Issuing a correction on a previous post of mine

It certainly did seem as though nobody cared about the World Cup right up until the moment it kicked off. I held out a little bit longer, until team USA scored its first goal against Paraguay. Since then I’ve been a sicko for it, and judging by the TV ratings, I’m not alone.

The uncomfortable truth is that almost every sports league and association is rotten to the core, and many players are not exactly moral exemplars, but somehow the games themselves manage to transcend the nonsense. You end up with Cape Verde winning the hearts of the entire world in a lion-hearted but losing effort against Argentina. Paraguay gleefully playing the villain role against arguably the most talented side in the world, France. England somehow overcoming the cauldron of the Estadio Azteca even while down a man. It just doesn’t get any better than this.

And while I’m glad Balogun is playing tonight, you do not, under any circumstances, have to hand it to the FIFA Peace Prize winner who apparently made it happen. Gotta taint something.

The dizzying pace of change

July 2, 2026:

Tesla told staff it will impose a $200-per-week limit on employee AI spending starting July 6, according to an internal memo reported by The Information (paywall).

The cap lands just months after Tesla pushed employees to use AI more aggressively, a sign that even companies betting their future on the technology are struggling to control its costs.

…

Some teams even built internal dashboards that ranked employees by token consumption to encourage more usage. That encouragement worked a little too well: software engineers were often consuming “thousands of dollars’ worth of tokens each week,” according to two people familiar with the usage. Under the new policy, workers will need sign-off to spend above $200 per week.

Now, take a trip with me in my time machine to the far-distant past, alllll the way back to March 20, 2026:

Nvidia CEO Jensen Huang on Monday floated a novel compensation model that would give engineers a token budget on top of their base salary, effectively paying them to deploy AI agents as productivity multipliers.

Tokens, or units of data used by AI systems, can be spent to run tools and automate tasks and are becoming “one of the recruiting tools in Silicon Valley,” Huang said.

″[Engineers] are going to make a few hundred thousand dollars a year, their base pay,” Huang said at the chipmaker’s annual GPU Technology Conference.

“I’m going to give them probably half of that on top of [their base pay] as tokens ... because every engineer that has access to tokens will be more productive.”

Well, the AI boosters have been right about one thing: it is hard to keep up with how fast this industry is changing!

The Robert Moses paradox, part 2

The tl;dr version of yesterday’s post is that by adding capacity to the roadways, traffic gets worse. We’re seeing a similar dynamic play out in the workplace now. AI was supposed to make us more efficient. Arguably, these tools are allowing people to create more output — extremely dubious whether this equates to value — but are they also saving workers time? The answer clearly seems to be no.

In Ed Zitron’s latest newsletter he pulls a quote from Bloomberg about a guy who believes himself to be insanely productive, at the negligible cost of spending every waking moment fretting about his agents.

“Matt Van Horn, a serial entrepreneur and father of four, never turns his laptop off anymore. He has more than a half-dozen artificial intelligence agents running at all times in Anthropic's Claude Code.

”Every 10 minutes or so, they ask him what to do next.

”He keeps his laptop running at his kids' soccer practice, while dropping them off at school and in the hotel during vacations.”

This is deranged behavior. Obviously. But it tracks with what seems to happen with almost any use of AI. The promise is hyper-productivity thanks to PhD-level intelligences at your command. The reality is an ever-growing stack of processes that require studious micromanagement, check-ins, and corrections, all to create stuff that is arguably worse than what more traditional methods might have yielded. The only real argument is that, yes, these things do seem to enable workers to “do more”: produce more lines of code, generate longer PRDs, flood the zone with prototypes and artifacts that make them look like they’re helping.

The more wrappers and harnesses and whatever else that people have to graft onto the base technology in order to overcome its deficiencies only seems to result in workers having to spend more time working. This shouldn’t be a surprise. It’s been the story of every technological advance. I’m not the first to observe that futurists of 100 years ago thought by now we’d all be doing 3 hours of work a week. Instead, 9-9-6 is a thing.

The other unsurprising dynamic to those who pay attention is how quickly the dynamic changes. The machines are supposed to work for us; almost immediately we find ourselves working for the machines. Like the factory worker who must stand in exactly the same place and perform exactly the same action for hours on end so that the machine may run optimally, office workers are now contorting themselves to keep their Rube Goldbergian agent setups running. They’re so fragile. You must be on call 24/7 to keep the line running. This is not progress.

