Y’all are making yourselves crazy about AI writing.

It’s true the internet is filling up with oceans of meritless text, but that’s been the state of the internet since well before AI came along. Folks are making themselves mental out there looking for writing that smells like it had a robot’s help, but is it really worth all the fuss?

has declared himself AI sheriff. As far as I can tell, he does it by gut. Meanwhile, brought in her own robot to sniff out robots.

But it is just very hard for me to care about this latest “threat” to the very important business of writing as a profession. Since the 90s, this body of work has been shot in both hands, both feet, the left arm and the right leg.

If someone has their gun out and taken aim at our left leg can we really be that concerned at this point? Why not just aim it at our head and get it over with?

Writing is no big deal

Before AI, there was a vast ocean of writing produced by hand every day that had nothing special about it. Though completely composed without the aid of LLMs, these compositions still had no notable destiny. They served no purpose beyond their fleeting utility, at best as datapoints in history.

Most writing matters no more than the email you just sent your boss explaining why you have to be late next Friday. It is written quickly, read once, considered briefly and forgotten forever.

Just as the same paint can be used to compose a gloppy mess of a still life in some college junior’s elective art course as might have been used to paint Van Gogh’s Field with irises at Arles.

Though a category of stuff might occasionally be assembled in a special way, that doesn’t mean we need to be precious about the stuff itself.

There will never be an edition of LinkedIn for Dummies that will join the Western canon. No memos about properly filling out time sheets using the new HR software will be preserved for their humane poetry. No account of the victory of the West Ralph girls JV basketball team over Chesterton will be savored by future generations.

So much writing can easily be written by AI, and it probably might as well be.1

Writing does not inherently matter.

Writing is mad. It is stringing together a series of lossy symbols in an attempt to recreate moments from a wildly complex and ever changing universe. Writing never even comes close to achieving its intention. And yet we keep doing it. We love it. Even though it’s almost always mostly wrong, to some of us it feels so right, because our minds are wired to operate using these same inept graphical representations that you are perusing here.

And sometimes there is writing that matters. When Martin Luther King, Jr. was in jail and decided to write a letter, that mattered. That was a piece of writing we couldn’t do without.

But Paul Revere warned his neighbors that the British were coming on the back of a horse, but that doesn’t make all the hundreds of thousands of other horse rides any less pedestrian.

The fact that the rare piece of writing matters very much does not make any old bit of writing any more important.

Support writers

In the age of TikTok, it’s hard for me to care that much if a new ocean of slop comes to the web. The slop was up to my nose before Sam Altman ever got to work, frankly.

Would I like to live in a new era of appreciation for the handmade? Of course. But it’s hard for me to believe that hating things will manifests it. It seems obvious to me that the era of appreciating the work of humans can only be borne of appreciating humans — not from vilifying the inhuman.

And yet writers are out there with long rhetorical knives, gleefully hacking and slashing at anyone who appears to defect from the presumed consensus that we must all be sure to produce every single sentence with our own precious, brilliant, irreplicable fingers.

And what bothers me most is that anyone — human or computer — can be so arrogant2 as to believe that they can actually really know what is or isn’t AI generated, when all AI does is imitate us, and the main thing we do (frankly) is imitate each other.

If you go looking for AI slop, you will surely find it — even if it’s not really there. Slop policing is digital phrenology.

The house of literature has been leaking, its windowpanes have been rotted and the siding has been falling off all around us ever since Craigslist went online. I’m not sure literature actually can be saved, but writers attacking other writers isn’t going to save it.

I know that.

If you care at all about writing, whether as an art form, as a mode of communication or as a business model that’s swirling down the drain of sustainability, let me tell you something you can do about it that will mean 1,000 times more, every time, than any instance of accusing someone else of using the wrong tool:

You can be nice about someone else who is doing a good job.

• No, really. All of them.

Some lost more. Some lost less. But they all lost money.

Late last week, published an interview with former CEO of Binance, CZ.1 CZ is one of the richest people in the world, almost certainly the richest person in crypto, an ex-con, pardoned by Trump and creator of the Giggle Academy — a free online learning platform.

