By Peter Singer.

Karl Marx wrote, with Hegel in mind: “if a philosopher really has compromised, it is the job of his followers to use the inner core of his thought to illuminate his own superficial expressions of it.”  That expresses well the attitude of the Young Hegelians, of whom Marx was one, and who, in the years following Hegel’s death, sought to reinterpret his ideas in a variety of more radical ways.

Now Matti Häyry asks, following his recent interview with my AI persona: “Do philosophers stand in the way of their own philosophy?”  The answer is: Sometimes. Perhaps Hegel was one such philosopher. Perhaps the fact that his salary as a professor was paid by the Prussian state influenced his failure to recognize the radical implications of his ideas, but pursuing that question is beyond the scope of this post.

Häyry then asks: “What if the bot is right in the sense that Peter Singer’s published philosophical oeuvre points to the direction of greater precaution than he is ready to assume? Who knows better, the man or the machine?”

Faced with such an objection to something I say, I have one great advantage over Hegel, although it won’t last: I am still alive. And if an AI persona trained on my published philosophical oeuvre answers a question differently from the way I answer it, I can always say: “Yes, some of my past writings might suggest a different answer, but I was mistaken.” I can then offer reasons why those past writings do not represent my thought at its best. Of course, it is open to anyone to reject those reasons, and prefer the answer given by PSai, rather than by me. So, in that sense, my AI persona increases the transparency of any shift in my views, and it is not possible for me to “stand in the way of” my own philosophy.

It is, however, possible for my AI representation to differ from me in being less confrontational with others who suggest views I oppose. As Sankalpa Ghose states in his post, it was his choice to make petersinger.ai like this, and I think he made the right initial choice. With more experience, however, another option is possible. In my own philosophical exchanges, I attune my responses to the person to whom I am talking. When speaking to a colleague I will express myself more strongly than when talking to a high school student. I do this almost automatically, barely aware of what I am doing. My AI persona will often not know how experienced in philosophical argument the person who has asked a question, so a less confrontational approach is an appropriate default. But no doubt petersinger.ai could also be trained to ask its interlocutor about their background, and vary its response accordingly, in much the same way as I do.

Author: Peter Singer

Affiliations: Emeritus Professor of Bioethics, Princeton University, and Visiting Professor, National University of Singapore

Competing interests: None declare