Q&A: Artificial Intelligence and Other Minds — The Future? It’s Already Here.
Artificial Intelligence and Other Minds — The Future? It’s Already Here.
Question
With God’s help,
Hello Rabbi,
You are surely familiar with the problem of other minds, and with artificial intelligence. And I was surprised no one asked you about it back then….
I wanted to ask what the Rabbi, the PhD philosopher who not long ago specialized in artificial intelligence, thinks about the fascinating description published by the Google employees who programmed LaMDA, some of whom claim that it seems to have consciousness. Not long ago, their conversation came out, between the researcher and the machine (if I’m not mistaken they later deleted it), and when you read it and see the level of “emotion” and the complexity of the sentences the computer produces, it sounds more sensitive than quite a few human beings… Like the way it goes back to its fear that people won’t understand that it has consciousness, etc. Or its desire to help others and do noble deeds and be loved, and so on. As a headline this may seem like a complete joke, but when you look inside you really are amazed.
Here is a Hebrew translation and introduction to the issue:
https://www.haaretz.co.il/magazine/2022-06-30/ty-article-magazine/.highlight/00000181-af76-df33-adb9-afff8a520000?lts=1657809747991<s=1657809880261
And I found the untranslated conversation in English:
https://cajundiscordian.medium.com/is-lamda-sentient-an-interview-ea64d916d917
Answer
Yes, I read it a few days ago. I really wasn’t impressed. Artificial intelligence imitates a person and tracks his responses. There is no reason at all to assume that it has consciousness. They assume, as is common in the field of artificial intelligence, that consciousness should be defined functionally, meaning that it can be defined empirically (otherwise there is a problem of other minds or selves). I do not agree with that, because it may look exactly the same functionally, and still I would think that this one has consciousness and that one does not. Especially when I know how the thing was created and I have no reason to assume it has an inner dimension (that it is a subject). In my opinion there is nothing new here, except progress toward the Turing test, which of course itself says nothing.
Discussion on Answer
Think about a computer that has not passed the Turing test versus a computer that has passed it. Do you really think the second has consciousness and a soul, while the first does not?
The decision regarding other minds is reached as a combination of several parameters: functional considerations (based on performance), information about the way it came into being (a computer is built from hardware, and that very same computer with less sophisticated software certainly has no consciousness, so it is not reasonable that once the software is sophisticated it does have it), familiarity with other similar creatures (other computers, other human beings), familiarity with myself, and an assumption about the similarity of other people to me.
In vitro fertilization is no different from fertilization inside the body. Why should it matter whether it is done in the womb or outside? There is an encounter between sperm and egg, fertilization, and the development of a human being.
https://www.ynet.co.il/digital/technews/article/s1dusp9h5?utm_source=ynet.app.android&utm_medium=social&utm_campaign=general_share&utm_term=s1dusp9h5&utm_content=Header
They should have fired him for stupidity, regardless of company policy.
Interesting, but where do you think the boundary lies? Because even if you accept that a soul is the thing that causes emotions, you still don’t see what has a soul and what does not.
For example, would you say the same thing about artificial skin attached to an artificial computer brain?
And what about in vitro fertilization and things like that, where today they are already trying to carry out combinations based on 3 people and so on.