The occasion for the interview was CZ’s new book, his memoir, Freedom of Money.2

Early in the interview, brings up the fact that SBF got a 25 year sentence for his multi-billion fraud conviction. “Do you think the law was fair to Sam Bankman-Fried, or do you think it went too far?” she asks.

CZ dodges, mostly, just pointing out that SBF didn’t tell people he was using customer deposits to make investments, which meant that, in a liquidity crunch, it wouldn’t have everyone’s assets.3

“I think some people wonder about it because my understanding is no one lost money,” she says, “So, was it overreach?”

Again, CZ dodges on judging it, but he does quibble with the notion that no one lost money. He did not, however, go far enough. I will.

Let’s increase the understanding — once again. Let’s clarify how literally every FTX customer really did lose money.

How people lost money

Most people lost money because they sold their claims before the bankruptcy resolved. One estimate puts the sales at 70% of claims.

Final payments at the end of the bankruptcy went to distressed debt investors, not the actual investors. Distressed debt investors take the risk of losing out on a bankruptcy by buying claims from claimants before the process resolved.

So maybe FTX users sold early and got 15¢ on the dollar, or maybe they sold later and got 60¢. Whatever. They lost money.

• And, look, you can say: Well, that was their call but:

  • A. It doesn’t matter. If the question is: Did most people lose money? Yes, they lost money.

  • B. Was it? It was SBF’s call to use their money to make investments without disclosing it. They thought their assets would always be there. They couldn’t afford to wait.

Second, even if they didn’t sell, they lost money two other obvious ways: opportunity cost and inflation. They had to go 2.5 years without money that they thought they had in 2022. Who knows how that money might have been useful to people in that intervening time? But not having it, as any economist would tell you, is a real material loss.

At a bare minimum, they could have tucked it into U.S. Treasuries. Recent years have yielded the most generous payout period on those instruments in most people’s lifetimes. That’s real.

And then obviously the value of a dollar has been taking an especially bad beating over the last couple years. So getting back the 2022 value isn’t really the 2022 value any longer.

Third, they didn’t get their crypto back. They got the 2022 dollar value of their crypto deposits.

In the U.S., bankruptcies pay out in the dollar value of whatever was lost. They pay out in dollar value at the time of the bankruptcy. Mostly this works out well and it’s less of a logistical nightmare. But there was a huge cost because this bankruptcy took place as the crypto market turned around again.

• Bitcoin is worth about 3X what it was in Nov. 2022 today.

• Solana, one of the favorite coins of the FTX crowd, is worth twice what it was in 2022.

• Tron is the world’s second-hottest platform for stablecoins. Its coin, TRX, is up 5X since then.

So that’s a real loss for those users. Many people hold coins longterm because they have conviction in them. FTX users didn’t get the opportunity to make that choice.

SBF was very bad

As I wrote at the end of my book about SBF/FTX, the worst people in history are the ones who do the wrong thing for the right reason.

SBF thought he was smarter, more moral and more discriminating than everyone else. For that reason, he was able to justify taking mad risks with other people’s money, because he believed the empire he would build would do so much good for the world that it would justify any past malfeasance.

Like so many before him, he talked himself into becoming a monster.

So, let me go ahead and answer Siewert’s question about whether or not the law went too far: Yes, SBF deserved his sentence.

He’s one of the worst white collar criminals in history. He robbed people of billions of dollars.

To paraphrase Judge Kaplan at the sentencing: It doesn’t matter if you steal $1,000 from someone, bet it all on Craps at the casino, quadruple your money and go back to split your winnings with the victim. What matters is the fact that you robbed them.

• It was one of history’s all time biggest robberies. It merits a gigantic sentence.

For those of us who watched the trial, every day was a new jaw dropping revelation of just how terribly he behaved.

And if that wasn’t enough to convince you, the flagrant way he’s demonstrated a complete lack of remorse since being sent away should do it.

I don’t think we’ll ever have a full picture of how far SBF’s grand narcissism went or how far it could have gone, but I’m glad the good people of this world will never again be subject to SBF’s attempts to save us all.

Back in the bad old days of crypto trading, when there were hardly enough buyers and sellers out there to repopulate an American ghost town, a once major exchange called Poloniex had a feature called the “troll box.” In the troll box, users could discuss their trades and their perspectives on trading with each other.

