421: YouTube’s $40bn Revenue, Zaslav’s Power Law, Samsung vs Google, Apple + Broadcom, Spotify, Neeva, DragGAN, and Ozempic
"...one emerges from the other."
Like resilience, self-organization is often sacrificed for purposes of short-term productivity and stability. Productivity and stability are the usual excuses for turning creative human beings into mechanical adjuncts to production processes.
-Donella H. Meadows, “Thinking in Systems: A Primer”
🎥🎬👴🏻🧔🏻♂️💸 How much did Martin Scorsese spend on de-aging ‘deepfake’ technology to make The Irishman?
I know the total budget was $159m, but I don’t know how much of that was specifically for those digital special FX (a lot was for actor salaries, period sets, costumes, digital set painting, etc).
I wonder, how much would it cost today, and how much better could the de-aging be executed?
Or to isolate variables: how much less would it cost to get the same level of realism that they achieved or how much more realistic would the same effects be today if they spent the same amount on them as they spent back then?
It wasn’t even that long ago — the film came out in 2019, so it used 2017-2018 tech, but it feels like 10-15 years ago because of the pandemic and the rate of progress of AI.
That’s the kind of thing I randomly wonder about… ¯\_(ツ)_/¯
💡 I really enjoyed my recent podcast conversation with Kevin Kelly so I’ve been looking for more.
This conversation he had with Tyler Cowen is completely different from the one we had, but very complementary, so I thought you might enjoy it too:
📺 😬 💥🪂🤳 Do you think it’s easy to attract enough attention to make a living on YouTube?
Well, this guy apparently crashed a plane on purpose for the views and is now facing 20 years in prison for covering up the crime… The video is still up!
📖🔍✍️🕵️♂️ I recently watched a wonderful documentary about Robert Caro and his editor, Robert Gottlieb. It’s called ‘Turn Every Page’ (2022).
It was inspiring to see how Caro dedicated most of his life to studying just two people — Robert Moses and Lyndon B. Johnson.
But what it really shows is how everything is interconnected, and if you dig deep enough into one thing, it gives you a lens to study lots of other things. Caro’s books were never just about these two people, but they were his gateways to larger worlds.
The film gave many examples of important facts that Caro could only discover by working diligently for years, by turning every page of primary source materials himself, rather than relying on what some other writer said about something, or by talking himself to the people involved, living where they lived, etc.
The main question that this film raised for me was: What about all the other subjects that don’t have a Robert Caro?
How many important facts and stories will never get discovered because nobody is working on them with this level of dedication and rigor? How many erroneous stories and ‘facts’ won’t get corrected?
You can choose to see this as sad, or as an inspiration to do better and support the rare people who do this important work!
For more in Caro, see friend-of-the-show Dwarkesh Patel’s excellent post on the Lyndon Johnson series.
💚 🥃 You can get more and support the project by becoming a paid sub:
🤖 A Word From Our Sponsor: ChartORG.AI 🤖
Take your ChatGPT experience to the next level with ChatORG.ai's real-time collaboration-enhanced platform. Our powerful tool allows everyone on your team to share and collaborate in multi-player ChatGPT conversations, in real-time on a platform built on OpenAI’s GPT-3.5 and GPT-4 🚀
Preserve insights: ChatORG.ai keeps all your ChatGPT chats neatly organized with shared folders. 🗄️
Ensure consistency with a unified approach: If your team frequently uses specific prompts, the prompt library is a game-changer. 📚🏆
To cap it all off, ChatORG.ai offers custom assistants 👩🏻🍳👨🏻🔧👩🏻💼👩🏻🚀. Tailor these AI-powered helpers to suit your team's unique needs and make them an integral part of your workflow 🤖 🛠️
⭐️ Start FREE today at ChatORG.ai/Liberty ⭐️
🏦 💰 Liberty Capital 💳 💴
📺 YouTube’s $40 billion in revenue 💰💰💰💰
These numbers are just bonkers:
YouTube generated $40 billion in revenue across all its products in the last 12 months ending March 31, including subscriptions from YouTube Premium and TV, which it previously had not included when reporting revenue, CEO Neal Mohan said Thursday at a conference hosted by research firm MoffettNathanson [...]
