How to understand the impact of AI systems on society? Most people do not encounter the technology of AI on a daily basis. The algorithms, the data sets, the computers running the software. We do encounter the applications of AI in very many ways. Think of health care, where algorithms can predict how people will recover from a coma or algorithms assist doctors in diagnostics. Or think of culture where the algorithms of platforms like Spotify have a major impact on people’s choice for music and on the development of their taste. But also, think of art producing algorithms that raise many questions about the nature of creativity and artistic property.
AI plays a role in sports where extensive data analytics can help in strategically deploying your talents, or it plays a role in education where learning analytics can give insights in the effects of teaching and learning strategies and also help to define implicitly then what good teaching and learning are. Or in law where AI systems help judges to reach a verdict based on earlier verdicts and similar cases. Or, as a last example, the field of Human Resources, where AI systems can help to select candidates based on earlier selection processes and potentially without the implicit bias that often plays a role between people.
Even though this can also play a role in AI systems, as became clear in the famous case of Amazon, where AI was used to select candidates only on the basis of the match between their CV and the job, but where the AI system ended up discriminating against women because it was trained with the CV’s of people who had been successful within Amazon in the past with an overrepresentation of male employees.
This brings us to some applications
that are also highly contested: killer robots for instance. According to many, decisions about life and death should never be delegated to a machine. It should always be a human decision, Or facial recognition. A technology that could be the end of freedom in public spaces, the end of privacy, maybe even, and also raises a lot of issues about accuracy, surveillance, democracy, consent. So not all AI systems are as desirable as they might sometimes seem.
So before we can dive into the ethical questions connected to AI, it’s important to sort out how to understand the role of AI in society. What kind of influence does it have? How does it have an influence and what’s at stake? AI systems seem to ‘do’ things that until recently could only be done by humans. And this sometimes raises the fear that technologies might start to take over, as it were. Do we really want to have an AI judge instead of a human judge is a question you could ask yourself. And do we want AI doctors making choices about life and death? And what to think about AI systems taking over human jobs leading to technological unemployment?
From the perspective of the theories that we’ve discussed over the past weeks, this framework of AI systems and robots, replacing human beings is not entirely adequate. Instead of analyzing the relations between human beings and AI systems in terms of technologies taking over, we should rather aim to analyze in which ways AI systems mediate the relations between humans and the world. In practice, AI systems do not replace doctors and judges or teachers, but they mediates practices in health care, in law and education. Just like the technologies that we discussed in week two, AI systems mediate the interpretations, the actions, the decisions of human beings at the individual level.
And in doing so, at a societal level, they help to shape interpretive frameworks and social practices like education, health care and work.
So the next step then is how to understand this mediating role of AI systems. How do they mediate? If we compare AI to the mediating technologies discussed in week two, AI systems bring a new form of ‘intentionality’ into the relationship between human beings in the world. A new form of being directed at the world, as it were. And this is the case because they have their own kind of ‘technological intentionality’ you could say. like the example of medical diagnostics. When a doctor uses an AI system to help in diagnostic work, both a doctor and the AI system are intentionally directed at the patient. The doctor has her or his findings and the AI system as well.
Comparing these findings with the early findings of other patients and other doctors. Moreover, the doctor and the AI system are intentionally related to each other. The doctor has to interpret how the AI system makes sense of the patient and the AI system, how the doctor makes sense of the patient, and maybe even of the AI system itself. Seen in this way, the AI does not replace the doctor, but rather mediates her or his work. AI is a ‘cognitive mediator’ if you want. It helps to shape how we think, how we understand the world. And as we saw before this is not the first time in history that technologies do this.
Technologies influence how we think, they influence our minds, our thinking. So the ethics of AI needs to take into account the role of technologies in practices and that role they play because of the influence on people’s understanding of the world and being a basis for the choices that they make. You will learn much more about that in the next step of the course.