18: My Thoughts on Spotify & Podcasting, TikTok Madness, F-16 A.I. Dogfights, and Amazon Cancels Iain M. Banks TV Series

"this echo of their brain waves out in the world, replaying in other people’s brains"

I was gratified to be able to answer promptly, and I did. I said I didn’t know. —Mark Twain, "Old Times on the Mississippi"

One of the things that I like about some podcasts that I’ve been following for a very long time, especially the “ongoing conversation” format, is that after having someone in your ears for hundreds of hours, you build up a huge shared knowledge base (even if it’s uni-directional and they know nothing about you).

It’s a tangled web of inside-jokes and details about people’s personalities, anecdotes about their lives or opinions about various things, their sense of humor, how they phrase certain things or think about the world or solve problems. Your mental model is getting to be of a pretty high-resolution after a while.

I was thinking that the same kind of thing may happen if I keep doing this newsletter for a while. There’s bound to be a few people who will have been reading from the start (and maybe have been following me on Twitter or forums before that), and after having read thousands and thousands of my words, we’ll have a shared knowledge base.

The effect is probably not as strong with text as with the human voice, but still… Anyone who puts stuff out there has this echo of their brain waves out in the world, replaying in other people’s brains. Kind of strange when you think about it like that.

Investing & Business

Facebook Warns Advertisers of the Impact of iOS 14

In June at WWDC, Apple announced that iOS 14 would require apps to ask users for permission to collect and share data using Apple’s device identifier, something that advertisers aren’t used to doing (with both pros and cons… A lot of “pro privacy” regulations sound good on paper, but end up like GDPR, mostly annoying everyone with buttons to click… People rapidly become blind to them and just go “yes, yes, yes”).

Facebook is now warning its customers (the advertisers… the users are the products, from a business model perspective), in a blog post, that ad performance for its “Facebook Audience Network” (which allows the targeting of ads outside of FB properties — you can see examples of what that looks like here) is about to take a dive with users of iOS, which is the most lucrative segment of the market.

We expect these changes will disproportionately affect Audience Network given its heavy dependence on app advertising. Like all ad networks on iOS 14, advertiser ability to accurately target and measure their campaigns on Audience Network will be impacted, and as a result publishers should expect their ability to effectively monetize on Audience Network to decrease. Ultimately, despite our best efforts, Apple’s updates may render Audience Network so ineffective on iOS 14 that it may not make sense to offer it on iOS 14. We expect less impact to our own advertising business, and we’re committed to supporting advertisers and publishers through these updates.

In a different post they write:

in testing we’ve seen more than a 50% drop in Audience Network publisher revenue when personalization was removed from mobile ad install campaigns. In reality, the impact to Audience Network on iOS 14 may be much more, so we are working on short-and long-term strategies to support publishers through these changes.

I wouldn’t be surprised if over time a new standard emerges that allows ad-targeting but with built-in privacy features and controls. Cookies and device IDs were created a while ago and weren’t designed with the modern world in mind (kind of like email and phones suffer from spam and privacy issues because they weren’t built for a zero-trust environment).

Apple has a big app business and has incentives to keep allowing the app install ads to flourish, and the in-app advertisements to keep working well to help finance developers. I think we’re in a transition period.

We all deserve more security, privacy, and control over our data. But as Jeff Green would say, we also all benefit from the quid pro quo of the internet where massive amounts of content is paid for through advertising, and there’s counter-pressure to maintain that exchange of value.

Bill Gurley on SPACs

Here you go.

Amazon Announces Halo Wearable & ‘Fresh’ Grocery

Amazon is aiming to crack open a new segment of the U.S. grocery industry with the launch of its Fresh store, bringing the e-commerce giant into more direct competition with chains like Kroger and Albertsons.


Amazon Halo Band is purpose-built to focus on your health and wellness—unlike smartwatches and fitness trackers, it doesn’t have a screen or constant notifications. The small sensor capsule delivers highly accurate data, and includes an accelerometer, a temperature sensor, a heart rate monitor, two microphones, an LED indicator light, and a button to turn the microphones on or off


Features that I’m most curious about: Sleep tracking (how accurate is it vs an Oura Ring?) and “emotional” tracking (via voice tone).