Most of YouTube’s revenue continues to come from advertisements played before and during YouTube videos. But Mohan said more is also coming from ads on its short-form YouTube Shorts as well as its paid subscribers. Some 80 million of its users subscribe to YouTube Premium, which sells access to ad-free music and videos, while YouTube TV, its television streaming service, has more than 5 million paid subscribers, Mohan said. YouTube accounts for more TV viewing than any other single network or streaming service (Source)
Isn’t it about time for Google to start disclosing more about YouTube in its financials?
In the same way that Microsoft should break out Azure fully, it’s getting a bit ridiculous to have what would be one of the world’s biggest media businesses be publicly traded yet so opaque.
📺 Zaslav’s Content Power Law 🎥 🎬 💸
The media space is in a strange transitional phase these days, caught in-between old models and paradigms and the messy, fuzzy shape of what is emerging to replace them.
Video content is going through some of the same changes that newspapers and magazines first went through, then music. Everything is atomizing, bundles are fraying, and it’s not yet clear what is replacing them. This makes the managers of large content companies uncertain of what they should focus on.
David Zaslav, who is CEO of Warner Bros. Discovery, recently said: "Five shows are 90% of what people are watching on HBO Max; and if we take it to 20 shows, it's 98% of what people are watching."
I’m sure it’s technically true, but what you do with that information is what ultimately matters. It’s easy for someone with a purely financial or engineering mindset to think:
Well, why are we paying for all these other shows then? Let's just get rid of all that, keep the top ones, and we'll barely lose anything while dramatically lowering our costs! That'll boost profits!
What makes this difficult to execute is that if people knew how to only make blockbusters reliably, everybody would do it. There would be plenty of studios with only successes.
Don’t get me wrong, it’s possible to find a formula that works well on average, and sometimes when the stars are perfectly aligned with the right creative people in the right places, you can have impressive streaks of success, like the early Pixar films, or the Marvel Universe blockbusters. But that never lasts forever. Pixar eventually made some mediocre films and is now more hit & miss, and the Marvel formula is running out of steam... You can’t squeeze the same lemon forever.
Trying to reduce everything to a few dimensions risks losing a lot of important nuances.
Anyone who tries to generate just the peak of the power law without the long tail risks having neither, as one emerges from the other.
To build something, intensity and time matter.
When a formula runs out of steam and you haven’t been experimenting, allowing a new generation of talent to gain experience, and trying lots of things in the meantime, you haven’t put in the work to grow your next franchise or formula. Instant hits happen, but many of the best creators (writers, directors, actors, etc) need some time to reach escape velocity.
You also need something from the audience: people who intensely love something give you the signal that you’ve hit something interesting. These people will stick around, spread the word, evangelize the thing, and over time it can grow into something big.
I think that a lot of modern executives would've canceled The Wire and Breaking Bad after 1 season (if they had greenlit them at all)... Even Seinfeld had a slow start during its first couple of seasons but grew into a monster over time. Does anyone at big studios have the patience for that today?
🇰🇷 Samsung to Google: Just kidding!
There were headlines for a while that Samsung was considering switching to Bing as a default search engine on its phones.
I always suspected it was a negotiation tactic to try to extract better terms from Google, but who knows, Microsoft has been a lot more aggressive in trying to gain market share ever since they incorporated OpenAI LLM technology into Bing, so maybe they were making Samsung an ‘offer they couldn’t refuse’. 🐴
We may never know for sure, but apparently, this round is over:
Samsung, the world’s largest smartphone maker, has suspended an internal review that had explored replacing Google with Bing on its mobile devices, the people said. [...]
