The newly formed German government has recently signed a coalition agreement that emphasizes a new digital agenda. One of the proposals is to investigate the use of blockchain technology for land title registers and priorities of the digital strategy include AI and quantum computing.
At the same time, we observe that for contact tracing during the COVID pandemic “data [was] often transmitted on paper by fax and then manually typed” into Excel spreadsheets despite the availability of more appropriate tools.
All politics aside, this discrepancy highlights a typical dissonance between the ambitious goal-setting practiced by high-level decision-makers and the much more profane nature of the most urgent problems in the field. I would suggest that this is completely normal and a simple consequence of human nature. Rather than complaining about decision-makers getting distracted by buzzwords, let us accept that these are the rules of the game and play it in order to solve the real problems.
Let us first be clear that this effect is not unique to government and public administration. Many practitioners have observed that data science and machine learning projects in the industry turn out to be 10% about building and applying the most appropriate machine learning algorithms and 90% about making sure that data is available at the right time in the right location, that data from different sources and time ranges use the same conventions, that idiosyncratic data collection artifacts are accounted for, that the right features are extracted and that this entire pipeline is working continuously.
For some of these projects, it may very well turn out that having cleaned-up and timely data available for basic analytics is more valuable than any fancy deep learning based prediction engine running on top. However, that will not stop an organisation from labelling the initiative as a Machine Learning project.
Similarly, de correspondent recently observed that popular blockchain projects could have been implemented without the blockchain without any loss of functionality. As shown in the article, this did not stop people labelling the town of Zuidhorn as a digital pioneer.
This is completely normal. Every human endeavor has a facade presented to the outside world and a reality underneath. A couple will not present itself to the outside in exactly the same way as they interact with each other in private, the way a person presents themselves on Facebook and LinkedIn paints an idealized picture, and dealing with a company as a customer is very different from experiencing the inside.
The entire idea of branding is based on the presence of a polished outside facade which does not reflect all of the complexity hidden inside an organization.
Why should individual initiatives be any different?
The objective problem with specifically launching a blockchain project or an AI project is that both of these anchor down areas in solution space rather than looking at the problem space first. As I have discussed earlier, the goal of a software project is to find a high-value match between problem space and solution space. Asking “which problems can we solve with blockchain technology” does not seem like a winning strategy to achieve this goal. Asking “what are some repetitive, time-consuming tasks in public administration that could be automated” seems more promising. Furthermore, it may turn out that all the necessary technology is already available and the main work that needs to happen is integration (and a ton of change management).
Unfortunately, the more productive approach is also neither sexy nor inspiring. Rolling out decades-old technology in offices around the world makes for less of a catchy campaign slogan than “leveraging AI and the blockchain”.
So maybe we just need to accept that this is the game we need to play to solve the real problems? A trojan horse that sneaks in real solutions under the cover of a hype technology? I suspect that in many places this is already what happens.
The Hype Layer and the Substance Layer
It is maybe time that we evaluate every project on two layers: The hype layer is the aspect of the project that can be sold, which will make headlines, and which can position your organization in its knowledge leadership role. The substance layer concerns the problem that really needs to be solved.
There are some ethical concerns with this, of course. Picking up research grants for a blockchain project and then using this money to install a postgres database while one of the developers writes blog posts about smart contracts does not sound very ethical.
Then again, it is certainly not uncommon for organizations to bend the truth about their activities in a project just enough to fit a narrative.
I think the ethical standards that apply here are similar to the ones for marketing: You cannot outright lie, but everyone expects you to represent the truth in the most favorable light.
If you are selling a blockchain solution, then there really should be some blockchain in there somewhere. But nobody needs to know that it only represents a small fraction of the value actually delivered.
A big question that I’m asking myself is: Who needs to be aware of the hype layer and the substance layer? In other words: Who needs to be in on the scheme for this to work? The project team should be, I suppose. But also parts of your organization should probably know about the distinction. After all, there is an impact on hiring: If the blockchain represents only 10% of the value, it does not make any sense to select most of your team mainly based on their experience with blockchain technology.
Instead of complaining about buzzword driven development, let us maybe accept that this is just part of the game and make it part of our design process to find creative ways to build hype technologies into the solution so that their blast radius is limited – and introduce technology that solves the real problems under the cover of the hype.