Human-Less Computing

Please note: This article is only available in English.
The digitization is blocked by the human factor. What needs to be done to succeed here?

Has anyone wondered lately why life is becoming more complicated every day? Hey man, the promise was to simplify things! Now, instead of having a common way of doing things, we need to handle multiple things in multiple ways only to deal with one single objective. This is even more apparent in organizational activities involving official government processes.

The paper-less office still remains a dream. Today, I have the need to print out more things than ever before. Forms over forms. There is no way to submit them online - via a simple webpage or via email. Luckily, taxes are already made electronically (at least in Germany - and at least under most circumstances). But here stupid forms with no connection between different points need to be filled out again. I have a feeling that these electronically created forms are (at least to some extent) printed out again in the offices of the financial institutions.

The problem has multiple origins. One is the historical side, which makes a paper form mandatory. However, even if we would skip such a requirement we wouldn't be able to circumvent the problem. Another big factor is the transferal of information. Usually, one form has to transform into another, i.e., one set of data has to be transformed to another. Not always are the fields equivalent. Most of the time form B (destination) only includes a subset of form A (source) with additional information given by a third-party (usually the institute working on the form). The data transformation from form A to B is non-trivial.

Now let's suppose we replace the human clerk with a program, e.g., sitting behind a webpage that contains a form to be filled out. Even though this electronic clerk would know how to handle the data in form A (verification, aggregation, response) it potentially wouldn't know how to transform form A to form B. It would have to be an expert in form B as well. The idea now is that, of course, there is potentially some other electronic clerk who is an expert in form B. The question is now reduced to the interface for bringing form A to form B.

This scenario sounds artificial, but the reality is that today we have a lot of services who are in expert in something, but not in something else. That's alright - they should not be. However, the interface between these services to enable going from one state to the other needs to be developed. It should be generic. It should be automatic. It should be extensible.

The bot framework that I've written about earlier could provide such a solution. However, it demands that every service also comes with a layer that allows tolerant communication. I doubt that everyone will supply a service and a bot layer. Therefore, I think the bot layer also needs to come from the interfacing layer - whoever is providing that.

Wouldn't it be nice to just go online, authenticate with a government issued identity (whatever that may look like - security experts wanted) and adjust details, e.g., like the location. Done in a minute. No paper wasted.

Created .


Sharing is caring!