i feel dizzy (again)
y’all wanted a personal essay? okay, here’s a personal essay
I saw someone say that these are supposed to be embarrassing or they’re not worth writing. I think they have a point, honesty is often uncomfortable. Unfortunately, I am incapable of defacing myself for the sake of a few dozen paragraphs, I have appearances to maintain.
This one isn’t a regular “let me explain a part of the world, as I see it” post either. It’s an essay that pretends to be about technological change and its consequences, but really it’s just about me, and what I think and do and feel. This is a genre I never really want to write, let alone publish. But I ended up finishing most of it in one sitting, and that hasn’t happened in a very, very long time.
It also came out as more pessimistic than I really am. I should therefore clarify whole essay is an emotional reaction and should not be read as financial/life/epistemic advice.
Oh, there’s also a suggested reading soundtrack for it1.
I
“Professions” as you understand them will cease to exist in the medium term future - Manav Ponnekanti, 10 minutes ago.
I started writing this because I caught myself in a mood that I’d noticed thrice before. The lazy description is “existential”, a more literal one is “woah, alright, okay, steady now. I’m dizzy.”
The first time it happened was when the Dall-E 2 demos dropped, the second time was after they gave GPT eyes, the third happened when I saw it turn sketches into games. And the one that led to this essay was the video generation model that has somehow solved for rudimentary physics simulation and object permanence.
I was three hours away from publishing, and it happened AGAIN. This time, it’s finally robots.
These are “merely” impressive demos of capabilities I really thought would have some kind of hard limits. These are “just” high-dimensional predictions based off of an impressive (though incomplete) model of reality. These are “simply” a bunch of clever hacks and large matrices. A shiny demo isn’t everything.
It is still vertiginous.
The vertigo might just be because my beliefs about information and reality were proved very, very wrong. Everything that I thought was “incompressible” proved to be a matter of scale, it’s more than a little ridiculous.
There’s only so many reminders of how uncalibrated I am that I can absorb before feeling like I’m standing on ground that I shouldn’t trust. These days, my default assumption is “you can just compress good data enough, and it will spit out a good model”, which helps with the dizziness, but doesn’t change the fact (or the feeling) that the ground is shifting.
II
“So what do we do? I don’t want to be the guy training the models”- Telegram message from Nihal in reaction to Gemini 1.5, 16th February.
Same, I don’t care enough about GPU code or training runs. Nihal is a lawyer who does research at a legal policy and governance think-tank, he enjoys reading and thinking about legal policy and governance more than any reasonable person would. I wake up and design really simple software, and the occasional beautiful website. I write my HTML by hand, because I do this as art and not to create shareholder value. This is what I want to do, am blessed enough to sometimes get paid to do, and hope I can do it for a while longer.
I have apps I want to make. They’re not particularly impressive ones, almost all of them are about putting text and images onto the internet. I feel rushed, because part of me still feels like there isn’t long before they become as trivial as three-digit multiplication.
It’s not over, of course. It’s very far from over. We work in pure simulacrum: words and ideas. Of course we’re intimidated by the word-generator. All our friends are upper-middle class (as defined by having a college degree) and want to get jobs that keep them there. Of course we’re worried about white collar workers. We would both like to make undeserved amounts of money by just thinking about things. Of course we fear digital automation.
We were both overreacting, but it’s still a real reaction. You’re going to see a lot more of it soon enough.
Most people don’t know how far along we are yet. I avoid telling them, it feels like asking for panic. And part of me believes that their panic will be unjustified, only I know how to panic properly. Only I have enough context, only I know just enough about how things work that I can see all the correct reasons to be the correct amounts of optimistic and worried. They wouldn’t get it.
III
“Why are they teaching us linear regression when we have computers that can talk?”– A friend, after I showed him what GPT 3.5 could do.
The truth is, even I don’t know who really should be worried right now. The people who are most vocally concerned are the ones who spend their time creating things that live on screens: programmers, writers, graphic designers, and video editors. Everyone else can’t be bothered to care, they’re too busy studying for medical exams and building plumbing systems. Moravec hasn’t been proved wrong yet.
If you’re in college studying for a computer science degree, you’re going to be afraid. Part of me finds this beautiful, considering all the jokes that were made at the expense of humanities grads. Part of me is sad that we might lose one of the easiest, risk-free paths to wealth that the world has ever known. Who’s to say if economic mobility is worth trading for a chance to see the nerds fret a little?
Writers are also worried, but not really. We still don’t have AI that will go out and interview people. We don’t have AI that has a Twitter account and friends in the right places. Heck, we don’t even have AI that is capable of having a strong opinion. The competition for most jobs is still other people.
If you’ve already graduated from “thing-doer” to coordinator/planner, congratulations! You’re probably making a lot of money, and will continue to do so. If you have a bullshit job, congratulations! Nobody knows why you’re here but they’ll probably keep you around as long as the economy keeps growing. If you have a job that involves being responsible for managing liability…good luck…? Computers cannot be held accountable yet, so just hope they don’t find a cheaper human to replace you.
