Artificial intelligence is considered a groundbreaking technology. Now we have to set the course so that AI does not become a risk, says philosopher Christian Uhle in an interview.
And suddenly there is a new presence in the world: Artificial intelligence (AI) permeates our everyday life; it is in smartphones, cars, in laboratories and doctor’s offices, in images, films and music. Some see AI as a promise of salvation, others see it as a harbinger of dark science fiction dystopias. One thing is clear: it is time for us as a society to ask ourselves how we want to deal with the endless possibilities that AI can offer.
The philosopher and author Christian Uhle has dealt intensively with this question and written a book. Title: “Artificial Intelligence and Real Life”. In an interview with IPPEN.MEDIA He explains what risks and opportunities lie in AI, why it is much more than a tool – and why we could quickly become addicted to it.
Mr. Uhle, could your new book about artificial intelligence have been written by an AI?
It was important to me to link the philosophical considerations with perspectives from future research and with small anecdotes, literary stories and my own experiences. So to get involved myself and find my own language. An AI can’t do that.
What if you specifically trained an AI to behave like Christian Uhle?
Even if an AI analyzed all the texts I wrote and wrote a new text based on that, it would only be a simulation. It would never be the text that I actually wrote.
Artificial Intelligence: “Perhaps we will end up in a scenario where AIs read the works of other AIs”
There should be authors who at least occasionally get inspiration from AI. Was that an issue for you?
There were passages of text that I discussed with an AI, and sometimes that inspired me to make additions. When it came to actually writing, I relied on myself. But the possibilities in this field are getting better and better and I can imagine that AI will lead to more books being written in the future. Any person who has ever started writing a book knows that there is a very long way between starting and finishing. Perhaps there will be an increasing number of people who let an AI write parts when they get stuck themselves. Then there would be even more publications than before. Also in science, where there is a lot of pressure to publish. But who else is reading all of this? Perhaps we will eventually end up in a scenario where AIs read the works of other AIs and cite them in their own publications.
To person
Christian Uhle, born in 1988, lives in Berlin
The philosopher is the author of several books and numerous scientific articles
Among other things, he contributed to the Arte series “Street Philosophy” and, as a lecturer at the Karlsruhe Institute of Technology KIT, led philosophical seminars on the question of what influence artificial intelligence has on social interaction
Sounds scary, right?
Not necessarily. Ideally, AI systems will become human assistants that support us in our work and enable us to have more time for our core tasks and achieve a higher level of quality – and, above all, to have more time for interpersonal relationships. But the danger is that we become AI’s assistants.
To what extent?
At work, AIs could become people’s assistants and take on tasks for them, but they could also take on the role of superiors and delegate tasks between people as efficiently as possible. AI must be specifically designed and integrated in such a way that it strengthens autonomy and does not weaken it. This also means that we humans can use our time independently.
Can’t we?
Studies show that people feel like their lives are more stressful today than they were 15 years ago. The question is: Can more technology help us reduce the hustle and bustle of everyday life, or is it already part of the problem? So far, it has often been the case that we do not use the time that we save with the help of technology to relax or concentrate more intensively on our core activity. We use them to get even more on the chain and to complete even more tasks.
ChatGPT is a mass phenomenon: “AI is more than a tool”
Why is that?
Social structures like ours, which are geared towards improvement, acceleration and growth, literally force us to use the freedom that has arisen for even higher productivity and higher output in some form. We should therefore now intensively consider the question of what we really want to use the time for.
In 2023, ChatGPT has become a mass phenomenon. In the book you outline the rapid pace of development of this AI system. Is this a milestone like the Internet?
I think the jump is even bigger. Technology has worked for hundreds of thousands of years with people making and using tools. Be it a hammer, a car or the Internet. With AI, we have not only created a new form of tools, but almost something like a new species that can use tools itself. AI is more than a tool.
What do you mean?
Let’s take a writing program on the computer. I can use this to write texts myself. Things are different with ChatGPT. Here I give an order: I don’t write a text, I let the AI write it for me. This is fundamental and represents a unique development in human history.
In one chapter you mention Goethe’s poem about the sorcerer’s apprentice whose magic gets out of control. Contrary to the poem, we as a society do not have an experienced master who could ultimately fix everything if we lose control of the AI. Is that a danger?
I don’t believe in the dystopian idea that AI will eventually turn against its creators. The much greater danger is that AI works very well and does exactly what it was designed to do. And that this is not the well-being of all people, but rather the particular interests of individual companies or political actors. With ChatGPT’s expanded language mode, which is now available in most countries, this takes on a whole new relevance.
How come?
For several years, intensive work has been carried out to ensure that AIs can hardly be distinguished from humans. Development has already progressed so incredibly far that one would not have thought possible a few years ago. AIs can simulate emotions and be conversation partners to whom you can confide feelings. We live in a society in which a quarter of all people feel very lonely. This is a huge potential target group for AI offers. There is a great danger that these offers are aimed at maximizing usage time in order to increase profits. And that increases the risk of users becoming emotionally dependent.
#created #species