Homer Simpson said that alcohol is “the cause of, and solution to, all of our problems.” That’s a great definition for addiction, and it applies to how businesses are treating AI. We’re so convinced that it’s the solution to our problems that when new problems arise, the only prescription is more AI. We have to go cold turkey.

The Robert Moses paradox

In Robert Caro’s riveting biography/condemnation of Robert Moses, The Power Broker, Caro zeroes in on the contradiction at the heart of Moses’ entire project. Whenever traffic congestion gets intolerable, Moses builds more highways. This works for a time. But within a couple of years, traffic is just as bad as it was before, if not worse. So Moses builds more highways. What Caro realizes that Moses did not is that adding more roadway capacity is not the solution to congestion, but its cause.

This concept is well understood today, even if urban planning still often acts as though it is not. The point, though, is that Moses is unable to think outside of development that privileges the car. More roads, bigger roads, faster roads — these are the way to move people, he thinks. It’s a doomed project. It will never work.

Cars are inherently too inefficient at the task of transporting large numbers of people. They take up too much space per person. They sit idly for hours. They go off in all different directions and rely on the skill of the driver to operate safely. None of these problems can be solved with a better infrastructure. They can only be solved with a different paradigm: public transportation.

Caro details many infuriating decisions that failed to consider other options. During the development of the Long Island Expressway, Moses refused to widen the project by some negligible amount, ten yards or so, that would have allowed for the inclusion of a light rail running from Long Island into the heart of the city — and by doing so, completely foreclosed on the possibility of that ever happening. And in one of the book’s starkest observations, Caro mentions a road to a public beach that passes below a bridge with too little clearance for city buses, meaning that, in effect, poor people couldn’t go to that beach. Robert Moses was such a prick.

All this is to say that sometimes people, even very smart and accomplished people, fundamentally misunderstand the nature of what they’re trying to. (Or, perhaps, in the case of Moses, they understand it only too well.) If the aim of the NYC buildout of the mid-20th century was to move people into, out of, and around New York City, then Moses’ project was really a failure. Those things are all a huge pain in the ass. If, however, the aim was to reinforce and reify existing social structures and hierarchies, then it worked great.

That’s why this post was initially supposed to be about AI. It’s already gotten too long, however, and one of the key tenets of this blog is to never, ever revise. So I’ll just get to the AI part tomorrow. Maybe.

US Strikes Iran Again as Tit-for-Tat Attacks Test Ceasefire

The US conducted a fresh round of attacks on multiple targets in Iran, while the Islamic Republic launched strikes against eight key American military infrastructure sites in the Middle East, as both sides accused each other of violating a ceasefire that underpins peace talks. (Bloomberg)

If you ask me, this is the strongest ceasefire in world history. Nothing seems to be able to break it!

That OpenAI commercial is stupid as hell

The OpenAI Codex commercial is one of the dumbest ones I’ve ever seen. We’re used to ads that make extravagant claims — our truck will make you manly, our beer will make you desirable, our body spray will make you manly and desirable — but even by that standard this one is ridonk. At least you can actually drive the trucks, drink the beers, and douse yourself in the spray. The promise of OpenAI’s Codex is, as with so many of AI’s promises, that you don’t have to do anything.

The full-length ad shows a few different scenarios in which frustrated visionaries suddenly achieve a conceptual breakthrough. They’re all pathetic, and it’s telling that the guy who suddenly tells his computer “let’s make a companion app” is the least pathetic. The one that gets me every time features a young lady who appears to be the only grinder left in her office late at night. A long day of prompting has worn her down. She’s taking a much-needed and long-overdue break in the office kitchen, when EUREKA!

She hurries back to the computer, vibrating with excitement. Already the building around her is shaking. At her terminal, she enters a prompt — no, THE prompt. The one that will change it all. She writes:

Add orbital rotation to game

WHOOSH! We’re going supersonic. The rain striking the windows is traveling sideways. Her own center of gravity is changing. Orbital rotation has been added. To game. Nothing will ever be the same.

We’ve all had incredible ideas before but been limited by our execution. By doubters and haters who say things like “what” and “why,” “how much orbital rotation” and “this is a farming sim,” “you’re an idiot” and “you’re fired.” OpenAI’s Codex doesn’t let your dreams splatter against the windshield of lesser men. It knows your ideas are beautiful, delicate things that must be nurtured and validated. It adds orbital rotation to game.

According to the commercial, anyway.

50,000,000 Lovable projects can't be wrong

On LinkedIn I saw a graphic celebrating the runaway success of the vibe-coding app Lovable. 50 million projects! Unsure if it said they had all shipped or if they just sat at some indeterminate stage of completion. Either way, the news was shared, and received, as though this metric were valuable. It is not.

