448: Tyranny of Marginal Users, Twitter Paywall, Airbnb AI, Softbank, Microsoft + Nintendo?, Renaissance Tech, and Rodrigo y Gabriela

"How can you improve things if you can’t change things?"

Oh words, what crimes are committed in your name?

—Eugene Ionesco

📺👩🏻‍🏫✏️🚌🖍️ This is a fun *and* thought-provoking one!

OSV Fellows Nat and Martha Sharpe have created a free mini-documentary (30 mins) about the ZigZag ALC “un-school” in Asheville, North Carolina.

What happens if kids of various ages are put in an environment rich in learning opportunities and they are trusted to do what they want with very few limits (almost none, compared to a traditional school)?

Whether you have kids or not, I highly recommend watching this video — I saw it twice, once with my family — it’s a great conversation starter and made us question all kinds of things that we were taking for granted.

Both my kids spontaneously asked if there was a school like that where we live.

It’s very easy to forget that there are different ways of doing things — sometimes radically different — and even if this exact model isn’t for you, it’s healthy to question what is “normal” and look far and wide for ideas that can be incorporated into your family’s own unique way of doing things.

You can read more about ZigZag’s model on their website.

I love that they focus on building trust, self-confidence, creativity, curiosity, and the meta-skills necessary to learn and solve problems.

Most people lack these tools even after spending a decade+ in traditional school — their natural curiosity has been largely extinguished and they’ve mostly learned to “memorize the answer the teacher is looking for” which isn’t how things work in the real world.

I’m not saying this model — or other models like Montessori — are perfect, but I suspect that mass education needs to move past the rigid, authoritarian industrial daycare model and into something that fits better with human nature and helps more kids explore widely to learn to know themselves and get closer to their true potential.

🛀💭 📊 🩺👨🏻‍⚕️🩻 What is normal?

When it comes to health, a lot of biomarkers and metrics use population averages.

But if a very large faction of the population is unhealthy, should the average really be the target for what is “okay”?

For example, if someone is in the 50th percentile for strength compared to the rest of the population today, what does that say about how strong the average human was over the evolutionary periods when we used our bodies like they have evolved to rather than sitting in chairs and cars all day? 🤔

🚨🎙️🗣️🗣️ I had a great time talking with friend-of-the-show Trung Phan (🇨🇦🇻🇳) about everything from solving the Twitter puzzle to what we learned from Rick Rubin. Check it out:

• 🎧 Trung Phan: Solving the Twitter Puzzle, What He Learned from Rick Rubin, Christopher Nolan, AI, and more ✍️🤣🎥🎬🇻🇳🤖

💚 🥃 🐇 What has value in your life?

Is it the stuff? Is it the people? Is it the ideas? Is it understanding how things work? Is it discovering new art that makes you feel something? Is it playing pickup basketball with your buddies?

Hey, I’m just asking, I can’t tell you! 💡

Everyone has to figure this out for themselves, it’s crucial because it has an influence on so many other decisions that will then shape your life. You end up in different places based on these forks in the road. 🚦

But if you value what I’m sharing here, I hope you’ll become a paid supporter so that you can get even more of it and help me keep this steamboat in the water!

➡️ There’s a 7-day free trial if you want to explore! ⬅️

---

🏦 💰 Liberty Capital 💳 💴

🤴 The Tyranny of the Marginal User 👑

Interesting piece by Ivan Vendrov:

Nearly all popular consumer software has been trending towards minimal user agency, infinitely scrolling feeds, and garbage content.

We have to be careful of the “good old days” effect, but there’s certainly some of that going on.

Even that crown jewel of the Internet, Google Search itself, has decayed to the point of being unusable for complicated queries.

Reddit and Craigslist remain incredibly useful and valuable precisely because their software remains frozen in time. Like old Victorian mansions in San Francisco they stand, shielded by a quirk of fate from the winds of capital, reminders of a more humane age.

How is it possible that software gets worse, not better, over time, despite billions of dollars of R&D and rapid progress in tooling and AI? What evil force, more powerful than Innovation and Progress, is at work here?

The Tyranny of the Marginal User!™️

Simply put, companies building apps have strong incentives to gain more users, even users that derive very little value from the app.

Sometimes this is because you can monetize low value users by selling them ads. Often, it’s because your business relies on network effects and even low value users can help you build a moat.

So the north star metric for designers and engineers is typically something like Daily Active Users, or DAUs for short: the number of users who log into your app in a 24 hour period.

By definition, a marginal user is someone who is barely interested enough in a thing to use it.

The slightest annoyance or moment of boredom could make them leave, so everything has to be simple, easy, immediate, and have plenty of triggers for dopamine.

Add confetti and scratch-off lottery ticket mechanics, and you’re golden! 🎉

What’s wrong with such a metric? A product that many users want to use is a good product, right? Sort of. [...]

in practice, the design of popular apps caters almost entirely to the marginal user.

Vendrov argues that this drives services and products towards the lowest common denominator, turning what may have been a complex meal full of colors and aromas into empty-calorie hyper-palatable processed junk food.

I don’t think this model is 100% correct and explains everything, but I think it’s *useful* (“All models are wrong. Some models are useful.”).

It partially explains a lot of what we’ve been witnessing online over time as platforms scale up and essentially become the internet equivalent of broadcast TV or those magazines near the checkout.

Elon Musk: Twitter X to become a paid service ✋ 🔐 💵 🐦

I’m sure you’ve heard about this by now:

[Musk said] X would start charging “​​a small monthly payment for use of the X system” in order to combat bots and spam on the platform.

I’m not sure what I think of it.

Ben Thompson (💚 🥃 🎩) has long been suggesting they should try this — or at least, he did in the early days, a lot has happened since…

Twitter Blue hasn’t exactly set the world on fire. It has an estimated 1 million subs, which is less than 0.5% of the ˜250m DAUs.

BUT

The incentives to get Twitter Blue aren’t very strong for most users and the (re)launch was bungled so poorly that it turned into a social-signaling war and a referendum on Musk personally, rather than what it should have been: Simply a useful product that makes the experience better.

It got so bad that the company had to add the option of hiding the blue check mark because, in some sub-communities, it became an in-group/out-group marker and users were peer-pressured into not subscribing and ridiculed for doing so. (what is this, kindergarten? 🙄)

It’s interesting that Musk claims that a paywall is the only way of dealing with the spambot problem. I don’t think Meta is going to make its properties pay-only any time soon…

If Twitter becomes paid-only, I see three main risks:

1. Unraveling the network effect. Twitter is like nothing else when it comes to the interest-graph. However, if a paywall drives enough users away, it’s possible that the utility for remaining paid users will be significantly reduced.

   Past a certain point, more of these newly marginal users leave, which further reduces the vitality of the graph for users further up the curve, and now they become marginal and may decide to leave, and so on, until even power users start questioning things and exploring alternatives.

   The network effect is incredibly powerful on the way up because it scales super-linearly. But that cuts both ways 🗡️, and value also goes down super-linearly when you enter shrinking mode. 🕸️

2. Creating another synchronized moment for everyone to move over to Threads. Part of what makes services like Twitter sticky is the coordination problem. Every time an alternative comes around (Mastodon, Bluesky, Threads), many users will check it out, but their friends, favorite follows, and audiences aren’t there, so they come back.

   However, if the whole group could somehow coordinate and move over at the same time, they would find everything that they’re looking for on the other side and they’d stick around. The launch of Threads was one such opportunity for a lot of Twitter users to coordinate an uprooting. Some sub-communities seem to have migrated, though not Fintwit or most of the big ones.

   It’s hard to know exactly why, but I’ve argued that Threads was too successful too quickly, and everyone’s first impression was an app missing too many features and with a bad timeline algo.

   If a synchronized migration attempt were to happen again in a few months, prompted by Twitter going paid-only, people would find something a lot more mature, with more and better content and a better-tuned algo, a web version, chronological timeline option, and hopefully by then DMs and other crucial features.

3. The number of paid subs could be small enough that it doesn’t make up for lost ad revenue. I know that Twitter was never very good at monetizing via advertising and that since Musk’s takeover ad revenues are down 50-60%, so the math may be easier to make work than if, say, Instagram went paid-only.

   But still, this is a high-risk bet. There’s no precedent for such a large social network putting up a paywall. Are 5% of users going to pay? 15%? 30%? ¯\_(ツ)_/¯

   How much will they make per user?

   The higher they charge, the fewer total subs. But if they charge something very low, a huge % of that will get eaten up by credit card fees and possibly platform fees (they can’t possibly get everyone to sign up on the web to avoid paying Apple and Google… They can have the dual price structure that Twitter Blue has, but the extra complexity will drive away a lot of users).

Bottom line: I’m nervous about this.

I love Twitter, and while it sucked that the platform was mismanaged pretty much since it was first created in 2006, at least most of its failings were in the direction of being unchanging, which meant that you could count on it, kind of like a utility — it was a bit like a protocol, like email.

Email may not be innovating or fast-changing, but that’s part of what makes it useful, something that you can build on for the long-term.

If Twitter/X really is to be the world’s “town square”, in Musk’s own words, then if there’s a paywall, it’ll keep a lot of people out — especially the poorest ones.

TIL Airbnb has an “anti-party” AI system and teams of professional party-poopers 🎉🥳🍻✋🛑👮🏻‍♂️🚓

[Naba Banerjee is the person in charge of Airbnb’s worldwide ban on parties and she] spent more than three years figuring out how to battle party “collusion” by users, flag “repeat party houses” and, most of all, design an anti-party AI system with enough training data to halt high-risk reservations before the offender even gets to the checkout page. [...]

Some measures have worked better than others, but the company says party reports dropped 55% between August 2020 and August 2022 — and since the worldwide launch of Banerjee’s system in May, more than 320,000 guests have been blocked or redirected from booking attempts on Airbnb.

If you ever see this when trying to make a reservation, you’ve been busted (or are a false positive 😬):

This reminds me of friend-of-the-show Jimmy Soni’s book ‘The Founders’ about the early days of Paypal, and how the business was mortally threatened by fraud so they had to invent all kinds of ways of mitigating it (though it can *never* be fully solved, it’s both an arms race and a Red Queen race).

Here are some of the things Airbnb’s algo looks at:

The AI models look at hundreds of factors: the reservation’s closeness to the user’s birthday, the user’s age, length of stay, the listing’s proximity to where the user is based, how far in advance the reservation is being made, weekend vs. weekday, type of listing and [...] whether the listing is located in a heavily crowded location rather than a rural one. [...]

“When we started looking at the data, we found that in most cases, we were noticing that these were bookings that were made extremely last-minute, potentially by a guest account that was created at the last minute, and then a booking was made for a potential party weekend such as New Year’s Eve or Halloween, and they would book an entire home for maybe one night. And if you looked at where the guest actually lived, that was really in close proximity to where the listing was getting booked.”

They basically created a “fingerprint” for party bookings! Clever.

Now if only they figured out how to deal with all the crazy fees that are tacked on every reservation in the final steps and inflate the cost vs what you see when you first look at listings, that’d be wonderful. kthx

🇯🇵🤖 Masa Son’s Big AI Bet, Shot & Chaser Edition 🥃🍺

Shot:

Five years ago, SoftBank chief executive Masayoshi Son announced that he had narrowed his focus and picked out artificial intelligence as the single most important investing theme.

Chaser:

SoftBank’s Vision Fund said that it sold its entire stake in Nvidia, worth more than $3bn, in 2019. Since then, Nvidia’s share price is up more than 1,000 per cent.

Oops I guess ¯\_(ツ)_/¯

🎮👾🍄 Microsoft’s Board Ok’ed Acquisition of Nintendo in 2020

Microsoft's board of directors was "fully supportive" of a potential purchase of Nintendo if the opportunity arose, according to a newly leaked email from the company's head of gaming. [...]

In an August 2020 email to Takeshi Numoto, Microsoft's commercial chief marketing officer, Spencer pondered the possibility of Microsoft buying Nintendo.

But there was no catalyst pushing Nintendo to do a deal and they didn’t want to go hostile, so Spencer said they were “playing the long game”.

I guess in 2020 they didn’t yet know how painful the process of acquiring Activision would be… That may have thrown some cold water on the Nintendo idea.

🧮 Interview: Peter Brown, CEO of Renaissance Technologies 💰💰💰💰💰

The CEO of RenTech is usually pretty quiet publicly, so this was a bit of a surprise:

• 🎧 A conversation with Renaissance Technologies CEO Peter Brown

Here’s a highlight showing that they clearly understand the talent power law:

I remember when I was at IBM, there was this attitude that programmers were like plumbers.

If you need a big project done, just get more programmers. But I knew that some programmers were, like, ten times or more productive than others.

I kept pushing IBM management to recognize this fact. But it did not. I remember being in an IBM managers meeting and some guy from corporate headquarters was explaining how they created something called their headlights program. The goal of which was to identify the best programmers in the company and pay them 20 percent more than the other programmers.

Now, I figured this guy from corporate was making, like, $300,000 a year. So, I raised my hand and suggested they increase the pay of their best programmers to $400,000 a year. And he was stunned.

He said, "What? More than me? You've got to be kidding me. Well, if the guy's Bill Gates."

I said, "No, Bill Gates was making, like, 400 million per year. Not 400,000."

Anyway, they just didn't get it. We don't make that mistake. We pay our programmers a ton in accordance with the value we place on the infrastructure they produce.

h/t friend-of-the-show Frederik Gieschen

⚾️📝🤖 AI-written sports recaps pulled after widespread mockery

Large language models may be coming for commodity writing, but apparently, we’re not quite there yet when it comes to sports recaps:

The writer dubbed a football game between central Ohio’s Westerville North and Westerville Central a “close encounter of the athletic kind.”

Another story about a game between the Wyoming Cowboys and Ross Rams described a scoreboard that “was in hibernation in the fourth quarter.” When Ayersville High School staged a late comeback in another game, a write-up of their win read: “The Pilots avoided the brakes and shifted into victory gear.”

Yikes, the cringe is strong with that one 😬

But it’s not just bizarre turns of phrase, there’s also this ‘blue screen of death’ type of error:

The Worthington Christian [[WINNING_TEAM_MASCOT]] defeated the Westerville North [[LOSING_TEAM_MASCOT]] 2-1 in an Ohio boys soccer game on Saturday

This is a good example of why, at this stage, using these tools to augment what humans are doing works much better than trying to cut people entirely out of the loop.

---

🧪🔬 Liberty Labs 🧬 🔭

🇬🇧🔌🏭💡 How the UK's grid has changed since 1920

From close to 100% coal to close to 0% coal!

But a lot of the slack was picked up by natural gas, and nuclear has declined.

Here’s a different view, showing absolute numbers:

The UK is burning about as much coal today as in 1920, even if the percentages are very different! It shows the importance of not just looking at relative numbers, especially in countries that are rapidly industrializing (unlike the UK in recent decades).

‘In the time since California started one high-speed rail line that will never be completed, China built this entire high-speed rail network from scratch’ 🚅

15 years after the start of the project, there’s no track laid between LA and San Francisco, and the cost of the project is estimated at between 88 and 128 BILLION dollars. 🤯

Something really has to change in the US (and most Western countries) where it’s impossible to build anything big.

How can you improve things if you can’t change things?

That doesn’t mean China is getting everything right — far from it — but the optimal way of doing things certainly isn’t how the US is going about things right now.

h/t Noah Smith (the title quote is him)

🏰🎮👾 Digital Museum of Old Video Games (revisit game environments in your browser)

This is incredible! A digital museum of old video games that you can explore in your web browser:

• NoClip

So far it has lots of old Nintendo games, but I hope more keeps being added.

I mostly played with the Portal levels. It’s incredible how well it works (you have to click outside the menu to get the full view). You can move around with your keyboard and mouse, go through walls like a ghost, and explore until you’ve had your fill of nostalgia.

There are so many virtual worlds, either in 2D or 3D, that I spent a lot of time in as a kid. Some I thought I may never revisit!

---

🎨 🎭 Liberty Studio 👩‍🎨 🎥

🎶 Rodrigo y Gabriela (2014 Live Show) 🎶

I’ve long enjoyed these two (my fave album of theirs is the self-titled from 2006, with the close-up of the snake eye).

Here’s my recommendation: Play the video above on a big TV in your living room while you hang out with your family or eat dinner or whatever, and enjoy dipping in and out of it, almost like you were watching them in a small club while having food and drinks.

🎥 🎞️ Do you like documentaries? Do you need recommendations? 🍅👍

I love this!

Rocumentaries is just a bunch of recommendations for good documentaries. That’s it. Simple concept, obvious value, no need to overcomplicate things:

I launched Rocumentaries as a small side project in 2016. I'm no expert, just someone who loves a good doc (and apparently watches a hell of a lot of them).

I wanted to build the thing I wish existed: a place where I could find awesome docs and save time watching bad ones.

It doesn’t claim to be exhaustive or unbiased, it’s just one person’s take. If you happen to have similar taste, amazing! If not, no harm done, it happens…

It reminds me of the web in 2004, back when small indie projects like this were everywhere.

You can sign up for their mailing list to get recommendations delivered to you, or you can browse the archive by genre or distribution channel (ie. Netflix, Prime, Youtube, BBC, etc).

I mean, where else would I have found this mini-documentary about an elderly couple experimenting with VR to “travel” to places around the world?