Saturday, December 16, 2017

Emotional intelligence for artificial beings? With a potential application to Shakespeare

2AI Labs, Tim Barber and Mark Changizi, has announced a model of emotion: VICTOR Socio-Emotional AI.
We at 2ai Labs have spent the last decade deciphering the hidden language of emotions, and have developed a unifying theory of emotions, of their meaning and the machinery underlying emotional reasoning. The theory emanates from a first-principles approach to the fundamental issues pre-linguistic animals face in communicating so as to settle disagreements. It has two key pillars.

I. Comprehension of emotions: The technology understands the specific meanings of emotional signals, including (a) what I want, (b) how compromising I feel I am being, (c) my hand strength in a situation, (d) my opinion about my opponent’s hand strength, and (e) my confirmation receipt of my opponent’s signals.

II. Emotional reasoning machinery: The technology consists of an “emotion engine,” and allows the construction of AI personalities to suit one’s needs. The machinery allows personalities to vary along socio-emotionally sensible dimensions, and so one can, for example, vary their level of aggression, willingness to compromise, and degree of hospitality. Social intelligence also requires understanding social currency, or cool, and that a person’s emotions, and an AI’s emotional response, are bets; this is part of VICTOR’s machinery.
This lab note is accompanied by some interesting emoji-type diagrams, which you can see in this tweet from Changizi:

Color me intrigued. I know little about Tim Barber beyond the snippets available in his CV and the fact that he's smart enough to team up with Changizi. I regard Changizi as one of the best psychological theorists we have, and was happy to blurb his book, Harnessed: How Language and Music Mimicked Nature and Transformed Ape to Man (Benbella, 2011).

I'm wondering what their theory would tell me about Shakespeare, in particular, what it would tell me about Much Ado About Nothing, a comedy, Othello, a tragedy, and The Winter's Tale, a romance. With these plays Shakespeare has, in effect, presented us with something of an experiment. We have something we can call a Drama Engine. This Drama Engine has a variety of knobs and sliders we can use to set values for the parameters which will determine the matrix, the donnée, for our little play. One of these knobs operates at a very high level, simultaneously determining various aspects of the nascent drama. It is the position of this knob that seems to determine whether the play will be a comedy, a tragedy, or a romance. For the sake of argument let us call it G, for genre.

In each of these plays a man wrongly suspects his beloved of betraying him with another man. The following table shows some of the effects of G as they "trickle down" to the initial situation for each of our plays:

Shakespeare Triad

Along the left I've listed five dramatic functions which some character must play. In the columns for each play I've indicated the character that takes these functions. The point of the diagram as the we move from one genre to the next (in this order) one function seems to disappear. Othello has no mentor, but he has a deceiver; and Leontes has neither a mentor nor a deceiver. What I think is going on is that functions are, in effect, being absorbed into the protagonist. At the play's opening, Othello is senior enough in the world that he has no need of a mentor. Leontes is king in his world, so there is no one higher. As for being deceived about his wife, he does that to himself, no external agent required.

There's more. In the comedy the deception happens between betrothal and the marriage ceremony. In the tragedy the deception happens between the marriage ceremony (which apparently has been a secret one) and the consummation of the marriage. In the romance the marriage has been produced a six year old boy and, at the beginning of the play, the wife is pregnant with another child – who turns out to be a girl. There are others things in this complex as well, but this is enough to give you a feel for what's going on.

But how does our Drama Engine work? What parameters does it have, and how is G related to the others? I set out to answer that question some years ago and got no more than part way there: At the Edge of the Modern, or Why is Prospero Shakespeare's Greatest Creation? Part way is better than no way, and at least the issue is on the table. But it's unsatisfying.

Can VICTOR take us further along the path, perhaps even to its end? I note that "other awareness" is at the center of the model and our protagonists are deficient in the obverse of other awareness, self awareness.

Something else. The first chapter, "Quantitative Formalism", of Moretti et al., Canon/Archive: Studies in Quantitative Formalism, suggests that Shakespeare's genre space varies on two dimensions, not the one I've indicated in that table. I'd wanted to say a few words about that in my already too long review, but I ran out of time.

However, I do have an idea of where to look for that other dimension. Both the comedy and the romance, but not the tragedy, have two plots. The two plots run simultaneously in the comedy and successively in the tragedy. The Claudio/Hero plot in Much Ado About Nothing runs in parallel to the Beatrice/Benedick plot. The Beatrice/Benedick plot is very different in character from the Claudio/Hero plot, and that difference derives from the characters themselves, their temperaments. But I believe that the same psycho-socio dynamic drives both plots. In The Winter's Tale the tragic Leontes/Hermione plot is followed by the comic Perdita/Florizel plot, and that comedy redeems the apparent tragedy, transforming it into a romance.

VICTOR, it's your move.

No comments:

Post a Comment