4. Preface
4.3 Digitalization
Digitalization has already significantly changed large parts of our everyday lives. With the help of computers, even existing societal systems can be organized much more efficiently than previously possible. Without having to change anything about their fundamental structure.
One example is submitting your tax return digitally instead of sending it to the tax office as a stack of paper. The processes are exactly the same. But because the data is available digitally (a file with structured data) rather than analog (paper), large parts of the necessary tasks can be automatically performed by a computer, instead of a civil servant having to manually perform them for each individual tax return.
The result of this is a leaner and faster administrative machinery. Furthermore, it's cheaper (far fewer salaries need to be paid) and fairer (as long as good rules have been established—the computer automation never decides arbitrarily, but purely based on these rules).
The potential for improvement through digitalization alone is enormous, and gradually realized. But the societal systems in which we currently live (social security system, education system, justice system, government, ...) all originated long before the beginning of digitalization. Just as digitalization can enormously improve existing systems, it also opens up many possibilities that would not have been conceivable without it.
Since the possibilities of data transmission and processing by computers will play a major role in my futurities, I would like to take a moment here to explain more generally the great potential that digitalization truly offers when it is incorporated into the design of systems from the start.
As long as the collection, processing, and evaluation of data can be carried out fully automatically—without requiring the working hours of a government employee—the ongoing costs (software maintenance, server operation, system monitoring) are negligible. Systems that would have meant an absurdly high effort in analog form can suddenly become the best possible solution to a problem when digitized!
The key building blocks of this digitalization, which will repeatedly appear in my futurities, are:
1. How the data gets into the computer system. The source can be data that has already been collected and stored. Or data retrieved from other systems. The source could be an application that citizens use to interact with the government, or an application that government employees use to do their work. Or sensors that measure something somewhere. Or cameras that constantly film a location.
2. According to what rules decisions are made based on this data. That is the first huge advantage that digitalization offers: as long as I can clearly formulate how a decision is to be made based on this data, this entire decision-making process can be automated, without human intervention or effort. Arbitrarily large amounts of data can be used for this, and the calculation can be almost arbitrarily complex without incurring noticeably higher costs. It's just the one political decision at the beginning: what exactly that decision-making logic looks like. Additionally, there needs to be monitoring for irregularities (program errors, exploitation of edge cases and loopholes) and monitoring to ensure that the result of the entire system is actually the desired one.
3. The use of AI based on neural networks to evaluate data and identify problems. Given that the capabilities of AIs are currently increasing significantly year by year, it is not very useful to list here exactly what is possible and how exactly it works. AI is constantly getting better at learning from large datasets, drawing conclusions, and making predictions about future developments. Within the data of a digitized societal system, these insights can then be used to detect fraudsters, improve the rules by which automated decisions are made, or notice emerging problems early on. There are many warnings about the consequences of data collection mania and the insights that can be gained from it. But like any other technology, this is just a tool. Of course, it can cause a lot of damage. But in the hands of a responsible state, it also opens up enormous positive possibilities.
Important: AI should never replace the clear and human understandable rules (⇒2.), according to which the overall system functions. Otherwise, it will become a black box whose decisions cannot be controlled!
4. Providing a public API* (Application Programming Interface). An API is an interface, a specialized language, with which a software system can be addressed by another software. This makes it possible to use the data that the state has collected in a wide variety of ways. This allows, for example, companies, NGOs (non-governmental organizations), or individual programmers to offer alternative applications or apps to interact with a state system. For example, to file a tax return. It is the API's responsibility to ensure data protection, so that access and modification are only possible if the user has securely identified themselves. Furthermore, it ensures that the data is only transmitted securely encrypted. APIs keep the government offering from being a monoculture. Instead, the best software prevails for interacting with the system. The data is not a silo, but can also be used for other purposes and combined with other data.
Wherever possible, software for societal systems should be developed as open source*. So that the source code is checked for errors by many eyes, many people can contribute improvements, and society as a whole has confidence in it.
How about a (definitely not serious) example?
The citizens have elected a party to power that has promised to give away state-funded ice cream as a reward for exercise. Every citizen is to receive one free ice cream cone at any ice cream stand once per week—provided that the stand is at least a minimum distance away from any other stand where the citizen has received free ice cream within the past seven days. This is intended to encourage citizens to get out of their own four walls more, and to make them happier.
The state provides two apps (and corresponding APIs, in case someone wants to offer alternative software). One app for citizens who want to eat state-funded ice cream (users), and one app for the ice cream stand operators. In both cases, the person must log in to the app, so that the state can identify them.
The app for operators: It provides a QR code*, which can be printed out and displayed at the ice cream stand. The operator also stores the GPS location of their ice cream stand in this app, and their bank details for money transfers from the government.
The app for users: This app allows citizens to see the locations of participating ice cream stands on a map, and which ice cream stands they have already received free ice cream from in the last seven days.
This app has a feature to scan the QR code of an ice cream stand using the smartphone's camera. If a citizen does this, the ice cream stand operator's app shows that this customer is entitled to free ice cream, the operator hands the ice cream out, and his counter for provided free ice cream cones increases by one.
The state will transfer a fixed amount of money for each ice cream cone to the operators account. And it logs that this citizen got ice cream here at this time, for later decisions of whether or not they can get a free ice cream cone. Additionally, citizens can also rate how satisfied they were with the ice cream in their app. If the ratings are too low on average, the ice cream stand will receive less or no money from the state. That protects against fraud, and it motivates the operators to offer the best possible ice cream, to be friendly, and to provide a good atmosphere.
And then, of course, there's an AI that looks at the collected data, analyzes it, and checks for irregularities. For example, it might notice that an ice cream stand is constantly changing its GPS coordinates—something fishy is definitely going on there. Or that the customer profile of an ice cream stand is completely atypical: Is the ice cream stand even there? Is it really giving away free ice cream cones? Every fraud will lead to irregularities in the data and can therefore be noticed by an AI.
How much effort does this new social system create for the state? First the API and the apps need to be programmed. Then there are ongoing maintenance costs for the servers on which the API is offered and on which the data is stored and analyzed. There are costs for maintaining the apps. There will be service staff that can be contacted in case of problems by ice cream stand operators or citizens. And there will be employees who will review the anomalies reported by the AI, visit the corresponding ice cream stand, and, if necessary, initiate criminal proceedings or otherwise react to misuse of the system.
What definitely does not exist, however, are any processes that must be carried out by a state employee for every ice cream stand. No permit is required to allow an ice cream stand to offer free ice cream19. No government employee has to manually enter bank details or GPS coordinates. No government employee has to transfer money to the operator. There is no need for regular or even one-time verification of whether the ice cream stand actually distributes free ice cream cones (so long as no suspicion of fraud arises due to the AI). In other words, compared to any possible analog system, the state administration of a well-designed digital system can be orders of magnitude more efficient.
The building blocks of digitalization, which we described in general terms above, are used here as follows:
1. The data is transmitted into the state's computer system through the two apps. The time and location where each QR code is scanned are the crucial data points.
2. The algorithm determines whether the citizen receives the ice cream based on temporal and spatial distance. And how much money the operator receives, based on the satisfaction rating.
3. An AI is used to detect issues (e.g., fraud attempts).
4. There is an API for alternative software or integration into existing software (for example into a hiking app).
As mentioned, this was a trivial, non-serious example. But these concepts can be equally applied to serious societal systems.
With this, all essential concepts have been explained and all the necessary tools have been laid out to begin designing concrete futurities.