Samsung shipped roughly 260 million smartphones last year, according to tech-market researcher Counterpoint Research, accounting for roughly one-fifth of the worldwide total. Google has served as the default search engine on Samsung’s smartphones since its first model—the Galaxy S—was launched in 2010. (Source)
Maybe it was Google that made them an offer they couldn’t refuse?
The change was always overblown anyway. It wouldn’t have changed that much, since it would’ve been for Samsung’s own browser, which isn’t what most of its customers use anyway. Google has distribution tendrils at multiple levels, and the Chrome browser is a powerful one.
🍎 Apple makes deal with Broadcom for wireless chips ‘Made in USA’ 🇺🇸
Today Apple announced a new multiyear, multibillion-dollar agreement with Broadcom, a leading U.S. technology and advanced manufacturing company. Through this collaboration, Broadcom will develop 5G radio frequency components — including FBAR filters — and cutting-edge wireless connectivity components.
The FBAR filters will be designed and built in several key American manufacturing and technology hubs, including Fort Collins, Colorado, where Broadcom has a major facility.
This is part of Apple’s ongoing push to diversify suppliers out of China and to eventually replace Qualcomm for its cellular modems.
🎙️🇸🇪 Interview: Spotify CEO Daniel Ek on Podcasts and Audiobooks
A really good conversation! Friends-of-the-show David (💚 🥃) and Ben went to Sweden to sit down in Spotify’s studio with Daniel Ek:
As much as I dislike some aspects of Spotify’s podcast-listening experience, I’m very happy that they’ve grown the podcasting pie and introduced the medium to so many new listeners.
Starting from way behind with ~zero market share in 2018, Spotify has now aggregated the [podcast] listener market and amazingly surpassed Apple as the world’s largest podcast platform — including close to home with the Acquired audience, where it has 60%+ market share among you all!
It will be interesting to see if they can shake up the audiobook market and bring some much-needed innovation to the space, both in terms of the business model and the content.
🪦 🔎 Neeva shutting down consumer search, will focus on enterprise search
One more notch on Google’s belt when it comes to defending against new entrants.
Not that it was ever very likely that Neeva would get so much traction that it would hurt Google, but a few years ago when money was free and optimism high, it seemed possible to counter-position and build a niche ad-free/subscription-based search engine.
Customer acquisition is apparently the real killer:
throughout this journey, we’ve discovered that it is one thing to build a search engine, and an entirely different thing to convince regular users of the need to switch to a better choice. From the unnecessary friction required to change default search settings, to the challenges in helping people understand the difference between a search engine and a browser, acquiring users has been really hard. Contrary to popular belief, convincing users to pay for a better experience was actually a less difficult problem compared to getting them to try a new search engine in the first place.
There are rumors that Snowflake may acquire Neeva. This isn’t over.
🇨🇳 China's imports and exports
In 2022, China’s imports totalled $2.71 trillion, and exports $3.59 trillion, leaving it a hefty $878 billion trade surplus.
It’s no surprise that the USA is China’s largest customer, importing $582 billion worth of goods and services from China in 2022.
🧪🔬 Liberty Labs 🧬 🔭
🎨🤖 DragGAN: Deforming images using generative AI
There’s nothing like seeing this one in action, so please check out the video above. It’s an incredibly powerful way to modify images and I expect it’ll become huge.
If you want to dig deeper into the technical side, the paper is here.
💊💉🤷♀️ Ozempic: ‘Did Scientists Accidentally Invent an Anti-addiction Drug?’
Interesting piece in The Atlantic about the side effects of the weight-loss drug Ozempic (which is a brand of semaglutide) on addiction-related behaviors:
All her life, Victoria Rutledge thought of herself as someone with an addictive personality. Her first addiction was alcohol. After she got sober in her early 30s, she replaced drinking with food and shopping [...]
Earlier this year, she began taking semaglutide, also known as Wegovy, after being prescribed the drug for weight loss. Her food thoughts quieted down. She lost weight. But most surprisingly, she walked out of Target one day and realized her cart contained only the four things she came to buy. “I’ve never done that before,” she said. The desire to shop had slipped away. The desire to drink, extinguished once, did not rush in as a replacement either.
