“Assured and Believable”

I’m in a production of Agatha Christie’s ‘Murder on the Nile’ this month. I’m playing a psychopath on honeymoon with his very rich new wife. Anyone who knows their Christie will already have worked out how all of that goes...

When I was first cast as the sinister evildoer, my wife read the script and sagely agreed that the director’d cast it very well. I asked her what she meant by that and she said I knew exactly what she meant. And she’s brought this up most weeks since, usually at dinners with friends, where they laugh. If a touch uneasily…

Thinking about it, though, our director’s given me plenty of notes on projection, on where to stand and on not leaning too hard on the bar, but in our several months of rehearsals, he hasn’t yet given me much advice on how to play a superficially charming man who’s one disappointment away from doing something unforgivable. I can only assume he thinks I’ve got that part covered. Also, now I think about it, a number of the cast do tend to take their drinks with them when they leave the green room...

Might be best not to dwell, actually.

Excitingly, though, just this week, we’ve had a reviewer in. Obviously, I scrolled straight down to find where my own name was mentioned and was delighted to see that I, apparently, give an “assured and believable” performance as the murderer. Of course, my joy was brief as my beloved wife pointedly enquired if “believable” was really a word I’d want bandied around in this particular context.

That, however, wasn’t the only thing about the review that now feels a bit off.

Firstly, the whole thing seemed to spend an inordinate amount of time praising the set. Then praising the set again, with a few more adjectives. Then it moved on to the costumes and the lighting, before returning to another whole paragraph on how good the set, the costumes and the lighting had been. Now, don’t get me wrong, our set is utterly fantastic and the people who built it deserve every approving word. It’s just that the words kept coming back round. And round. And round again.

Secondly, the review gave one of the main parts to an actor who actually isn’t in the show, which was rotten luck for the actor who’s been playing it all along. We think, having done some digging, that the name used might’ve come off an old cast list on a deleted website page somewhere. Nevertheless, the correct name was, obviously, in the programme that the human beings were handed on the way in.

Basically, the more I’ve looked at that review, the more it’s struck me that there’s a decent chance it was written with AI. Clearly, I might be doing somebody a real disservice, but the more I’ve thought about it, the more the idea of an AI theatre review bothered me.

Peter Brook opens his ‘The Empty Space’ by saying “I can take any empty space and call it a bare stage. A man walks across this empty space whilst someone else is watching him, and this is all that is needed for an act of theatre to be engaged.” And the bit that matters there is the part about “someone else is watching him”. Especially, as Peggy Phelan argues, “Performance’s only life is in the present.” You see, our little show disappears every night and a well-written review will be one of the few things that we’ll have left when it’s gone.

A review is constructed for that reason. It’s a testimony whose whole claim is: I was in that room, on that night, and this is what I saw.

I think Ted Chiang puts it best when he explains how “art is something that results from making a lot of choices.” When you write, he says, you’re “making a choice about almost every word you type”. On the other hand, when you stick some hasty notes into a chatbot and ask it to sort a quick review out for you, “you are making very few choices”. The chatbot fills in the rest, and one of the main ways it does that is “to take an average of the choices that other writers have made, as represented by text found on the Internet; that average is equivalent to the least interesting choices possible, which is why A.I.-generated text is often really bland.”

And, well, that’s pretty much what our review is. It explains the old cast list (and I do hope the other chap enjoyed his notice), the three goes at lionising the set, and all that generic praise that could’ve been about any production of the play since 1946.

Chiang’s verdict about this sort of thing is the bleakest thing I’ve come across this week (and remember, I’ve spent the week playing a man who murders his wife). “The task that generative A.I. has been most successful at is lowering our expectations, both of the things we read and of ourselves when we write anything for others to read.”

It “reduces the amount of intention in the world.”

What does it say about the nature of the arts that a theatre critic could watch a whole live production and then, rather than respond to it themselves, merely feed their notes into AI? Brook only asked for one thing for theatre to exist. That “someone else is watching him”. So it makes me pretty sad that, as far as our show’s review is concerned, I’m not sure anybody really was…

Still, I’ll be keeping the “assured and believable” line. After all, it’s for a role my wife’s repeatedly suggested I’m well qualified for. And she didn’t need a chatbot to tell her that.

Sources

Peter Brook, The Empty Space (London: MacGibbon & Kee, 1968).

Peggy Phelan, Unmarked: The Politics of Performance (London: Routledge, 1993), p. 146.

Ted Chiang, ‘Why A.I. Isn’t Going to Make Art’, New Yorker, 31 August 2024.