TikTok: Masa Hates Feeling Left Out, and Out of Nowhere, Walmart Appears…

The Information reports:

As President Donald Trump’s September deadline for TikTok to sell its U.S. business approaches, investors are scrambling to assess or participate in a deal. One party that has recently emerged as a possible buyer: SoftBank. [...]

President Trump has made it clear he wants TikTok’s U.S. business in U.S. hands, so it is unlikely that SoftBank, a Japanese company, could lead an acquisition. [...] It couldn’t be learned whether SoftBank was exploring teaming up with another entity on an existing bid or creating a new effort. SoftBank’s founder Masayoshi Son is close to Oracle co-founder and chairman Larry Ellison.

Then TikTok CEO Kevin Mayer left the company… Some reports say that it’s because he was “excluded” from the merger talks.

Soon after that, it was reported that “TikTok is nearing an agreement to sell its U.S., Canadian, Australian and New Zealand operations in a deal that could be announced as soon as next week” with a price in the “$20 billion to $30 billion range”, but the crazy part was that:

Walmart had been working with SoftBank on a bid for TikTok

What? Walmart?

Then The Wrap reported that Oracle was close to a deal, with White House support:

The proposed deal would comprise $10 billion in cash, $10 billion in Oracle stock and 50% of annual TikTok profit to flow back to TikTok’s China-based parent company, ByteDance, for two years, according to one individual with knowledge of the deal.

But then, another twist! Walmart said it was working with Microsoft:

“We believe a potential relationship with TikTok US in partnership with Microsoft could add this key functionality and provide Walmart with an important way for us to reach and serve omnichannel customers as well as grow our third-party marketplace and advertising businesses,” it said. “We are confident that a Walmart and Microsoft partnership would meet both the expectations of US TikTok users while satisfying the concerns of US government regulators.”

And then we learned that Alphabet/Google was also going to be part of the Walmart consortium..

Rapidly turning into this:

Salesforce Q2

  • Second Quarter Revenue of $5.15 Billion, up 29% Year-Over-Year, 29% in Constant Currency

  • Raises FY21 Revenue Guidance to Approximately $20.7 Billion to $20.8 Billion, up Approximately 21% to 22% Year-Over-Year

The stock was up over 25% on the news, getting close to $250bn market cap… Lifted up the whole software sector in sympathy.

Someday, I should go back and read threads on value forums from about a decade ago when CRM was the poster-boy for “insane valuations” and “an obvious short” that “isn’t making any profits”.

Always useful to go back and see what people were really thinking at the time and not colored through today’s hindsight bias.

Charter’s Tom Rutledge on his Way to the Billionaire Club


The stock reached that level at the end of July and has been above since. So a little more is required, but his odds are looking mighty good right now (hopefully I’m not jinxing it).

Spotify Honeymoon Over for Joe Budden & My Thoughts

One of Spotify’s podcasting exclusives is getting out, with this to say about the experience:

“September 23rd, I cannot tell you where this podcast will be,” [Budden] says. “But as it stands, I can tell you where it will not be, and that is Spotify.” [...]

He claims his show exceeded Spotify’s audience reach expectations by 900 percent to the point that his listeners crashed the platform.

Still, he says he never received a bonus, and the company wouldn’t allow him and his team to take vacation days on Christmas and New Year’s Eve, because that would have required them to miss two episodes. While the company wouldn’t pay them actual bonuses, it offered to give them Rolexes instead, only to say the watches they picked out were too expensive. Then, he suggested Spotify give money away to their fans for Christmas instead. The company declined.

“That was the first time it dawned on me that Spotify is pillaging,” he says. “You pillage the audience from the podcast, and you’ve continued to pillage each step of the way without any regard for [the fans.]”

I don’t know the whole story, but these are strong words about the experience that may make other creators with large audiences think twice…

He says compared to two years ago, when he signed to Spotify, podcasters can find better deals from multiple companies, and he suggests that Spotify’s only interested in finding new popular shows rather than feeding the podcasting “ecosystem.”

Everybody’s not looking to feed the soil, some are just looking to take the fruit,” he says.

My Worries about Spotify & Podcasting

My primary problem with Spotify’s foray into podcasting is that when it comes to creation and innovation and exchange of ideas, I like an open ecosystem, like the open web, much more than a walled-garden, like Facebook.