As one redditor described it, “It lo…

Read more

I apologize to staff of the White House.

For a while now, I’ve intimated that prediction markets inevitably engender public corruption. I have suggested that members of the White House staff might use insider knowledge to make bets on huge events, thereby earning a profit to make up for their lackluster public salaries.

We have no actual indication that any of them have done this! It’s just — y’know — believable. What with everything.

After all, it has sure looked like people with privileged information have been placing bets.

But it never occurred to me that it might maybe be members of the U.S. military that could be making bets on globally significant events. We expect soldiers to be honorable public servants with enormous discretion, but something in the system that inculcates honor among our warriors might have broken down.

The Department of Justice alleges that a member of Army special forces made bets on Polymarket ahead of the ouster of Nicolás Maduro in Argentina.

Someone made over $400,000 on that latest little enforcement of the Monroe Doctrine. U.S. Attorney Jay Clayton1 now thinks he knows who, and he’s not happy about it.

One prediction down

Obviously, the accused is innocent until proven guilty.

But let’s just discuss this as if Clayton did really get the right guy, as the Southern District usually does.

One of my predictions for the year was that we’d all feel weird about prediction markets before we got to the end of the year. Here we are!

If it is really true that a member of the military was placing bets on prediction markets, that’s unsettling. We expect politicians and political employees to be a little corrupt. Political folks tend to be driven by narcissism and hubris.

Public servants are meant to be different, though. They are the ballast on the ship. We may or may not have crazy people captaining from time to time, but they come and go. It’s the career public servants we hope will balance their depredations.

Janus-face

The Greek god Janus didn’t just have two faces. They looked in two different directions.

Prediction markets may surface crucial information that would otherwise stay hidden. They may offer signals to people that would have otherwise stay hidden. This is one face.

But there’s another face. The money. The biggest problem of prediction markets is always their greatest asset: if you’ve got tea, the tea is much more tempting to sell than it is to tell.

And that’s why it’s far too facile to simply say prediction markets must be an unqualified good.

Here’s what’s tricky about this case: I’m not sure I really care about insider trading, as a rule. My personal ethical jury remains out on that question.

But here’s something I am sure about: I am sure I don’t want soldiers trading bets on military operations. That heuristic sits on solid ethical and pragmatic grounds.

Prediction markets might have been beneficial and, ultimately, inevitably, but let’s not put blinders on. Prediction markets also put corruption nothing but a click or two away.

Jeff Bezos has already shown how could bring in more revenue for writers — he’s just gotta take the leap.

Amazon, the online store that started with books but almost seems to have forgotten about them, actually created a massive platform for independent writers that most folks on here would probably disdain to even think about.

But those proletarian scriveners are getting paid, my friends, and Substack could easily follow where Amazon led.

Substack already solved our problem of scaling email and payment management. That was very nice of them. Now let’s see it innovate to grow the pie.

Friends: Let’s talk about Substack Unlimited.

Lies, damn lies and big 5 publishers

Back in 2016 and 2017, writing for the New York Observer, I got a bit obsessed with the Amazon Kindle as a reporting topic. E-readers, cheap publishing and e-ink all looked very good to me.

Yes, there are trade-offs. I get that and I got that then, but Kindles extended the reach of literature, lowered the barrier to entry and kicked-off an indie book subculture.

But back then, mainstream reporters were doing stories every year or so that said, basically, “No one actually buys ebooks.” Which was weird, because Amazon kept putting out new versions of the Kindle and it didn’t seem like that was slowing down — which indicated that it was probably working out.

Something was up!

So, I dug into it, and I figured it out. All those reports were — well — basically lies. At least, they weren’t telling the full story. When the headline read: “People aren’t buying ebooks” what it actually meant was “people aren’t buying ebooks from Big 5 publishers.”

This was a very important distinction!

It wasn’t hard to work out why they weren’t buying those books! Publishers were jacking up the prices on ebooks! Amazon wanted to sell ebooks for $9.99 each but the Big 5 didn’t. After a long court fight, they upped the price, sometimes to the point that an ebook cost more than a paperback.