But what if you like doing things? What if your ethic is Protestant, your dedication is to the craft, and you just can’t live without the need to be useful?
What if you’re like the programmers who are not mad about the possibility of losing their jobs, as much as they are sad that they can’t spend the rest of their career typing ASCII characters onto a screen while thinking hard? They don’t care about “prompting” an opaque model into creating business value, that’s not why they got into this.
You want problems. You want to feel smart about solving them, this is your jam. Some of you are actually sad that you cannot “grind” anymore, that things will be made too easy. (The grind is dead, long live the grind). I think this is as real a sadness as any other.
IV
“Yeah, but this has always happened. New technology comes along, people have to adapt.”
“True. I guess I could do more to use these tools.”– My in-laws talking, during a belated New Years barbecue, 6th January.
They’re not wrong about the history, but there’s a chance that this is different, and they’re completely unprepared. I think a part of them knows this, because I also heard the words “little bit of Python” and “more skills”. Phrases that are associated with bootcamp advertisements and the rest of the industry of grift. There is doubt creeping in here, even if they want to believe otherwise.
I have no advice here, nothing that I want you to “take away” from this. If you want me to tell you how I think about the economy as a whole, and what you should do with your life, ask me to write a separate essay about that (don’t actually do this, I have nothing to say except “get good at doing things and convince people to pay you”). And that might just be outdated.
For what it’s worth, I don’t think we’re going to see a sudden instance of crazy job losses across the entire economy. I’m not yet worried about the end of economic mobility and the permanent cementing of the upper class. I’m mostly concerned with slow but steady shifts that some people are going to be hurt by. Some of this hurt is going to be due to questions of meaning, and uncomfortableness with uncertainty. Losing your job is pretty bad, but even anxiety is a pain as real as any other.
V
“I’ve spent most of the last two days lying on my bed staring at the ceiling.”- Twitter DM (as I remember it), early 2022.
A few of you will remember that I tried running a small micro-grants program back when I was… 19. That’s crazy, I had no right to try doing something like that. I was supposed to write an update on the what and why of the whole thing when I shut it down, but that has stayed in my drafts too. Maybe I’ll get around to publishing it someday.
Anyway, I mention this because, near the end of the active part of this project, I sent out a small grant to a young artist in Indonesia. All she wanted was the money to buy an iPad (secondhand) so she could start doing commissions. This was around a month or so before Midjourney and Stable Diffusion made their announcements.
The quote above was how I remember the DM she sent me a week after those “tools” launched.
I don’t think it’s the end of her life, but I understand how something like that can fuck you up.
This is my current phone wallpaper. It’s a screenshot of Simon Bailly’s Instagram feed. There’s nothing particularly technically difficult happening here, but I really like how she uses colours. They’re not particularly sophisticated or subtle, you could even describe them as “borderline garish”, but I think they’re really neat. So does the New Yorker, who has hired her in the past. Somehow, I don’t think Midjourney will stop them from doing it again in the future.
I could use this to make a point about how “art is about having a point of view”, and how large companies will probably always be hiring for bespoke services. But you already know this. I only brought this up so I could have a cool thumbnail image for this post.
I play the guitar (poorly). I already know we’ll have a model capable of turning “indie fingerstyle open E tuning instrumental” into an audio clip I can’t hope to match (at the moment). I don’t really care, playing an instrument and songwriting is cool as hell. Always will be. I might spend my next payment on a learner’s drum kit.
VI
“…you should publish more, you have some bangers.”– Jihad Esmail, on a video call where I was designing his new site. December 20th, 2023.
I have a writing voice and I don’t like it very much.
It’s too ‘internet’ — there is no weight to it and my sentences collapse into snappy, faux-honest confessionals far too often. There is a rhythm to it, I’ll admit; but it’s both predictable and not particularly musical. I have no doubt you could get GPT to imitate it, and write about the things I generally talk about (economics, ideas, the internet).
But other people seem to like it. I’ve had half a dozen people tell me I should write copy, and more than a few compliments re. the essays I have here (y’all really say the nicest things). Apparently “taste” is the differentiator and I don’t have to worry about GPT stealing my readers because what I do is one of a kind by definition. It’s nice to know this writing thing is still worth doing.
I can also boast of an $80 subscription from Substack’s patron saint, Marc Andreessen himself. Last year he read The Problem of Too Much Money and thought it was good enough to become my first (and only) subscriber. But it’s an annual subscription and he’s got almost nothing for it so far.
So I’m writing a post for him. It’s about learned helplessness and insurance, the societal trend towards “shifting the blame” culture, and entire industries that are built on the backs of it. It’s called “Covering Your Ass-as-a-service” and I will try to complete it by the end of the month. Hopefully he sees it and decides his subscription was worth it.
Maybe you can sign up too. Unlike a subscription, it won’t cost you anything. Yeah, this was a teaser and like + subscribe request. Sorry, couldn’t help it.