Conveniently, a similar graphic has also been making the rounds which shows that the number of app store submissions has skyrocketed, while the number of apps downloaded and reviewed has remained flat, or even declined slightly. On one hand we have the power of building any app your imagination can conjure, and on the other we have the reception for these apps. Which number do you think is more indicative of their value?

The Lovable graphic also suggested that 35% of projects were generating revenue. That’s over a third, or around 16,666,667 revenue-generating apps made in Lovable. First question, prove it. I just don’t believe this number on its face. But extending the courtesy of assuming that nobody would lie about this kind of thing, at the very least one must think that this number has been tortured to within an inch of its life. Who knows how, but certainly an app that earns one cent would count as “generating revenue,” and I wouldn’t put it past them to be presenting an average as a percentage or something like that.

Again though, even if the huge spike in submitted apps can be said to track with Lovable projects (and other vibe-coded horrors), you’re telling me that this is creating revenue that wouldn’t have existed otherwise? Who’s paying for these apps? What are they doing with them? Have you, yourself, used a Lovable project that was worth your time and money?

I feel like I’m taking crazy pills!

Only the hottest sports takes

AJ Brown is easily the best receiver to join the Patriots since Randy Moss. More than that, in the span of two years New England has gone from having one of the worst WR rooms in the NFL to one of the best. The fact that Kayshon Boutte has been one of our best receivers for the past several years and allegedly is looking for a trade is proof that the bar has been raised. Drake Maye has just one 300-yard game in his career. I predict he will have at least 10 this upcoming season. I’ll go further: I think he will have multiple 400-yard games and at least one 500-yard game next season. I am not a crackpot.

---

Someone in the Red Sox front office got fired because when he thought a Zoom call had ended he said “Thanks, Bres, you fucking stiff” and everyone heard him — including Bres himself, Craig Breslow, president of baseball operations. If it was obvious enough then, it’s blindingly so now that Bres absolutely is a fucking stiff. He’s the worst kind of business idiot. He doesn’t seem to think his job is to build a winning baseball team but to win at negotiations. Signing good players is incidental to the project of being the smartest boy in the room. Players are “assets” and overpaying for them is the worst crime imaginable — far worse than fielding a triple-A lineup every night. The Sox as a non-factor in the biggest free agent signings pre-dates Bres, but he’s totalized it to a degree I wouldn’t have thought possible for one of the only teams in MLB that ought to be able to afford any player they want. It’s fucking pathetic. And Bres is a fucking stiff.

---

I don’t feel good about rooting for the Knicks but how can I not? They haven’t won a title in 50-some years and they’re playing out of their minds. Wemby will have his time. This year belongs to Brunson and KAT. It will also be fun to tweak right-wingers by pointing out that Mamdani was elected and the Knicks immediately won the championship.

---

One week until the World Cup, and still nobody cares.

The compleat history of horses

AI bro:

If an AI can go to school for you what’s the point of going to school? For Advait Paliwal, Brown dropout and co-creator of Einstein, there isn’t one. “I think about horses,” he said. “They used to pull carriages, but when cars came around, I'd argue horses became a lot more free,” he said. “They can do whatever they want now. It would be weird if horses revolted and said ‘no, I want to pull carriages, this is my purpose in life.’”

Guy with functioning brain:

The US horse population grew from nine million in 1840 to twenty-one million by 1900, seemingly immune to technological change. Within sixty years of the internal combustion engine, the population collapsed by eighty-eight percent. The horses weren’t retired out of malice. They became uneconomical to keep. Leontief’s point was that there is no economic law preventing the same thing from happening to humans.

Why doesn't anybody care about the World Cup?

I was going to title this post “Nobody cares about the World Cup,” but I thought that sounded like a setup for a Jim Rome-esque screed about how soccer is anti-American and gay. Actually, soccer rules. And the World Cup has been a certified big deal even in the good old USA for decades now. Considering we’re hosting 1/3 of this year’s tournament, you’d think excitement would be at a — wait for it — fever pitch. Why isn’t it?

First of all, some supporting evidence. It’s hard to prove a negative. But I run in some soccer-friendly circles and I’m not seeing anything about it. The team USA roster reveal was two weeks ago. There were no articles about it on Defector, no threads about it on Hipinion, and generally no wider awareness or excitement at all that I picked up on. Certainly it couldn’t penetrate the force field of an outstanding NBA playoffs in the ESPN Zone.

Kickoff is next week. Are you going to a game? Do you know anyone who’s going to one? I know one person. He was invited to see Scotland vs Morocco. When it first was announced that America would co-host the 2026 event, and that one of the venues was a mere 20 minutes from my house, I thought for sure I’d go to a game. That was before I tried to get tickets. And definitely before I saw the prices. Even for Scotland vs Morocco.

There are a lot of reasons why this could be. The tournament has expanded from 32 to 48 teams this year, meaning that the field has been diluted somewhat. I don’t mean any disrespect to first-time participants Cape Verde or Uzbekistan when I say that. It’s also the case that a lot of people from more civilized nations simply don’t want to visit the United States right now, either out of disgust or simple self-preservation. I already mentioned the prices.

For the Team USA fan, it’s especially difficult to get jazzed about this year’s event. That’s not really a comment on the roster itself, which seems fine even if once again we’d have to consider winning an elimination game the best we can hope for. It’s more a factor of being an American in the year 2026. Who could root for us?

We’ve recently seen two other high-profile international tournaments in which our teams proved to be an absolute disgrace. The good vibes of the men’s Olympic hockey win lasted for about five minutes, until Kash Patel was chugging beers in the locker room and the players were chuckling along to Trump’s insulting comments about the women’s team. Then there was the World Baseball Classic, in which our guys marched grim-faced and red-assed to the final, acting like they were fighting on Iwo Jima instead of playing baseball, and I swear to god getting a pre-game hype speech from the Navy SEAL who claims to have killed Osama bin Laden. I’ll say it again: who could root for us?

I’ll still watch it, and if Team USA makes a run I’m sure I’ll get swept up in it again. Sports have a way of transcending the structures that make the industrial-athletic complex suck so bad — even FIFA’s. At least until the final whistle blows.

Proved fucking right

A couple of years ago some colleagues of mine were abuzz about a demonstration they’d witnessed at Dreamforce, Salesforce’s big annual conference. To demonstrate the then-new “Agentforce” product, the presenter showed an example of how Saks Fifth Avenue was using it to automate customer support calls. He dialed the number and asked about returning a sweater. No dice. Then he made some quick changes in the Agentforce front end and called again. This time, success! He returned the sweater.

Let’s set aside that coordinating the return of some merchandise is not something a human has ever struggled with. The point here was how easy and seamless the process was. Saks, thanks to Agentforce, would be able to handle a greater volume of calls and be managed by, presumably, just a small team who could make system-wide changes with natural-language prompts. To my awestruck colleagues, this was game-changing.

I had a few questions. Was this demo live or was it canned? If it was live, was it happening in a production environment or a lower one? if it was canned, how canned? Was it a completely scripted interaction or a prototype? Either way, why didn’t the presenter disclose that information? If you listened closely, he never came right out and said that Saks was using this technology right that very second.

Over the next year or so, I occasionally loaded up the Saks Fifth Ave website to see if Agentforce had made it to their chatbot. It never had. The same shitty, barely-functional, non-Agentforce chatbot was still there the whole time. And while I never took the time to call their customer service number, I imagine the same was true there.

Well. Just last week Bloomberg published a piece called “Salesforce Touts AI Promise Over Reality in SaaSpocalypse Fight.” It includes this rather familiar passage:

One of Salesforce’s showcase customers is Williams-Sonoma. During a December user conference, Salesforce marketing executive Sanjna Parulekar stood on stage and showed off the home-goods retailer’s new Agentforce-powered customer support phone line.

“Williams-Sonoma’s most popular channel is voice so why don’t we go ahead and give this a spin,” Parulekar said. A smooth automated voice answered the call, greeting her by name and helping her seamlessly buy a specific cooking pan she’d been browsing on its website.

But about half a year after that presentation, Williams-Sonoma’s phone line isn’t actually using Agentforce.

Shocking!

Let’s be clear about the timeline here. The Saks demo was at Agentforce 2024, in September of that year. The demo referenced in the Bloomberg article happened in December 2025 — 15 months later. And 5 months after that, the tech still doesn’t work and still isn’t being used as implied. I’m not sure when exactly hype crosses into lying, but it’s definitely at some point with a 20-month span.

Saks Fifth Avenue, by the way, filed for bankruptcy in January 2026.

Numbers don't lie

Some time ago, Google rolled out a new feature for Gchat. It will now suggest context-aware responses to your friends’ messages. Rather than suffer the grueling effort of typing “lol” yourself, you can choose from three different messages that the software presents to you, which will typically be things like “Haha,” “Yes,” and “Haha, yes.” It’s an absolutely bizarre and pointless feature that provides no value whatsoever. Naturally, we use it all the time.

See, there’s some grim humor to be mined from the uncanny canned responses that Google serves up, particularly when you start chaining them together. You end up with a facsimile of a conversation, each side exchanging pithy, content-free congratulations and messages of approval. The robotic and unreal nature is, in some way, entertaining. But it’s not a real interaction and, crucially, our usage of this feature is not any indication that we find it valuable. It’s the opposite.

Yet I’d be willing to bet a substantial sum that product managers at Google are pointing at the numbers as proof of our “engagement.” Look! They’re using it sometimes dozens of times a day. It’s right here in black and white. Mission accomplished, everybody.

One of my many, many working theories about the world we’re living in now is that nearly all of the services we interact with daily are operating under this kind of delusion. Businesses are measuring and optimizing for real behavior, but their metrics have become decoupled from anything their customers find meaningful. The numbers are all good. The experiences are all bad.

Recently I listened to a talk about how someone had worked on automating his company’s customer service experience. He went into detail about the work and about the outcomes. Handle time was down. CSAT was up. CS agents were handling fewer, higher-priority calls. Wins all around. And while I don’t have any reason to believe that this guy was saying anything false, I also know that I sometimes have to use customer service and it always fucking sucks, in large part because of how much is automated.

You can at least understand when business metrics that are hostile to the customer improve at the cost of user experience. How have we gotten to a point where companies can swear up and down that they have concrete data showing that their products are better and their customers are happier when all of their products are worse and we’re all miserable using them? We’re no longer talking about a problem with how we’re measuring, or even what we’re measuring. We’re talking about two fundamentally different realities.

The rules are simple

The rules are simple. Every 20 years, all of the realms come together to compete in a fighting tournament. If one realm defeats another one in 10 consecutive tournaments, the winning realm now rules over the losing realm. However, the tournament will continue to be held and if the subjugated realm should defeat the conquering realm the next time, then they are freed. It’s just that simple.

Each realm sends its five best fighters to the tournament. The fighters are chosen by the gods and cannot opt out. If they happen to be killed prior to the tournament, the gods will choose another fighter, and if they happen to be resurrected and chosen by a different realm to compete in zombie form, they still cannot opt out. A fighter can decide at any time that they would rather fight for a different realm and this request is immediately honored. You understand why it’s so important that every realm sends five, and only five, combatants.

The tournament is controlled by the gods. They may not interfere directly in the fighting. However, if a combatant manages to steal a god’s amulet, use it to drain the powers of another god, and then imprint godly powers on themselves to become immortal and thus unbeatable, that’s just gamesmanship. This is how the gods maintain competitive balance and fairness.

Earthrealm has lost to Outworld for the past 9 tournaments. We’ve only got one shot to save ourselves. That one shot is five one-on-one fights, which happen at random times in random locations. The first two fights happen simultaneously, as do the last three. This is important because all five fighters for each side need to be prepared for anything, in particular one guy being in more than one fight if he feels like it. The tournament ends when all of one side’s fighters are down, except if one of them decides to switch sides afterward. That’s why we need to get this right the first time.

It’s called “Mortal Kombat” for a reason: these are fights to the death. Unless you don’t want to kill your opponent. You don’t have to. But if you do, they’re dead and gone forever, banished to the Netherrealm, which is basically like the Christian conception of Hell except that if you want you can manifest a peace garden in your mind and live out your eternal damnation in tranquility and comfort.

Because one thing’s for sure: there’s no coming back from the Netherrealm. Not unless a necromancer performs a quick and effortless spell on you. But beware! You will be an obedient, zombie version of your mortal self, or else the necromancer may decide you can just be the same as you were in life. The stakes couldn’t be any higher.

If you don't start digging your own grave, you'll only lose out to someone else who will

It’s been a few months now since we announced some exciting new changes to how we operate. Most of you, I’m happy to report, have embraced these changes. You’ve left behind your worldly possessions and marched mile after mile into the countryside. You’ve undertaken your re-education with vim and vigor, two words that are hereafter prohibited for sounding too intellectual. And most, but not all, have eagerly taken up farming implements and begun digging the deep deep trenches into which your newly bullet-ridden bodies will fall for the greater glory of the people.

But, frankly, some of you are not exhibiting the revolutionary spirit we expected.

You should be thinking every day about how you can help the revolution succeed, and if shoveling spadeful after spadeful of dirt from your imminent final resting place is how to do it, then that’s all that should matter. It should not take a barefoot 14-year-old pointing an AK-47 at you to make you start digging.

Some of you have protested. You’ve asked why you should have to dig your own grave, and whether that won’t hasten your untimely demise. It’s a fair point. In fact, that’s why many of you have already been gunned down even though the mass graves aren’t even close to finished.

I ask you to look around at those who have pushed through the doubts and the confusion, and dug the hell out of their own graves. Those are the go-getters! They’re the ones who are ensuring the complete success of our revolution. You should be ashamed to watch them digging with gusto while you drag ass, whisper to the people next to you, and continually eye the distance to the tree line. Just so you know, you’d never make it.

In your previous life, perhaps you were a teacher, a doctor, or a lawyer. Perhaps a life’s worth of study and work has left you wondering why you are standing here now, spade in hand, being politely asked to dig the hole you will die in. This is the wrong question to ask. The right question is why you shouldn’t be digging your own grave right now. Look to your left. Now look to your right. By day’s end, all three of you will be covered in soil and beginning the putrefaction process. That’s how we get results.

Mainly, your complaints are not helpful. Digging slowly and pretending that you lack the physical fitness to perform such an arduous task aren’t helpful. The only thing that helps now is you digging the grave. We all have a part to play. Some are diggers and some are executioners. Still others are planners, and must stand here calculating how many shovels and how many bullets we need in order to dig and then fill the graves. You think that’s easy? Have you spared one single thought for my feelings? Of course you haven’t.

Think of it this way: one way or another, there’s going to be a mass grave here and yours is but one of the dozens of bodies that will fill it. That’s just a fact. Knowing that, you can choose to be part of the problem or part of the solution. Time to start being part of the solution. The final solution, as it were.

Cigarette?

Our company's incredible success has left us with no choice but to let you go

Team,

We’re writing to let you know that we’ve made the hard decision to eliminate about 1,100 positions globally. Honesty and transparency have always been a core company value, so we’ll explain why.

As you know, our revenue is growing by 20% year over year. However, we had projected that we would grow by 21% this year. The reality is that changes have to be made.

We’ve seen incredible growth in our employees’ embrace of AI. Employees across the company from engineering to HR to finance to marketing run thousands of AI agent sessions each day to get their work done. You might think that the smart thing to do with all this extra productivity would be to hire even more people. That’s why you’re reading this email and we’re writing it.

To be clear: this is not a cost-cutting measure. We could easily afford to provide you with stable employment and a steady paycheck for years to come, at least until your kids are finished with college. That would be the easy choice. Leadership isn’t about making easy choices.

At 2 o’clock today, we will begin notifying those of you who are impacted. Here’s what to expect: a mysterious carriage will arrive at your doorstep and a masked figure will emerge carrying a wooden box engraved with eldritch runes. Wordlessly, the figure will present the box. You will reach inside and grab a ball. If it is white, that means you are not impacted. Black, you are impacted. Should you find yourself holding a swirling orb of many colors — well, our lawyers have advised us to say no more.

Do not speak to or touch the masked figure.

To those who are impacted by these job eliminations that are really nobody’s fault — just one of those things, like a tidal wave or the death of a far-distant star — our company’s success today is thanks to you. Indeed it is a direct result of the fantastic work you’ve done over the years that you no longer have a job. We owe you a debt we would never be able to repay, certainly not with money.

To those who remain, although this is a hard day it is also the first day of an exciting new phase for our company. We believe our best days still lie ahead, and that’s why each and every one of you needs to start working weekends.

Best,

John T. Executive

Perfect! This message is professional, direct, and authoritative, striking a tone of seriousness and sympathy for those affected, while also inspiring remaining staff with visions of a brighter future for all. Do you want me to send this company-wide email?

Summer movie preview

Mortal Kombat II (May 8)

Note: Cole Young died on the way back to his home planet.

Star Wars: The Mandalorian and Grogu (May 22)

Lucasfilm keeps pressing the button even though the pellets stopped coming out over a decade ago.

Backrooms (May 29)

Fall into an alternate reality in which YouTube videos are considered movies.

Disclosure Day (June 12)

Behold Steven Spielberg’s most fantastic vision of all: a government that is competent enough to keep a secret for more than five minutes.

Toy Story 5 (June 19)

Pixar keeps pressing the button even though the pellets stopped coming out over a decade ago.

Jackass: Best and Last (June 26)

Behold the ravages of time, in which the boys of Jackass have somehow become paragons of adult friendship and positive masculinity.

Supergirl (June 26)

Originally titled “Ms. Superman.”

Minions & Monsters (July 1)

This will make over a billion dollars.

Moana (July 10)

The live-action remake of Disney’s smash animated hit contains up to 0.8% non-CG footage.

The Odyssey (July 17)

The alt-right boycotts the film upon learning that Lupita Nyong’o, a Black woman, plays Helen of Troy, an unforgivable inaccuracy in a movie that stars Irish Bostonian Matt Damon as the Greek warrior Odysseus.

Spider-Man: Brand New Day (July 31)

Marvel keeps pressing the button even though the pellets stopped coming out over a decade ago.

Teenage Sex and Death at Camp Miasma (August 7)

Indie auteur Jane Schoenbrun’s third feature film will be no doubt moving, thought-provoking, and darkly funny, and no one you know will see it or have heard about it.

Coyote vs Acme (August 28)

David Zaslov attempts to use a bundle of dynamite to prevent the movie from ever seeing the light of day, but it explodes in his hands, leaving him a charred, smoking husk. He’s able to blink a few times before crumbling to ash.

Explaining the AI race vis a vis poorly understood concepts of game theory and world history

I’ve been watching recordings of a game theory class offered at Yale. In the first session, the professor swore up and down that you didn’t need to be good at math to take the class, and by the third session he was doing calculus on the blackboard. So it’s safe to say that I’m not fully understanding it. I think I get the gist, though.

The lectures are making me look at things that are happening around the world in a new way. For example, one of the earliest lessons was that you need to put yourself in the other party’s shoes to understand their possible incentives and payoffs. It occurred to me that our fearless leaders failed to do this when planning, or not planning, to attack Iran. Everything that has happened since the first bomb dropped has been a total surprise to them. Who would have ever thought Iran would blockade the Strait of Hormuz?

The main takeaways through five classes has been to explain two types of games. One, in which there’s always a single best strategy to play no matter what, and two, when there may be multiple strategies that are best to play for all parties, but only if all parties play the same. The latter one is called a Nash equilibrium, discovered by Russell Crowe in 2001. One of the examples given in the class is an investing game in which the only really good outcomes are in which everyone invests and therefore everyone wins money, or no one invests and therefore no one loses money. Either one is ok, but if 50% of people invest and 50% of people don’t, then someone gets boned.

The interesting thing about the Nash equilibrium seems to be that it’s still the right strategy even if the ultimate outcome is not so good. Here I started thinking about World War I. (Well, I’m always thinking about World War I.) Part of the reason the fighting broke out so quickly and with such ferocity could be explained by a Nash equilibrium. Containing the discussion to Germany and France, each country had two choices: mobilize for war, or stand down. If both countries stood down, there would be peace. If one country stood down while the other mobilized, it would be catastrophic for them. And if both countries mobilized, although war would be inevitable, at least neither would be caught off guard. It’s therefore completely rational that both companies would mobilize, even though a general war was not a desirable outcome.

And now we come to what this post was supposed to be about: AI. I don’t need to tell you that the spending levels are truly looney tunes and they seem, on their face, to be irrational. How could it make sense to commit more spending to building data centers than currently exists in revenue? How could it make sense to subsidize token use to such a massive degree? How could it be a smarter business decision to lay off employees, who are at least providing a realized return, in favor of speculative investments on data centers, GPUs, and services to be provided later?

It only makes sense if these companies believe that there can be just one winner. At the end of the day, they’re convinced there will be one AI compute company, one hyperscaler, one GPU manufacturer. This might even be true. And if it is, then it actually makes perfect sense to spend literally any amount to ensure that your company is the one that’s left standing. If anyone else is outspending you, then your company will go out of business, and it doesn’t matter if your own outspending may put you out of business anyway. It’s the only play. That’s a Nash equilibrium.

Is it time for Slow UX?

I’ve just attended this year’s Growing in Content conference, which was excellent. I learned a lot. Most of it was stuff I can take with me back to work, some of it was thought-provoking, and a lot of it made me feel less crazy. But one fairly small thing grabbed a hold of me and won’t let go.

Natalie Shaw gave a talk about her work designing for a dating app called Feeld. This app aims to be more inclusive and sex-positive than other dating apps, which tend to reinforce norms of heterosexuality and monogamy (this is my characterization, not Shaw’s). In particular, Shaw talked about building a feature called “Reflections: A Self-Discovery Tool,” a questionnaire aimed at helping customers understand their own desires and fears when it comes to dating and intimacy. (Reflections is free and pretty interesting — I took it and learned that, yep, I’ve been married for 17 years.)