Users of the drug are reporting an effect on a wide range of addictions:
They have reported losing interest in a whole range of addictive and compulsive behaviors: drinking, smoking, shopping, biting nails, picking at skin. Not everyone on the drug experiences these positive effects, to be clear, but enough that addiction researchers are paying attention
While that sounds great, the fact that we don’t really know how it works and how it impacts the brain should make us careful, especially since it’s not a once-and-done type of treatment:
In diabetes and obesity, semaglutide is supposed to be a lifelong medication, and its most dramatic effects are quickly reversed when people go off. “The weight comes back; the suppression of appetite goes away,” says Janice Jin Hwang, an obesity doctor at UNC School of Medicine. The same could be true in at least certain forms of addiction too.
h/t my friend MBI (💎🐕)
🧠💡 Interview: Andy Matuschak on intellectual exoskeletons and tools for thoughts
I enjoyed this episode of Joe Walker’s Jolly Swagman podcast:
“Andy Matuschak is a developer and designer. He helped build iOS at Apple, founded and led Khan Academy's R&D lab, and now works as an independent researcher investigating tools for thought — that is, technologies that can transform human cognition and creativity.”
The discussion on “tools for thought” and mechanisms to improve memory retention, and how that’s linked to creativity and better thinking (since thinking is largely remembering and making connections between those things we remember) was fascinating.
The idea of building-in such tools into books is very interesting, and I hope it catches on.
🇫🇮🔌⚛️ ‘Nuclear power helps bring down electricity prices by 75% in Finland’
The start of a much-delayed nuclear plant in Finland has helped bring down electricity prices by more than 75 per cent in the Nordic country.
The Olkiluoto 3 (OL3), Europe’s first new nuclear plant in 16 years, began operating in April and is capable of meeting up to 15 per cent of the country’s power demand. Nuclear made up a third of Finland’s total electricity generation in 2021.
Average spot electricity prices in the country fell to €60.55 ($65.69) per megawatt hour in April from €245.98 per megawatt hour in December, a decrease of 75.38 per cent (Source)
Prices even went negative for a bit:
Early on Wednesday the market price for electricity dropped below zero cents per kilowatt-hour (kWh) and for hours after that the price was only 0.3 cents per kWh at its highest, according to the country's grid operator, Fingrid.
So many things went wrong while building this plant, and yet, it’s still pretty amazing. Mark Nelson puts it well:
Finland spent 19 years building the world's slowest, most over-budget nuclear reactor. Was supposed to cost €4 billion. Cost €12 billion. Construction companies and the reactor builder took much of the hit. Finns get dirt-cheap electricity, €45 per MWh, for a century.
Imagine if we got good at building nuclear power plants because we had a sustained program of construction to build many many units of the same design sequentially so that suppliers and construction companies could get good at it (rather than the one-offs that we’ve been doing).
This isn’t science fiction: Many countries did it decades ago when they were much poorer and still using slide rules. In the same way that we lost space rocket capabilities over time but regained it in recent years, we could do the same with nuclear power.
Combine that with more rational regulation, and we’d have abundant, cheap, clean power for everyone. Many believe we need fusion for that, but fission offers pretty much the same benefits if only we decided to go for it and stop self-sabotaging.
For more on this topic, see my conversation with Mark Nelson:
🎨 🎭 Liberty Studio 👩🎨 🎥
🎮 A Doom Map that is… CONCEPTUAL ART?!? 🤯
Liminal spaces, impossible geometries, concept art about love & loss, the nature of memory, and the human condition executed at a high level…. using a 30-year-old game!
This may not seem like much, but I encourage you to check out at least the first 20 minutes to get an idea (you can watch the video at 2x, so that’s just 10 minutes). But if you like the start, it only gets stranger and stranger after that…
I never thought a Doom map would make me want to read House of Leaves. If you’re not familiar with ‘ergodic literature’, this video will give you an idea of just how weird and disorienting it is.