So the fact that these books weren’t selling that well was just basic economics. Nevermind the fact that — all things being equal — lots of people like a physical object. It was actually even simpler even than that. People Discriminate By Price: News at 11.

Indie in the Amazon

But what those reports didn’t tell you was that there was this buzzing world of independent publishing enabled by Amazon Kindles, through its Kindle Direct Publishing program, that was totally popping off.

Now us smart Substack people have probably never read anything from KDP because we are all much too brilliant for it and we haven’t quite finished Infinite Jest. The kind of writing that moves on there is mostly romance, but also other genre stuff: such as sci-fi, LitRPG, and etc.

Back when I was reporting on it, authors confessed to me off the record that they were making millions of dollars writing on there. I’d ask them if they needed to take jobs teaching writing at Universities like all the New York Times bestselling authors you’ve heard of. They would laugh.

No, the best use of time for them was always going up into the attic, opening the laptop, and banging out the next chapter. Every time.

And what was the secret of all that money being made? It was: Kindle Unlimited.

Unlimit my bags

Now this might sound counter-intuitive, and that’s precisely why I suspect that if Substack tried to follow Amazon’s lead, they’d get rebellion from the big guns on this site. But hear me out.

Kindle Unlimited is a Netflix for books. Readers pay a flat amount each month (right now it is $11.99, in the US) and they get unlimited access to books in the program. Precisely how authors get paid out of this is a bit arcane (because: scammers), but basically it comes out to pages read in their books. It’s not enough for someone to download your tentacle porn urban detective story — they need to actually move through the pages. The more pages they move, the more your share of that $11.99/month is.

Amazon is spreading around more than $60 million per month these days among authors. We don’t really know how much Substack is spreading around, but I’m confident saying it’s significantly smaller.

Here’s the thing: You can still sell your books one-by-one on Amazon Kindle. That’s no problem. That gets you access to the light readers out there, and it works great.

But there’s also hardcore crazy readers out there. These are people who will find one of your books and, if they like it, they will crank through everything else you have written. This has a way of kicking your earnings into gear fast. These folks are all in on Kindle Unlimited.

Have you seen Silo on Apple TV+? That’s based on a self-published novel by Hugh Howey. Howey was one of my sources back when I was working this beat. Back in 2017, Howey told me over repeated conversations that going exclusive with Kindle Unlimited cranked his revenue into very, very high gear.

It seems risky? Why would you go into a shared pool when people are already buying your stuff directly at a pretty nice clip?

Why? Because price-elasticity-of-demand, baby. When it works, it feels so good.

Into the breach, dear friends

There’s a lot of talk about bundling on Substack. That’s fine. I’m all for it, but it seems like amateur hour to me.

Better idea: Mega-bundle. Substack Unlimited.

Substack Unlimited could be an opt-in program that any writer on here could join. You can offer people to subscribe directly to your paid content (like buying one book at a time on an Amazon Kindle) or they can get it if you’re in the Substack Unlimited program.

This could easily work the same way. There’s a big pot of money from all the Unlimited users each month and that gets shared out to everyone via pageviews.

LFG.

Leave a comment

Would this undercut the earnings of the big guns?

We can’t know for sure, but Howey was a big gun on Amazon Kindle Direct Publishing already when he decided to join Kindle Unlimited, and it didn’t undercut him.

But, also, they don’t have to join it if they are nervous. They can keep getting paid directly by users.

How can we know it would grow the pie?

We can’t until Substack tries it, but it worked for ebooks already.

Wouldn’t enabling bundles work better because it’s more thematic?

They aren’t mutually exclusive. I’m all for bundles, but, I submit, that the desire for coherent themes is one of those OCD impulses that makes sense in our heads, with their need for structure and order, but actually doesn’t really help that much.

As the great Jane Jacobs showed: Zoning parts of city as “industrial” or “commercial” seems logical, but by and large, letting chaos reign works out better over time.

Wouldn’t all the big shots eschew Substack Unlimited so it would just fill up with all the lesser writers?

That’s exactly what happened on Kindle Unlimited, yes. Stephen King is not in there. But Stephen King was already rich.

Of the hundreds of writers making mad bank on Kindle Unlimited, those of us who aren’t using may hear about one of them every couple of years. There are many more.

Thousands of minnows swimming around the phat lake of Amazon KDP yielded some... whale-like minnows. Sounds pretty good to me.

But, but... if Substack created an Unlimited program, wouldn’t that mean I’d be in the same pool as people I disagree with?

Buy a helmet, kiddo.

Look. A lot of us spend a bunch monthly on here, and, while we are glad to support people, comparing the quantity of content our dollars get us access to compared to, say, one subscription to The Atlantic, it’s no contest.

A lot of people out there are more price sensitive than I was when I got into this site. Basic economics says that if there’s a lower cost overall price entry point, more people will come in.

My means are more limited now than they have been for a while. I’m dying to upgrade my subscription to to paid, but I just can’t justify it at the moment. If we had Substack Unlimited, though and she joined? I’d be all over her archive.

At the end of the day, I think more people want Substack’s core product than they want Kindle’s product. There are more readers of topical blogs than there are of indie genre books.

Substack Unlimited could be even bigger than the half-billion dollar behemoth that is Kindle Unlimited.

Substack Unlimited would grow the pool for everyone.

Jeff Bezos showed the way, and he did all right. Have faith.

Tech platforms wrecked the business model for news. Ironically, though, tech platforms also rely on the news industry to keep their attention-based model chugging.

That’s why tech platforms should open their own independent news desks. By mining the data spit off by platforms, managing editors guide the editorial direction of reporters toward topics that interested platform users. And news would actually have a business model again.

Platforms might not quite be the new nations, but they are the land masses of the internet. It’s time to just deal with it.

• What I’m about to propose here will annoy just about everyone. Just ask yourself please about whether or your irritation compares it to the hypothetical ideal of a long lost era or to the present, extremely bleak and far from perfect reality, please.

How platforms wrecked news

Everyone knows the story about how Craigslist came along and yoinked classifieds from newspapers.

Then Google and Facebook came along with a wildly better way of targeting other kinds of advertising directly to the kind of consumers a given advertiser wants.

The can say: “Our users are broadly 35- to 55-year old, high net-worth coastal professionals” (or whatever they say — they aren’t pitching me), but Facebook could say, “Do you want a 35 to 39 year-old divorced guy in Columbus, Ohio, who loves Anime and no college education? Yeah. We know who every single one of those guys are, and we can show your ad to all of them.”

It wasn’t even close!

Everyone knows that Facebook and Google have become an advertising duopoly, and the only companies cutting into their duopolistic status are other platforms, such as Apple and Amazon. Media sites have no chance.

So you know that story. What’s talked about less, though, is how much the platforms crushed the other advantage news once had. News used to drive a lot of attention. People watched the news on TV. They read the paper. They listened to news radio.

None of that is really true any longer. Now they scroll and they stream. This is how media gets consumed. They don’t watch the 6 o’clock news because that’s what’s on TV. They watch whatever they want. And that’s not usually news.

• There’s some research out there backing this up that I’ll point to in an upcoming essay.

So now in this era when news attempts to pivot to a subscription-based model, it’s coming really, really late. People already very much like the infinite candy of social. The habit of the news is long lost.

If you can’t beat ‘em

But here’s the problem for the platforms: People like to talk about the news.

The News and the Media are often conflated. The Media is just about anything that holds your attention.

But the News is the business of going out and witnessing the world, then coming back and telling an audience what was seen. (It’s also talking to people doing interesting things, like writing laws and building robots, that aren’t conducive to photos).

News is telling the masses about the world, in short. On platforms, people want to talk about the world and what’s going on in it, but they need to know what’s going on.

But there’s fewer and fewer people out there going out and having a look, and those that remain are all going out and looking at the same things, too, because those are the things that are most likely to drive enough attention for them to keep going.

As people get frustrated with the quality of the information available, they also get frustrated with the platforms.

The platforms benefit from high quality information gathering. So the platforms should open news desks and hire reporters who establish beats that are responsive to the interests of platform users.

Substack News

The first concern I can imagine substackers having about Substack News would be that it would compete with them for attention and subscriptions.

But, wait!

Structured correctly, Substack News would drive more subscriptions for everyone.

What does Substack want? More paid subscribers. So here’s my idea: Make Substack News a premium feature that’s offered to anyone who pays to subscribe to literally any existing substack.

I suspect the toughest conversion for every user here is making that first paid subscription. Like, I’m a free user. I’ve never paid to see anyone. But I really really want to get access to every delicious thing that posts, however I’m just on the margin about subbing.

Is it quite worth it? I’m on a budget.

But ‘lo! Then you hear that every subscriber also gets access to this high quality newsroom that’s covering tech updates, foreign policy, climate change, economic development. All the big things.

All you have to do is buy one subscription to get access to all of it. You can have your cake (recipes) and (r)ea(d) (the news), too!

And others

Any platform that’s attempting to launch any sort of paid subscription should do the same thing: open newsrooms responsive to the main interests of its users.

X is the obvious other example of a platform that should open a news desk. It would probably take about 45 minutes before good reporters at X irritated Elon Musk, of course, but it’s a good concept in theory. People on X talk about the news constantly.

More reporting on more topics could drive more interesting engagement there. And obviously you’d only have full access if you bought an X premium account.

Netflix should obviously open a news desk. It’s the world’s new CBS. It should give its viewers an option to watch a new kind of nightly news.

Snapchat should add news to Snapchat+. Reddit has tried doing a publication before, and it failed, but it didn’t try News. Reddit Premium with News, focusing on reporting, not culture or criticism, could do better. It aligns rather than competes.

Some objections

Won’t these news desks be biased on behalf of the platforms?

Probably, but narrowly. The blind spots would be predictable. It wouldn’t take much media savvy to know which reports to give side eye and which ones are probably reliable.

• And it’s not like reporting isn’t crazily biased now.

But, okay, to deal with that skepticism, jobs on platform newsdesks should be contract-based, not at will. The Editor-in-chief should have a contract for multiple years with clear performance indicators and clear editorial independence. All of this can be made public.

To me, the bandwidth of bias here is important but it’s narrow. Reporters in these rooms might be biased on behalf of the platform or even tech. It’s a much bigger world out there, though with much to discuss.

If, for example, Substack realized it had a ton of readers in Chicago and it opened a Chicago desk, would I worry about the Substack-bias corrupting Chicago coverage? Not really.

Objectivity should be a north star, but it is largely a fiction. I’d rather have more truth seeking in the world, even if I have to be a little skeptical about some of the truth-seekers’ coverage, than less.

And if Substack News covers soft, count on X News to make up for it.

Idealism about this dying industry has a way of letting perfect oppose the good. Being biased toward Substack (or X or Reddit or whatever) doesn't matter, it seems to me, when you're covering trade policy or baseball contracts or the conflict in Argentina, but what do I know?

Won’t readers get mad at the reporters covering the platforms they love when their reporting contradicts users’ priors?

Yes. Hopefully! That still gives users something to talk about! And that’s what keeps users on the platforms.

Shouldn’t we want people to get off the platforms?

I’m just trying to deal with the world as it is, friend, not the world as I’d like it to be. If I were writing about the world as I’d like it to be, I’d just describe the 70s all day.

At the end of the day, News doesn’t have a business model any longer. Everything you dislike about “the media” is downstream from that.

• The New York Times business model is being the New York Times (+Wordle). No one else can replicate that.

But platforms have a business model, but it’s one that needs good news gathering to really prosper. And if that news gathering can also help a platform deliver on what it promises its strongest users, revenue, then even better.

There’s a lot of talk about whether news can be decentralized. So far, I see media being decentralized, but I don’t see news decentralizing yet.

People are reading and watching stuff, for sure, but just because it’s on a screen, that doesn’t make it news. Substack has proven it can give writers a way to make money off their expertise, but there’s not much news gathering happening on here.

There’s research. There’s analysis. There’s opinion. God is there ever opinion.

But not many people are going out, having a look and just telling you what they saw. And if they do it sometimes, they aren’t doing it in a particularly structured way, by and large.

Tech platforms wrecked the news gathering business. I don’t fault them for it. Everyone’s gotta eat. But they are going to realize before long that they also depend on a continuous flow of good reporting.

Just as tech destroyed this industry, the platforms may soon realize that it’s also in their interest to build it back up.

Come back to Backstage tomorrow for my take on Substack Unlimited.