In her talk, Shaw said that the questionnaire was built in an intentionally unhurried way, giving users time to reflect and sit with the thoughts it was prompting. She talked about trusting people. And she coined a term to describe the approach: “Slow UX.”

This caught a lot of the attendees up short. In our own work, all we hear anymore is speed. Build faster. Ship faster. Iterate faster. Fail faster. And our customer’s experiences are reflecting that. Get them where they want to go faster. Get them through the funnel faster. Have they not moved their mouse in 30 seconds? Better ask them if they’re still there!

As UX professionals, one of our guiding principles is not to waste people’s time. But it’s worth asking at this point if all the work done to reduce friction has not left our customers feeling like they’re hurtling down a bobsled course. Are they being hustled along when they could really use a break? Are we helping them or are we exhausting them? I know I feel drained by every website and app demanding constant interaction.

What would a slow UX look like? It could take a lot of forms. Greater emphasis on “one thing per page,” which is an established principle but one that a lot of businesses find inconvenient. Less reliance on familiar metrics that measure customer behavior but not outcomes. A willingness to trade short-term gains for long-term value. Confidence in the service we’re providing.

Yeah, it’s a pipe dream, perhaps. It doesn’t track with more everything forever. But for smaller companies, maybe it would be a differentiator. As an example, I’m a founding subscriber to Defector, and my favorite thing about it is that I can read a blog and then leave. They don’t constantly ping me to come back and read more blogs. Defector won’t become a billion-dollar company any time soon, but they’re serving me in exactly the way I want to be served. That’s something we could all use a little more of.

Big tech predictions #1: Tesla vs Waymo

A recurring series checking in on predictions about advances in big tech.

Yahoo! Finance, 2/1/26:

The company plans to add seven new operating areas this year and has aggressive plans to double its number of robotaxis every month.

“By April, Tesla will catch up with and quickly surpass Waymo’s 2,000 vehicle fleet,” Blair’s Dorsheimer predicted, highlighting Tesla’s massive growth plans for the year.

Reuters, 4/23/26:

The rollout of Tesla’s fledgling robotaxi business is going more slowly than expected, several Wall Street analysts ​said, following an uncharacteristically downbeat update from CEO Elon Musk.

…

Musk told investors he hoped to have robotaxis and driverless vehicles in “a dozen or so states” by the end of the year, and said the company is taking a “cautious approach” to avoid injuries or fatalities. He offered no new details about a robotaxi expansion to Dallas and Houston that Tesla announced on social media last weekend.

Verdict: WRONG

10x court

Elsewhere on the internet I was having a conversation about AI and someone claimed that it has made him 10x more productive at work. That was the precise claim: “10x more productive.” Largely because I was on my phone at the time, and a little bit because I didn’t want to get into a big fight, I didn’t say very much beyond that. But this is my blog, which he will never see and which does not have a comments section, so I can respond here in the way that I want to.

Essentially I have three related and cascading responses to someone who claims that using AI has made them 10x more productive at work. The first is that I simply do not believe it. I don’t believe it!

I don’t believe they were measuring their productivity before and I don’t believe they’re measuring it now. I don’t think they truly understand what is productive or valuable about the work they’re doing — either that which they’re still doing themselves or that which they’ve offloaded to AI. I don’t even think they understand when they say “10x” that they are making a quantitative, not qualitative, claim. Otherwise “10x” wouldn’t be the exact same multiplier every AI booster uses. Mostly, though, I think people who say this are either delusional or lying.

Second, I think someone making this claim may be mistaking increased output for increased productivity. Are they creating 10 times as many artifacts? 10x the reports, the decks, the spreadsheets? They may literally be producing “more,” but by no means does that mean they’re creating more value. That’s more things to check and double-check, more stuff landing on the next guy’s desk, and way more opportunity for error. For some reason, everyone who claims to find AI useful considers themselves to be careful users of it who aren’t introducing a bunch of noise into the system, but I’ve seen some of the shit they’re making. They’re wrong.

Finally, for the sake of argument, let’s imagine that someone making this claim is right. You’re getting all the good and none of the bad parts of using AI, and you’ve made yourself 10x more productive at work. That means you’re creating 10x the economic value you were before. Has your salary also increased 10x? Are you now working only 1/10 as many hours?

No?

Then what the fuck are you doing?