Tuesday 14 May 2024

Trust

 In the AI world that is unfolding around us, it is going to be increasingly difficult to know what to trust and who to trust. This issue has been known in science for a long time, but it is now becoming apparent in everyday life and politics. Von Foerster and Maturana gave this wonderful talk about it many years ago: 


Von Foerster makes the distinction between two words that mean "truth": the Latin "veritas" entails checking reality (German "wahrheit" derives from this). The English word "truth" on the other hand derives from "trust" - in other words, it becomes a matter of interpersonal agreement in terms of establishing truth. Since von Foerster was committed to arguing that there was no objective reality to check, the issue of trust and truth is central.

It is trust which is being affected not just by AI but by all forms of technological communication. Trust demands human intersubjective engagement - quite simply, looking into each others eyes - and most forms of technological communication do not allow this (text, phone, email, etc). Moreover, the interaction between people using these media of communication will unfold differently from how things unfold in interpersonal intersubjective communication. 

These days, we can't be sure about anything. We can't be sure that the sender of an email is who they say. We might assume, for example, when somebody calls us up, that the person on the other end of the phone line is the person they say: but it might not be. We can't be sure that a paper was really written by professor X, or that a student actually wrote the essay they submitted. 

This is central to the problems we are going to face in the coming years. So we are going to need new ways of establishing trust and confidence in communication.

I'm interested in what people think about this - about what we can do to establish new levels of trust. Since my googleform experiments are going reasonably well, I've created another one (I hope you trust me!). There is a very simple question: How would you establish the trustworthiness of a communication in the absence of being able to look someone in the eye?

Interested to hear the responses! 



No comments: