PROGRAM THE WORLD

One of my favorite things about playing CIV VI (yes, still) is the underlying mechanism of “fuck around and find out.”

“Isn’t that just life?” you ask, immediately skeptical of today’s post?

Yes, obviously! But hear me out.

Every single choice you make in CIV VI has a butterfly effect later, shoving you into (often geographical) corners (without fresh water) you can’t possibly build or fight your way out of. Depending on your difficulty setting, your embryonic little village might be slaughtered almost immediately, long before resembling any sort of actual civilization, or much later, never having witnessed the glory of the Modern Era, much less the Information Era.

Let’s talk about the Information Era.

Even after (actual) thousands of hours playing this game, I am constantly at the mercy of seemingly innocuous decisions I made hundreds of turns ago, and now, on the cusp of a paradigm shift in scientific application, my people are starving/thirsty/under attack/without religion/the arts.

With every choice to pursue animal husbandry at the temporary expense of pottery, or a code of laws before craftsmanship, I am programming the world. At least in CIV VI, I do this intentionally, having memorized the tech and civics trees a long, long time ago.

I am creating my own destiny! Well, my people’s destiny — even if I don’t know what the end game will be yet, or the journey I’ll take to get there. In CIV VI, my only clear measurable outcome is to stay alive.

In real life, there’s still lots of poverty and famine on the map but way less than there used to be. Religion ebbs and flows depending on where you live, while modern-day fundamentalists spent the 2000’s fucking just blowing up many of our oldest world wonders. The tech tree is happening live, and moving faster, while the civics tree is kind of going backwards.

We are beholden to a world increasingly and inescapably programmed for us.

But is it?

Considering the relentless bullshit about how AI will take over the world and/or enslave us, the new LLM’s (ChatGPT, Claude, Gemini, Copilot, etc etc) don’t actually do that (yet).

In fact, they do that measurably less than the old-school machine learning that otherwise controls every aspect of the rest of our lives, from dating apps to mortgages and face recognition. Many of these have unsurprisingly turned out to be an ethical nightmare, coded from the start with our most profound biases, ten blue links shoved off screen to make room for ten VERY SIMILAR LOOKING links, bought and paid for.

While I’m increasingly aware of the technical potential and drawbacks of AI large language models, I am (again, at least for now) significantly more confident in my ability to discern the morality of them — the shoulds, vs the coulds.

So I choose to use them. I still have a choice to use them, that is.

I don’t us them to write, mostly for research, knowing they are imperfect. And I mostly use Claude, either in the flesh, or by way of Perplexity and as my default LLM there. And it’s great! Even if we’re not 100% sure why it’s great.

A significant reason I choose Perplexity is because it provides citations for its responses or summaries or whatever.

That mechanism alone makes me feel somewhat less horrific about the tool taking a prompt from me, pilfering the best of journalism, and just spitting out my own personal Wikipedia answer.

Ahh, but I am better than that. I can make myself feel even less horrific when I actually click through to the provided sources, to either fact check what Perplexity/Claude have given me, because #TRUSTNOONE, or to actually link back to the original sources when building my own on-going research or writing one of these little ditties.

Look at me! Still visiting the websites where the journalism was originally published, usually paying for a subscription if it’s a reliable source and one is on offer, or at the very least, sending them a few clicks.

After all of that, I am still (I think) standing on the (burnt-out) shoulders of the world’s (last remaining) journalists.

I am still doing the right thing.

…right?

…even with the insane energy consumption?

…even as this collection of hard and soft technologies threatens to “explode” the economy? Or make not make anyone any money at all? Or just make the same people unfathomably richer?

One way I justify my use of LLM’s is because I’m using these tools to understand what the hell is going on (or what the hell happened to get us here), and make reasonable guesses at what’s coming* and what you and all of us can do about it.

But I know how the internet works (logistically, I stopped coding when <iframes> were the shit).

And so, lurking in the cobwebs of my brain, is at least one major question:

I still have agency over this, right?

…right?

Not because of cultural FOMO, of course — cancelled plans are like (what I imagine) cocaine to be, for me. There is no trend I am not up for sleeping through.

When it comes to tools and software in particular, I am almost by rule:

1) Extremely loyal to what works, however ancient it may be and

2) Absolutely ruthless about migrating away from what doesn’t

I simply don’t have time to play with every new toy anymore, much less to migrate my 4849 (actual number) Apple Notes, which were once Evernote(s), to yet another fucking note system.

Entire baby LLM companies have raised billions, come and gone since I downloaded their app, stashed it in my “Testing” folder, and then forgot about it.

No, I mean “agency” in the sense that, of everything else online, I haven’t been “nudged” towards these tools (or caught in their collective web) in any way except that I tried them and they have proven to make an extremely time-consuming part of my job a little more expedient and maybe even more thorough.

I am not “all in” on them because my work process is extremely weird, because I am not willing to submit most of (or at least any more of) my data to them, and because, unlike when I got started in this job (?), I have become far more skeptical of shiny new things (see, autonomous cars) and infinitely more excited about destroying the bottlenecks that prevent equitable, affordable access to actually-clean air, water, food, shelter, education, wellness, and health care.

As promising/reckless as these tools may be, there’s a lot of known unknowns, and even unknown unknowns.

One thing we do know? The internet we’ve come to know and loathe is, to paraphrase Cory Doctorow, “enshittifying”.

ZIRP and preying on your lizard brains combined for about fifteen years to make you think your devices are actually listening to you (they are). They did, and they sold all that data, and they are completely addicted to that model, and it’s really ruined everything. For every fun new toy (TikTok), there is the inevitable lure to become a marketplace, because there is simply no other way to make money online.

But also, here’s the thing — and maybe you don’t remember it because you’re like me and vital information goes in one ear and out the other unless it is put directly into my 4850th Apple Note, but — you actually searched for sweatpants a little while back.

I know what you did.

You don’t remember because Shōgun was on at the same time and if you weren’t watching the subtitles, you’ll never know why the gardener died.

Maybe you didn’t search for these exact sweatpants in your feed right now, and maybe not on this device, but come on. It’s 2024, you might not know how the ad networks actually work, but you do know they are working, all of the time every second of every day, even when you’re not actively inputting data with your greasy fingers. You get what you get and you don’t get upset.

You’re getting an Instagram ad for stretchy sweatpants made of plastic or a banner ad for those same exact sweatpants on a completely random website you’ve never visited before because…you asked for it.

Of course you did that — and you did it because it was shown to you because of something else you clicked on before that! And now we’re veering dangerously into free will or at least internet free will territory (which at least maybe has an actual start point or genesis story), but the point is:

You did that. Take some accountability.

Because do you know what that means? It means you actually do have agency.

It’s so much less than you should have, or that you had before, but you do.

I don’t want to rehash all of this, you know it, you get it. If you drove near an abortion clinic, anyone can buy that data, etc. Old news. I’m just trying to reinforce for you that you drove there even knowing your location was being tracked and sold not only by your phone company, but by your car company, too, and probably because you had to.

But I’m here to explain that your little decisions do matter. Even online! They add up.

Your agency really fucking matters, today more than ever. Because journalism built America on geographic and demographic print ad monopolies, and then desperately pivoted that gamble to the internet.

But they lost, real bad, because Google and Facebook won.

Google and Facebook sucking up all the ad dollars for the rest of our lives seems — or at least, seemed, twelve months ago — inevitable and too huge to even begin to undo.

Not unlike last Friday night when you read about the record-breaking ocean temperatures and then stared at your plastic salad container you just diligently rinsed out, and then tried to remember if it’s one of the types of plastic that’s actually recyclable (it’s not), and you read the headline again, and you stared at back at the Nancy’s chopped salad container, back at the headline, back at the salad, and said actually aloud — what the fuck am I supposed to do about this?

I get it. But as I quoted in that unexpectedly popular post about ignoring my kids:

I don’t want to minimize famine, clearly, but you should know that you, yes you, have the agency and resources to feed or diaper at least one kid, or give another kid a bed net, or help one recently-fired journalist keep their newsletter going for another month.

Just by subscribing, and then upgrading, or even just clicking on one of their ads*, you are proving your agency, you are participating in and furthering the civics tree.

The internet is being torn apart, the facade crumbling, the latticework being stripped for training data, but when you volunteer for a clinical trial or literally just pay for a worthy Patreon, you are exercising so much power.

You are affecting the world, back. No, bitch, this is what I want to read. This is who I want to support.

We always say one of the tenets of our Membership is that by choosing to pay for it, you’re saying “Yes, I want to see more of this in the world.”

With every click on an actual news website, that’s what you’re doing.

With every dollar you spend on an actual book or whatever, with every link back from your Squarespace to a group who recruits young progressive candidates, or provides funding for abortions, you’re not only taking back your algorithms, not only programming the internet, you’re programming the world.

It’s the Alignment Problem but user-facing, and for good! Less hand-curated recommendations, more algorithms, this is old news.

I have this dear old friend who works at one of the two huge music streaming places and who is an extremely talented musician and producer in his own right. And I swear to god, however much pop music I listen to (and not just because of my kids), I still get my best music recommendations by simply asking him “What should I be listening to?”

I never have any idea what he’s going to send me back, but I trust him to give me some actually-hand-curated music options (I have other friends who I respect and admire, whose professional competency makes them one of the most trusted leaders in the free world, but whose taste in music I have abhorred for decades (Drew) — but this is neither here nor there).

We are at the point now where, yes, subscription-fatigue is a very real thing, and it’s all really exhausting and unaffordable, but at the same time, because of the democratization of subscriptions, you have more power than you’ve ever had before.

Look, one the one hand, the NYT didn’t just start printing money because it made a conscious choice to appeal to a…wider…audience, or because got some of that ad money back from Google and Facebook. It did it because they have always printed crosswords, and made them part of the subscription, and people happily paid for their crosswords, and so they added a couple more games, and then they bought Wordle, and then they snagged Wirecutter, and all along the Cooking site has been amazing, and now their games team is bigger than most of the remaining newsrooms in America.

I am telling you all this not because hey this is great, but because you did this. If no one played and paid for the crossword and Connections or Spelling Bee, the NYT couldn’t pay their games team (or honestly, probably, their news team).

If Amazon hadn’t capitalized on the 1990’s hell of maintaining your own servers to build and scale AWS, the store wouldn’t have made them any real money and they couldn’t have blown billions on Whole Foods or become a pharmacy or own more planes and trucks than FedEx or paid for Jeff Bezos’ steroids and all in all provide the opportunity for co-founder and author MacKenzie Scott to make dude billionaires look as selfish as they really are.

(Of course the Amazon store is now a completely untrustworthy advertising behemoth in itself, which is the point!)

On the other hand, if you hadn’t — understandably — spent the past fifteen years kind-of-expecting most of your news to be free, we wouldn’t be here either. I know.

Well, it’s time to rectify all of that, friends. It’s time to take the red pill and believe.

It’s time to forge each of our personal Apple Pay Face ID authorizations into systemic ones where for-profit newsrooms like The Atlantic and 404 can become profitable, and non-profit places like Capital B and The 19th and ProPublica and The Markup can change the model entirely.

When they can pay the bills, they can — theoretically — hold the most powerful among us to task.

When they can pay the bills, shit, when they thrive, the corrupt insurrectionist former president with a hundred felonies to his name cannot simply merge his social network with a shell company, make billions on paper overnight, with every fucking intention of selling to pay his half a billion in bond, just in time for the election he is running in, again.

Order somebody’s book? You’ve got the power. Knock on doors for a fucking school board candidate? Untold power.

Copy and paste obv for individual climate, food, health actions that, added together, make for that sweet sweet Compound Action.

What you pay attention to — like, actively look at, which the internet takes as “Got it, this guy wants to see more of DIY native landscaping, now, go go go” — is what we get more of.

There is one very-likely sliding doors future where Trump wins, where Moms for Liberty gets their shit together, where Medicare never gets real purchasing power, where new LLM’s become as intractable from our lives as rudimentary machine learning is today, but we can each and all take this moment that we have, living on the edge, our civilization struggling to get past the great filter, to program the world for what we want it to be.