Don't let your anger cloud your judgement
AI needs to be understood, not shouted down
Don’t allow your anger about AI’s disruption to colour your understanding of what it is. It doesn’t serve you. You can be angry, who am I to tell you otherwise, but try to dispassionately to understand what is coming. Anger about trains is of no use when there’s one speeding towards you on the track.
So why do I say this?
I’m a conflict averse investigative reporter. It’s an odd combination, because my work forces me to confront people. I don’t really like that part of it, but it’s somewhat inevitable given the nature of the work.
When pressed, I tend to err on the side of caution. Regularly, I hear the phrase “Do I really know enough about this to comment?” echo through my mind. “This is perhaps someone else’s domain…”
Doubt plagues me.
It forces me to look harder at what I’m writing about. In the case of AI, it forced me to spend three years learning how AI agents actually work. I built them. Lots of them in fact. I ended up writing so much code that I became fully lost in that world. “Am I a writer at all?” I wondered. “Am I a coder now?”
Perhaps.
Whatever the truth may be, I now (just about) believe I can write about AI, within a narrow range, because I understand it just about enough to translate my understanding into english. English. The language used to program humans.
Having looped endlessly through the language of machines, I’d been away from writing ‘english’ for a very long time. I’d lost touch with the most endearing feature of the language; in english you can write stuff that doesn’t do anything at all. You can use english to write beautiful and convincing programs that are completely isolated from reality.
I started noticing this kind of english on substack. It can be great writing, and it draws you in, but it doesn’t really wrestle with its own subject matter. It doesn’t really interrogate its own premise. I’ve read AI articles from substack writers with huge audiences, and as I’m reading, my own self doubt manifests on their behalf. I think “I’m not sure this is good advice?!”
The target for this writing seems clear; you’re disgruntled about AI, and these words are going to speak directly to that anger. It reminds me of a great line from one my favourite comedians, Stuart Lee. Recounting an imaginary conversation someone has about his comedy, Lee says
“Did you watch his show...?
...yeah.
Was it funny?
No. But I agreed the fuck out of it.”
As the audience applauds, Stuart brilliantly continues his satire. “I’m not interested in laughs. What I want is a temporary liberal mass consensus that dissolves on contact with air”
Stuart is an expert in his craft. He knows the exact ways in which you’re already disgruntled about the world, and he speaks directly, masterfully to that existing despair. This is satire. It works when you accept, wholesale, the presupposition at the heart of the satire. It is funny. A shared experience to illuminate the absurdity of an idea or belief is top tier entertainment, but it’s not the vehicle to get in when you’re looking to understand something.
I’ve seen articles on substack about AI that veer very close to the satirical. Beautifully crafted, a bubbling anger at their heart, competently written, and desperate to point out all the absurdities and excesses of the industry (of which there are many). They’re great articles, but they’re nearly useless in helping you understand the train on the tracks.
One thing I’ve seen a couple of times is willfully wishing the technology away. It’s all silly. It won’t work. The people investing in it are egomaniacs. It wastes water, it wastes energy, it kills jobs. It’s true but incomplete. As the anger about AI grows, I think the audience for that anger will grow with it.
Be angry if you must, but seek out perspectives that aim to faithfully chart a course through whatever is beyond the horizon. Screaming at every wave, as it crashes over the bow, isn’t going to cut it.
When I’ve appeased my own self doubt, I’ll be sharing my thoughts here.
I like this write-up, especially the observation about the huge volume of “empty” writing, most recently obvious on Substack.
But not understanding AI is the crux of the problem. Most of us have a vague idea about how it works, which implies a tiny minority does. That minority will have nearly unlimited power, including the power to authorize different levels of operability to different levels of people. Not to mention it’s a resource hog (but you did mention that). But thanks again for a thoughtful article, definitely not “empty.”
Hi Phil, thanks for this post. It is important to be self-critical and to question one's assumptions and, importantly also the deeper reasons driving one's opposition or advocacy for an idea. I have two questions: 1) at what point is critique and rejection of the way AI is being implemented in society *not* driven by anger? Is there a legitimate and reasonable strong position against AI in your perspective? 2) Does your impulse to self-doubt also apply to your defence of AI?