I’ve seen this movie before. Years ago when Facebook started courting publishers and creators, they dangled some pretty good carrots. You make a page, you get tons of followers, we drive tons of traffic to you. Everybody’s on Facebook, so when they follow you, we push your content to them (they may otherwise not see it), your content can go viral, etc. You’d be crazy not to join, right?

So every creator and publisher got on Facebook. Once they were all in, the screws began to tighten. Organic reach started to go down and down and down… So if you have 100k followers, at first you publish something, and your 100k fans see it. Then only 20% see it. Then 10%. Then 1-2%. Oh, you want more of “your” followers to see what you publish?

Buy ads to reach them.

So Facebook basically took everybody’s audiences hostage and asked a ransom to reach them.

Spotify podcasters may think it’s “their” audience, but once you’re not in control of the RSS feed and the relationship with the advertisers anymore, Spotify is in between you and your audience. It’s probably a good business for Spotify, but for creators and listeners it may go in directions they don’t like…

I’m afraid that Spotify will look super nice to podcasters for a while, but over time, if they get a major share of the ecosystem, we’ll see the screws begin to tighten and they’ll do what’s best for Spotify, which isn’t necessarily what’s best for podcast listeners and creators.

Once you have one player that controls most discovery and distribution, you are at their mercy. Apple’s benign neglect of the ecosystem is exactly what allowed it to become great — if some company had tried to make it all proprietary and controlled and monetized from the start, it would be much much worse, just like if the open web had never existed and everybody had always only been on AOL.

Walled gardens inevitably change content rules over time so that certain types of content are excluded or significantly disadvantaged in discovery because it’s not what pleases advertisers... When it’s all decentralized and open, it doesn’t matter if certain things don’t please all advertisers, but when it becomes all centralized and any ad can go to any show, there’s leveling by the lowest common denominator.

What if they start looking at all content with machine-learning to look for certain keywords and play big brother so that their advertisers only support “approved” podcasts and every one else is demonetized? Ok… the chilling effect means that everybody starts self-censoring and content starts to be tailor-made to the black-box algorithm, kind of like all those Buzzfeed headlines and lists and clickbait that started a few years ago because it was the only thing that went viral on social media… So after a while, the ecosystem is a lot less vibrant, a lot less genuine and authentic, and it’s all crap made to please the Spotify SEO god that can make or break your podcast.

Oh, and if you don’t like Spotify’s player? Currently in the open podcasting ecosystem, you can use Overcast or Castro or plenty of others. It’s all RSS and Apple’s iTunes directory can be read by anyone, so anyone can make a podcasting app. They all compete against each other, a big incentive to become better.

Many innovate with smart playlists and UI, or dynamically removing silence between words so you can save time (according to Overcast, I’ve saved over 500 hours of my life just from that feature since I’ve used to app years ago), dynamic “remastering” of audio so voices are clearer (compression, EQ, etc). In a centralized world, if Spotify doesn’t have a feature you want, you’re screwed.

And currently, Spotify’s player sucks. Their speed setting goes from 1.5x to 2x without anything in between. That’s just stupid. And then from 2x straight to 3x? Do they even use their product?

They also don’t have dynamic silence removal, I’m not seeing any compression/EQ feature, I’m not even seeing show notes, though maybe they’re hidden somewhere… It’s really bad, especially for something that they’re supposed to be focusing on this much and have been working on for years.

If Spotify succeeds in dominating podcasting and becoming the controlling player, this may become a new IE6 moment for the internet. I’m not saying it’ll happen tomorrow, but it’s a possibility down the road.

Heico Fiscal Q3 Transcript Highlights

This is getting long and I’m running out of space, so I’ll just link off to my Twitter thread of highlights.

Science & Technology

MS Flight Simulator: Flying Above a Real Hurricane

How clever is the Flight Simulator team?

So many great features… Live weather. Of course it makes a ton of sense once you’re also streaming data from the cloud about the terrain! And speaking of clouds, how good is their cloud physics engine?

DARPA: AI vs AI F-16 Dogfighting, Then AI vs Human

Speaking of Virtual planes, DARPA sponsored an event called AlphaDogfight Trials where AI systems built by many organizations competed against each other in virtual close-range, visual-range dogfights. The winner then went on to fight an experienced human F-16 pilot:

Heron Systems’ F-16 AI agent defeated seven other companies’ F-16 AI agents and then went on to dominate the main event – a series of simulated dogfights against an experienced Air Force F-16 pilot – winning 5-0 through aggressive and precise maneuvers the human pilot couldn’t outmatch.

