On Wednesday of this week, the world umbrella organization for electrical engineers, the Institute of Electrical and Electronics Engineers (IEEE), published the Standard IEEE 7000. That sounds like a message for nerds and the initiated at first. The standard, however, is a revolution in thinking in the digital world. Because it turns the development process by 180 degrees and thereby makes corporations and start-ups more accountable.
If you want to reduce the standard to one thought, it is important to ensure right at the beginning of the development of digital products that there are no problems later. So far, it has been the case that corporations and start-ups bring digital products to market at top speed. In its first few years, Facebook put this to the point with its motto “Move fast and break things”. This is how the digital economy still works in large parts. Problems then have to be solved by civil society or politics.
The IEEE 7000 standard isn’t just a suggestion on how to change that. There are already countless codes, voluntary commitments and ethical regulations that regulate the applications of artificial intelligence in particular and usually only have limited effect because they are only limited to one group, one region, one specialist area and are often only checklists for common places. The IEEE is not only the world’s largest technical professional association, but also one of the few institutions that issue globally applicable standards and norms. For example, so that WiFi runs the same all over the world, connectors around the computer fit uniformly and access to the Internet works the same everywhere, at least in theory.
What if solving digital problems was that easy? There are many good reasons why the technology debates at the beginning of the noughties replaced political discourse as a socially relevant arena. A lot went wrong. Especially with the introduction of low-threshold artificial intelligence (AI) in social media such as Facebook and Twitter or in sorting systems of search engines, facial recognition programs and authorities, it became clear that people with digital technology are at the mercy of machines, which very often the interests of corporations or the Multiply the prejudices of power structures. The usual way of correcting this has so far been to hold politicians accountable, which lagged behind technical progress in terms of time and expertise. If AI now makes the leap from cyberspace into the physical world with the market maturity of digital everyday devices and vehicles, a control mechanism that not only reacts faster, but much more fundamentally to digital innovations will be imperative.
34 computer scientists, humanities scholars and system developers have worked on the new standard for five years
That is why the IEEE-7000 is not just a technical standard. It bears the somewhat cumbersome title “Standard Model Procedure for Consideration of Ethical Issues in System Development”. It took five years to work on it. 34 computer scientists, humanities scholars and system developers were involved. The project was initiated and promoted by business IT specialist Sarah Spiekermann from the Vienna University of Economics and Business. Your work has always been shaped by the humanities. For example, in her book “Digitale Ethik”, published two years ago, she drew a direct historical line from the philosophical works of the 17th century to the transhumanism of Silicon Valley. Again and again she argues against such worldviews, which regard humans as a source of error. And from this she developed the principle of “Value Based System Engineering”on which the IEEE 7000 standard is based.
It doesn’t work quite as easily as traditional technical standards. You can’t just pretend to include a few ethical values here in addition to the connectors and cables. Rather, the standard is a process that gives system developers a process at hand from the start of development that makes it easier for them to take into account a catalog of values that will usually result in the common good and the basic rights of users being preserved. Because it is the system developers who have everything in view in such design processes, from the devices to the programs and automation processes. That is why she did not chair the working group herself, but worked with Ali Hessami, a long-time safety engineer with British Rail, and a consortium of more than thirty scientists and engineers. Many had years of experience in enforcing industrial regulations and when a compliance officer should act.
The instructions for action are aimed at criteria such as the business model, the environment and the mental health of the users
In principle, the standard divides the development process into ten steps that provide clear instructions but no content. The direction of attack is already very clear, because the challenges that can be met with the standard are clear.
This includes not only the quality of the software and hardware, but also the business model and corporate culture. If, for example, it had been ensured in the early stages that social media was not funded by ads but by user contributions, the world would have saved itself a lot of trouble. Effects on the environment are a criterion. You can currently see this in the cryptocurrencies of the so often celebrated blockchain technology, which consumes an incredible amount of electricity. But also in the increasing number of AI applications in household appliances and vehicles that run on structures made with raw materials that will soon have to be mined under the oceans. The mental health of the users is a focus. Addictive behavior and depressive effects through the manipulation of feelings and self-worth are acute mass phenomena.
In the digital world, “a time of seriousness and responsibility” is beginning, says business IT specialist Sarah Spiekermann
The time for the standard is just right now, says Sarah Spiekermann. “The innovation of digital technologies has passed the phase of joy in playing, novelty and experimentation. This is the beginning of a time of seriousness and responsibility. This is not just a reaction to the end of digital pop culture and the end of neoliberal forms of capitalism but also on the real challenges that a mature IT industry has to face. “
The main advantage of such a standard is that it is globally valid. Laws, i.e. the usual political response to digital problems, only ever apply nationally or regionally. The road from a standard to general validity is still a long one. One advantage is that the IEEE, with its headquarters in New York and New Jersey, is primarily recognized in the USA, where most of the digital innovations are being developed. In Europe, the ISO norms and standards of the International Organization for Standardization in Geneva apply above all. But even in Europe, companies, organizations and authorities can require suppliers to develop their products in accordance with the IEEE 7000 standard process.
That would not only be a step forward in dealing with digital technologies. It would be a regaining of sovereignty that is threatened by the gradual loss of control in the comfort bubbles of the digital world. Precisely because decision-making processes will be automated in the future. They will be as banal as the answer to the question of which ingredients go into the stew in the evening, and as far-reaching as the specification of which therapy is used against an ailment. A process that guarantees not only technical but also ethical quality in machines could ensure that the next steps in digitization are real progress.