VII
“We are going to fuck up the internet and all that is true and beautiful.”– Companies working on content generation and summarisation, probably.
I am now a self-appointed steward of the internet, by the way. I just decided one day that I can do this, I can do things that make the internet a better place to be on2. It’s kinda like being a good citizen, you keep the sidewalks clean and share meals with your neighbours. Except I make reading lists and maps and tools that help people publish things online.
So of course I have opinions on copyright and intellectual property. I wish every SEO-hacking content generator a very pleasant “rot in hell”. I’m going to go up against the search products that chew up pages to spit out summaries with my own solution for making webpage discovery easier.
I have sympathy for people whose stuff was used as training data without consent, I think the copyright litigation is too little, too late, but a valid reaction. I’m especially sad about visual artists, they were the most fucked over.
I have sympathy for the people on Upwork. They don’t have the most enviable lives, but they had a brief moment where they could strike out on their own, free of the mess of their national job markets. I hope it doesn’t get too bad for them.
Eventually, I imagine I will have sympathy for more and more people. The back office workers and customer service, the admins, the translators and ops teams, all the new grads who need a foot in the door. Maybe even truck drivers and service workers, someday. Funnily enough, I think the lawyers will stick around for a while.
And, perhaps foolishly, I believe that I’ll be alright.
VIII
“No, no plan really. The world is…changing so fast. I try not to think about it too much.”
“Very fast! I don’t know how you kids do this.”– Me and my grandmother, early last year.
I think having lived through a bunch of major exponential curves might have given her some sort of perspective here. I had just graduated, and admitted I had no definite plan for what I wanted to do in 5 years from now. And I think she understood what I meant.
Much of the world is shaped by people plugging away on their dream despite change/uncertainty; bending the world to their will in a way that makes for an epic biography. But many of the great failures were failures to adapt to change, the world bent and they broke. So all I have is an anti-plan. Instead of charting out what I should/shouldn’t do over multi-year timelines, I just do the things and hope to not die.
I was also lying to her. Of course I think too much. I have to, nobody else is going to do it for me. People keep telling me it’s enough to work on things you love. I believe them, but not completely; nobody knows where this goes. And nobody wants to admit that it might just be over. Everyone has their own reasoning for why things will go well, we’re stubbornly optimistic.
But it isn’t the end of the world. Yet. This is still “just” a factor that I must plug into my analysis of the world now, a caveat to all advice that I give, a rogue variable to keep a wary eye on.
My career advice used to be “find something you like, and can spend lots of time on, people will pay for it once you get good enough”. I’ve now had to shorten it to “find something you like”. I can no longer make promises of what the payoff will look like, or that a “career” or “profession” is worth having for most people.
IX
“It’s always awkward to explain how I owe many of the good things in my life to Twitter”- Me, when I’m trying to explain how I owe many of the good things in my life to Twitter.
Here’s a secret that isn’t a secret, just something of a rarely-mentioned fact:
I’m using my friends as insurance reassurance.
They are people who know how to work and think. They also have hearts of gold. This is really useful, sometimes in a “their help is what puts money in my bank account” way, as well as in the “if I lose both my hands I don’t have to worry about being homeless“ way.
And now it’s in the “if I don’t make it, at least I’ll know people who have” way. I’m not sure if most people have this privilege, or explicitly acknowledge it when they do, but it’s a big part of how I think about life.
I try to be easy to help, I do what I can to be helpful so people know my intentions aren’t parasitic. Either I’m misreading everyone, or this is all it takes to make people be fond of you.
Here’s another secret fact: It’s really hard to fall out of your socioeconomic class. There are hundreds of small ways in which you’re supported, cajoled, and bullied into sticking around through opportunities and systems that are part of the environment that you’re born into. This is an undeserved blessing, the kind that traps you within a very specific life trajectory. But it’s a blessing nonetheless.
X
“Pick a number between 6 and 10”
“10”
“ffs”- Telegram conversation, me and Aarya. 17th February.
I was typing this essay out on my phone at 5:50AM. I had 5 sections written out in under half an hour3, I wasn’t sure how many more I should try to add. Thank you Aarya for deciding for me.
I intended for this to be a really comprehensive compendium of everything I had to say on this topic. But it turns out 10 sections isn’t enough, there’s always more to say about everything.
You will notice that I haven’t actually tried to give out actual object-level advice. I’m not going to tell you to “get in while there’s still time” because that’s not what I’m doing myself. I haven’t laid out my own optimistic scenarios (there are several), since I don’t trust my own estimates enough to thrust them onto other people.
I’ve also very conveniently skipped over what happens once we’re past the “AI can’t do X yet” phase, and enter a time where we can perfectly substitute our human labor with more compute. I had to, because that’s a whole other essay on its own. An essay that I am incapable of writing, because if/when that happens, all bets are off.
Jihad Esmail has very flatteringly described me as an Internet Craftsman
👏