The Serpent in the Grove

I was laughing about the AI short story being called “A Serpent in the Grove”. I suppose the idea that this title might suggest a writer of originality and inventiveness just struck me as very funny. And in general I keep hearing all these stories from friends working in “the arts” about how AI is popping up in their working environments. They’re always really funny.

A friend who works in film development told me her boss rushed in and asked her to “make AI look at the scripts” in a sort of panicked tone one day. “What for?” she asked. “Oh god! Just see what it can do!” said the boss. Many stories I hear like this. An esteemed, middle aged, translator was late on sending something in. He was asked for a sample, and sent one. A few sentences read strangely and so someone just put the text into Google translate and discovered this was exactly what he had done, with no added changes (I know). Publicist managers using AI to write emails and getting angry responses because the tone is so over familiar or strange.

It is always important to remember, I think, that a manager in this kind of job is a person in their 40s to 50s, who probably hated maths and STEM at school in general, studied English at university, and has done their best to keep their relationship with “the computer” to “sending emails” since computers became a major part of everyone’s lives. Which was not actually so long ago. I think it is fair to say computers have been everywhere since around the early 2000s. (Americans may put this earlier.)

In fact I know people who work with managers who don’t even send any emails. They dictate everything they wish to send into a tape and their secretary transcribes it all. Yes, this is still happening. If you asked this person, what is the future of AI? I am sure they would give a colourful and interesting response. But what would this be based on? Would it have much of a relationship with reality?

Whenever AI comes up in a conversation with people who don’t know much about it they always have these totally crazy and fantastical propositions for what it might be able to do. Genuinely brilliant feats of imagination. Or even what it might currently be doing, without any of us knowing. People are always saying, in dark tones, that it does all kinds of things nobody knows about. I have had conversations in which someone, with a completely straight face, has said: It might give language to all the pets! (???)

Another one: It might be able to fly around the air (??) and record all our conversations to perfect every accent in the world.

This is actually one of the things I find funniest about AI, that its relationship with human imagination is sort of paradoxical. In the sense that using it to do a task for you which you could do yourself, writing a bad short story, say, necessarily involves an outsourcing of your imagination. (And probably a dampening. If you keep using it to do things, you will probably stop being able to do them as well). Whereas talking and thinking about it seems to fuel people’s imaginations like almost nothing else I have seen in my life.

Again, paradoxically, I have always thought that the powerful imaginary properties of AI (by which I mean the properties humans invent and assign to AI, rather than its own powerful imagination, which I do not believe exists) are what make it so exciting and, correspondingly, so economically valuable.

There has always been an, I think deliberate, mystification of AI on the part of people who stand to benefit financially from its widespread implementation. This mystification works well both as a marketing strategy for AI (think of how the crazed coverage of it relates to attracting investment) and also as a defense whenever AI cannot do the things it has been marketed as doing. Much of the promise of AI relates, not to its current form, but to the potential brilliance of a theoretical future version of it. By this logic, whatever it can or can’t do right now is kind of irrelevant.

I don’t think it is the Emperor’s New Clothes, exactly. It does do some things. But I don’t think something branded as “an assistant tool which can help proofread a legal document” would garner anything like the attention or investment which “a magic black box which might accelerate itself beyond the scope of human intelligence and take over the world” has managed to.

It’s best to keep a sense of humour about these things I think. Even if AI does take over the world and ruin it, you’d at least have that.

A human could write a funny short story about all of this. I don’t think an AI could. But perhaps in the future, when it has learned how to fly and finished teaching the animals how to speak. Anything is possible.

Till next time xxxx