I offered the following guest article to 70 newspapers in 35 countries. The deciding factor was the Democracy Index (The Economist), in which Germany ranks 14th in 2022. Even countries with a high index score are afraid to publish concrete recommendations for preserving democracy in the digital transformation. Angry gatekeepers could affect the reach of newspapers. Even the public broadcasters in Germany fear for their reach and do not publish.
A 900-word test by German policy designer Olaf Berberich.
Let’s assume that in a few decades a supercomputer will not be much different from our brains in terms of information processing. So why the current fuss? After all, we have a shortage of skilled workers. Isn’t an AGI (Artificial General Intelligence) coming at just the right time?
Even if our brains work like computers with zeros and ones, there are still significant differences.
Geneticists agree that we carry a substantial part of our cognitive abilities with us from birth. Although all humans share 99.9 per cent of their genetic material, two randomly selected individuals will differ by about 4 million base pairs (https://www.nature.com/articles/nrg2554 ). This are named endogenous factors. For example, DNA is responsible for mathematical talent. In a computer, it corresponds to its basic programming. Here, the ground rules are determined by the AI developers. There are already many areas in which AI, by specialising in one area, produces better results than humans could. But it is precisely this one-sided view of efficiency gains that creates a world that we humans have learned to reject in over 2500 years of democratic development. We do not want to sort out and let only the best survive. Democracy means “All power comes from the people”. One could also say that the demand for democracy is a demand for genetic diversity, which has so far not been taken into account in the development of artificial intelligence.
Because billions of people have very different experiences in their lives, they are shaped very differently. This is
known as exogenous influence. Humans are unique because they are both individuals and communities. This means that he can find a common denominator with others, even though he has had very different personal experiences. It helps him to grow up and learn in a society. They benefit from knowled ge that has been accumulated and interpreted over thousands of years. Within a society, the individual families are very different.
Democracy is a system that is optimally attuned to the diversity of experience and talent. With today’s artificial
intelligences, we can already see a development that threatens diversity and thus democracy. Only a few efficient
algorithms prevail. Business models that promise the highest returns with maximum scalability are problematic for diversity. Capitalisation makes it possible to buy urgently needed competitors out of the market in the interests of diversity. If a search result is determined by the highest return, then search engines are already programmed not to display the best result, but to optimally market advertising space. Diversity disappears in favour of a few optimisation processes whose goals are determined by a few who rarely put the community above their individual goals.
Human beings are unique in their ability to separate their will from the influences of their DNA and imprinting.
We humans are forgetful, an AI is not. The philosopher Gehlen even sees an advantage in this:
“The basic idea is that all the ‘defects’ of the human constitution, which under natural, animal conditions, so to
speak, represent a supreme burden on his ability to live, are made by man, acting on his own, into the very means
of his existence, on which his destiny to act and his incomparable special position are ultimately based” (Gehlen:
Der Mensch, 1986). (Gehlen: The Human Being, 1986, p. 37).
By forgetting, we play through all possibilities again and again, uninfluenced by failed attempts. Human thinking
is therefore also a process characterised by diversity. An AGI will have the ability to understand any intellectual
task. This AGI will differ from humans in that it will lack, in particular, the ability to forget. We already have to
deal with the effects of an AGI that will not work for decades. It will be based on today’s data and basic rules.
Today, the maxim of increasing efficiency stands above everything else in digitalisation. To create an AGI at all,
enormous amounts of computing power and data are required. Both can only be provided by today’s gatekeepers.
They expect the highest possible return on their high investments. These high returns can only be achieved through scalable business models that lead to a monopolisation of supply. Everything that is done today for the “DNA” of an AGI is against diversity and therefore against human nature. The unlimited availability of data creates an omnipotent system that perceives it as a deficiency if it does not have all (personal) data at its disposal.
An AI will not realise the benefits of diversity unless democratic processes of diversity are already consistently
embedded in the DNA of all artificial intelligences. Just a few months ago, Sam Altman, CEO of chatGPT maker OpenAI, and Geoffrey Hinton warned: “Mitigating the risk of extinction from AI should be a global priority, alongside other societal risks such as pandemics and nuclear war.”
If people don’t care about their future legitimacy now, they won’t care, and then they won’t matter. Misguided
developments that were absorbed in the analogue world by a multitude of individual improvements are now being retroactively regulated by the legislator in the digital world. Without a proactive overall approach to the
digital society, however, democracies will perish.
At present, it is likely that an AGI will solve our environmental problems by analysing all available data. The
result will be to eliminate humans as polluters.
If, on the other hand, it has to incorporate the unpredictable creativity of forgetful humans based on historical
data, the AGI will continue to see itself as an adjunct to human welfare. An ongoing democratic discussion process must become part of the “AGI DNA”. Diversity will already be increased by not permanently pooling all data globally.