures can do the job of federation well enough. Next, blockchain isn’t really fast (whatever is advertised on Nasdaq’s electronic billboard shown above), and if you use proof-of-work you need to produce quite a bit of CO2 to build the blockchain.In other words: building and maintaining it in the financial world — as Hyfen tells us — is hard and expensive. And certainly, the risk of actual damage to the chain of trust, especially in the public blockchain uses, remains high.
Humans have intelligences that are hardware-optimised to the hilt. That machinery between your ears is geared towards extremely energy-efficient and fast operation. You must be able to react almost instantly to outside triggers to be successful. And not only as an individual, but also as a group. This requires stable convictions. Influencing us therefore works best when the signal is extreme. Extreme optimism triggers us. And even more: extreme pessimism (threat!) triggers us. That means that we are vulnerable for exactly the sort of don’t-get-left-behind hype-peddling we have seen around blockchain. We’re even all vulnerable of being an agent of such hype-peddling.
So, if someone peddles you a new hype, and especially if that someones sells it with a devoid-of-a-real-use-case-don’t-miss-the-boat argument, beware. Look at the three aspects above. The first aspect anyone in general management can check. The other two require you to ask your trusted IT advisor, one that understands actual technology, like your Chief Legal Counsel, but then on technology.
Anf finally, does all of this ‘ranting’ mean I think all blockchain initiatives are by definition worthless? Nope. Though I guess that proof-of-work blockchain will simply be too CO2-expensive to be acceptable, other approaches like proof-of-stake may actually work as a bridging technology between serious entities (like banks and asset managers) in trusted legal settings. There are some interesting problematic consequences having to do with the agility of organisations and society if we massively do that, but that is for another story. But so far, the hype of 5 years ago hasn’t come true. Not by far.
On a personal note: I see it as a fact that human psychology works that way. It is well supported by research (Kahneman, Ariely, Dehaene, and many others). Optimism sells. Pessimism sells. Woe the realist.
Featured image: Photo by Pascal Bernardon on Unsplash. Image from 2019. Note the interesting (false) claim that blockchain makes transactions fast. Or safe for that matter, because blockchain too has its weaknesses.
*) The Dutch pension system system is currently partly being dismantled. There are issues (like ‘doorsneepremie’), but in my opinion none that make this wholesale restructuring a necessity (the restructuring is based on idiotic assumptions about risk), and the whole operation is extremely risky. But that is not the subject here. Return to story.