Of course, these types of aerial fights are pretty old-fashioned, but it’s still a nice proving ground for AI tech, and it’s probably easier for an AI to shoot missiles from long-range anyway… Via Fat Tail Capital

The Arts

Amazon’s Iain M. Banks Scifi Series Gets the Axe

Sad news. As a fan of his works, I was really looking forward to this.

Looks like Banks’ estate has had second thoughts (which surprises me, you know, after selling the rights to Amazon…):

The Guardian writes:

The adaptation of the Scottish author’s sci-fi books was announced in 2018, when Amazon Prime Video acquired the global rights to a TV version of Consider Phlebas, the first Culture novel. The author’s estate was set to serve as executive producer, but in a statement to the Guardian on Wednesday, it said the “timing wasn’t quite right” for it to go ahead.

“The interest and devotion that Iain’s work continues to inspire is a testament to the enduring relevance of his ideas and imagination,” said Banks’s estate. “The estate is hugely grateful for all the care and creative energy that went into the early stages of the project.”

Not sure what’s behind their thinking here, or if there’s more to it, but I hope the timing becomes right in the near future…

Here’s writer Dennis Kelly’s to Den of Geek:

“We’d talked about it for two or three years and it went a certain way along,” says Kelly. “I’d written probably 20-30 pages of the bible, but once I got a sense that it wasn’t going to happen, I had to stop writing because you become emotionally attached to the work. You think, I can’t bring myself to write work that I think may not happen, especially not that, because I loved Iain Banks. I loved those novels, I thought they were incredible.”

In the end, I just think the estate didn’t want to go through with it. It wasn’t the material. They hadn’t seen anything [he had written], it was just because I think they weren’t ready to do it, for whatever reason. I’m a little mystified myself, to be honest.

“I don’t know why it fell through, because it looked like it was all lined up to be really, really interesting but it just didn’t happen.”

Back to the Guardian:

Banks died in 2013 aged 59, two months after revealing that he had terminal cancer.

I remember the news very vividly… I can’t wait until we cure cancer.

Amazon CEO Jeff Bezos described the series as “a huge personal favourite” when the television deal was announced in 2018. Bezos is not the only billionaire fan of Banks’s sci-fi series: Elon Musk named his SpaceX drone ships after starships from the Culture novels, including the vessel Just Read the Instructions.

It probably won’t make much sense out of the context of the books, but this fan wiki page has a list of all the ship names in The Culture series.

Some personal faves:

  • No More Mr Nice Guy (GSV — General Systems Vehicle)

  • So Much For Subtlety (GSV)

  • Prime Mover (Superlifter)

  • Gunboat Diplomat (LOU — Limited Offensive Unit)

  • Size Isn't Everything (80-kilometer GSV)

  • Very Little Gravitas Indeed (GCU — General Contact Unit)

  • Attitude Adjuster (LOU, “Killer” class)

  • Frank Exchange Of Views (ROU “Psychopath” class, nominally demilitarised, but in fact a fully armed warship.)


Kiki’s Delivery Service (1989)

My family's on vacation (pandemic-style, so mostly close to home), so we had more chances to watch films with the kids lately. So after watching two other Hayao Miyazaki films (‘Ponyo’ from 2008, which I mentioned in newsletter #16, and ‘My Neighbor Toroto’ from 1988, mentioned in #17), we figured we might as well keep going…

So a few days ago, I saw 'Kiki's Delivery Service' (1989, studio Ghibli) for the first time with my 6yo boy and 8yo niece. What a masterpiece. Just delightful. The characters are resourceful, nice to each other, go after their dreams, never give up, never surrender.


There's a lot to study in there, to figure out how they did a lot of things, big and small... the visual arts and scene design is amazing (that city!), but a lot of the animation is great in subtle ways. How they gave you the feeling of flying with hand-drawn motion is just stunning.

IMDB has this on the design of the city:

During the production phase, Hayao Miyazaki and his artists traveled to Sweden to research for the film. The photographs they took of Stockholm and Visby, formed the basis of the fictional city of Koriko. The city also contains elements of Lisbon, Paris, San Francisco, and Milan.


Even the way they give you a sense of place and of direction so you aren't lost in the world makes it feel so much more real... Lots of tricks like the high-altitude views of the city and the map