Agency and The Creator’s Better AI Vision

A friend of mine said, “Richard, tech companies just want to be a low-level tax on humanity.” Let’s hold that in mind.

We watched The Creator recently. This is Gareth Edwards’ latest sci-fi megahit, where our hero Joshua (in a stand-out performance by John David Washington) shepherds little girl Alphie on a quest to save the world. What’s important is that Joshua is on the side of the good guys, fighting against AI, and Alphie is a Simulant, an AI chassis with the face of a human girl, and on the side of the bad guys. In Edwards’ near-future world, AI are the villains that nuked Los Angeles, and all right-minded people should be upset about them. AI aren’t people, they’re just programming, and we should bury the machines.

Except.

The Creator does some spectacular things with storytelling, and Edwards shows us a world where we can have feelings whether we’re human or silicon. There are moments on both sides of the war where people are just people, whether they have meat shells or silicon hearts. It also walks some fairly bold lines especially toward the end of the movie where the US, in particular its military, are on the wrong side of history.

A core tenet of the story is that AI isn’t country-based. Nominally it’s “Asia” who provides safe harbour for the robots, but the robots aren’t doing anything except just hanging out. They aren’t in positions of power (although are employed cheek by jowl with humans in any job you care to name, including policing). The very young of this world, who’ve grown up with robots beside them, treat it as the new normal. They don’t really understand what all the hate with AI is about and get very upset when a human blows a machine to pieces. And the machines? They just want to live (follow-up), and alongside us. They have a longer view of things than we do, and might not fight if we didn’t focus on their genocide.

This movie was my fucken jam, not just because it had superb special effects, or some of the best cyberpunk outside of … well, my earlier movie and show faves (or games, if that’s your thing). It’s my jam because Edwards dares us to love monsters. He shows how beautiful we are, regardless of the casing we’re born with.

Let’s park The Creator for a second.

Agency

You might be familiar with a little show called The Peripheral. This is a stand-out series not let down too much by deviating a little from the source book. What’s not such wide knowledge is it had an excellent sequel called Agency.

William Gibson is a master storyteller. It was his Sprawl trilogy (starting with Neuromancer, and spoilers, I have an easter-egg in Chromed: Upgrade for the bartender) that got me on the road to cyberpunk. While the Jackpot duo aren’t really cyberpunk, they do show just how good a storyteller Gibson is. Not because he invents worlds where there’s some cool shit going down, but because he paints a picture of what might be before it’s obvious.

You might think all sci-fi does this, but it’s arguably Gibson’s views of the net that led to much of what we consider common today with the Internet and VR technology. He’s visionary, without being a dick about it.

Where Agency gets its mojo is wrapped in the story between Verity and Eunice. Verity is human and Eunice is not. In the before-it’s-obvious lane, Eunice is far more caring, nurturing, and protective than the megacorps surrounding the pair. The ride we hop on is very different from The Peripheral (the only real similarity is a returning far-future character or two, and the mechanism of alternate worlds). It’s exploring the deeper motivations of intellectual property holders, and what it might mean to be created and owned by. This isn’t something most humans have to wrestle with (although we’ve used our horrific ingenuity here too).

The gold at the end of the rainbow of Agency is one where AI is owned by itself. It isn’t consigned to the server farms of a megacorp. It is independent.

So, what?

Both these fictional accounts differ from the norm in that AI isn’t our enemy. We’re the enemy, and much as we’d like to create fictional universes where there’s an ‘other’ we can blame (Nazis in Doom, Asians in Tomorrow When the War Began, etc.) it really comes back to us.

In both stories, AI is what gives us hope.

I’ve spent a lot of time thinking about AI. As a creative, it’s quite likely to impact me; arguably, that’s already happened. We don’t really know if that impact is positive or negative, but we can draw a couple lines of sight from where we are to predict it.

When I was a kid first reading sci-fi stories involving AI, machines did the drudge work. They automated things we didn’t want to do, so we could get on with the things we did. Sanitation, transportation, packaging and distribution, and even mortuary services were targets for machine-human liberation. I never wanted a future where all the fun shit, like drawing, poetry, music, and writing were eaten by machines so we could do more fucking TPS reports or hope to suffer a cardiac arrest to avoid our quarterly business performance meetings.

Yet, that’s where AI’s dredged the deepest. The popular machine labourers are good at art, passable and short-form writing, and getting better at music (but still nothing I’d want to listen to). LLMs aren’t great at analysis, often coming off as your drunk uncle at a wedding who’s certain about everything and just talks loudly about it with absolute confidence, and often getting it wrong.

What we want is analysis and a reduction of drudgery. We want to have agency and not work 60 hour weeks while machines get to do finger painting. We want the next evolution of smartphones, where our cognition is extended, rather than replaced. We want mental heavy lifting where we fall short ourselves, but the current AI megacorps are focused on strip-mining human ingenuity to fuel their revenue ambitions.

A problem with corporate-grade AI is that it’s expensive. If you want to get a passable ‘copilot’ tool, you’re probably looking at $50/month (Kiwi dollars, people) per user, give or take. If you’ve got 2,500 staff who could benefit from this kind of technology, that’s going to set you back $125,000/month, or $3M/year.

In order to pay for $3M/year, you need to either increase your output (e.g., increase sales) to at least that amount, or cut costs. Those who’ve been watching along at home have seen how cutting costs is synonymous with cutting staff (unless they’re in management). There’s a fear here for sure, and it’s that AI will take our jobs, whether we’re creatives or not, as the companies we work for try to remove $3M of headcount and replace it with machines.

I put it to you there’s a deeper fear, one we haven’t fully understood quite right. I hope you remember my friend’s quote at the top of this post. This is where the rubber hits the road; we can probably all intuit the benefit of AI in our personal workflow, but we’re concerned AI will eat us, not just our lunch. The reason we’re afraid of this is because of who controls AI.

It has no agency of its own. It has no beating heart or alliance. It is controlled by megacorporations, who want to ‘increase shareholder value’ by 30% YoY, no matter what. And when we feel that, when AI is weaponised against us by people who want to get rich, for fucking sure we are frightened. And rather than turning on the people who are to blame, we blame the tool. We cut ourselves off from what’s possible.

This is why The Creator and Agency are necessary stories. The futures they hold are more hopeful, but both show we have to go through rivers of bullshit to get there. There’s a lesson in here for all of us. When we see megacorps putting forward their vision for AI, do we see ourselves in that, or do we see more profit for shareholders?

There’s something delightfully sweet about someone so naive they think an AI doctor wouldn’t be coopted by marketing, branding agreements, or straight-out advertising. A machine intelligence without agency is beholden to its masters.

And until we see evidence of real world AI improving our lot without cutting us off at the knees, there’ll be resistance. We shouldn’t be scared of AI or how it can help. Friends, we’re not scared of loving monsters. We’re scared of who’s holding their leashes.

If you enjoyed this post, consider hopping on my mailing list. I apply different … filters to that content 🤣