Blockchain. Never have so many people sought so much from a technology understood by so few. There is nothing particularly novel about a divergence between the depth of knowledge enjoyed by computer scientists on the one hand and wider society on the other. But the extent of the divergence with respect to blockchain – or distributed ledger technologies (DLTs) to use the generic label – is profound. What makes it all the more significant is the fact that many of the most ambitious claims being made on behalf of DLTs touch on one of society’s crucial keystones: trust.

For a more detailed overview of how DLTs work, this digital overview from Goldman Sachs is a good place to start, but for the purposes of this post a rough and ready summary will suffice.

In essence, when we talk about blockchain or distributed ledgers we are talking about (i) databases, that (ii) rather than being maintained centrally, are duplicated across a network in such a way that every change to the database is recorded and approved on each node on the network, following an exacting process of cryptographic verification.

It is this combination of distribution and cryptography that underpins the bold claims about trust that some advocates of DLTs make. If trust can be hard-wired into the architecture of the network, the argument runs, then no need to rely on the good intentions of individual participants, let alone the good intentions of a central authority with privileged access to everyone else’s data. The Goldman overview goes so far as to dub DLTs “the new technology of trust”.

On the face of it this is a tantalising, alchemical prospect: the creation of resolute trust among anonymous and far-flung participants in digital networks. Nor could this deus ex machina have timed its own invention any better: how useful for us to have hit upon a trust-shaped technology just as concerns spike about a deepening trust-shaped hole in public life.

But a little circumspection is warranted here. These are very new technologies that are fully understood by a very small number of people. No doubt in time, important and perhaps even transformational applications will emerge – the World Economic Forum is actively engaged in this area, as this recent white paper attests – but there is a hint of bubble-like behaviour in the wide ranging claims of radical transformation that now swirl around DLTs. Here are four reasons that might prompt us to take things a little bit more slowly.

First, we should keep an eye on the widening gap between the claims being made about potential DLT applications and the actual roll-out of such applications. Look past all the talk about technological revolutions and there is not yet much concrete change on which to base our assessment of these technologies. Moreover, anecdotal evidence suggests that a growing number of organisations are looking down the wrong end of the telescope at DLTs: instead of bringing their problems to the table and assessing whether DLTs might help, they are bringing DLTs to the table and looking for problems to which the technology might be applied.

Second, are we sure that we either need or want a technological fix for a decline in trust? While it is undoubtedly true that there has been a worrying decline in public trust in key institutions in many countries, we should be careful about catastrophizing this development. Countless interactions and transactions take place each day that rely on the presence of sophisticated networks of trust, suggesting to some that maybe DLTs are a solution for “a problem no one has”. But even if that optimistic picture is false and we are, in fact, in the midst of a structural decline in interpersonal trust, it doesn’t follow automatically that DLTs are the answer. Maybe our first instinct should be to prioritise the rebuilding of societal trust rather than accepting its demise and jumping to the roll-out of technologies optimised for a low-trust world.

Third, we should recognise that there are hard limits to the extent to which we can sidestep the need for trust, even in a world replete with DLTs. If nothing else, we need to trust the cryptography. We need to trust the network architecture. We need to trust the hardware on which the technology runs. And so on. For most of us, this means taking companies at their word when they tell us that we can trust the way they have incorporated DLTs into their operations. And for decision-makers within those companies, it means taking the word of the computer scientists when they say that their DLT implementations can be trusted.

Fourth, what might the unintended consequences be if the idea took root of replacing socially grounded methods of generating trust with technologically distributed methods? Perhaps the last few years should give us pause here. When thinking about decentralising trust, a useful point of comparison is the internet’s radical decentralisation of news and information flows. This has undeniably had profound democratising effects, but the legacy is not quite that simple. We are becoming increasingly aware of more troubling consequences, such as echo-chamber phenomena that have contributed to levels of fragmentation and polarisation serious enough to have prompted concerns about the health of democracy itself. This is a question we dwelt on at length in last year’s Global Risks Report.

The analogy between the internet and distributed ledgers is far from perfect, and none of this is to suggest that DLTs are a threat to democracy. But we should at least consider the possibility that trying to outsource trust to technology might entail complications and side-effects that we can’t currently envisage.