Hiding Likes, Podcasting Growth, and a $270 Toaster

Welcome to Links Variety Pack #1


I’m Ishtaarth, and welcome to Links Variety Pack #1.

After walking divergent paths to collect information in one place, this newsletter is my breaking point. Unless you’ve been living under a rock, the humble newsletter is seeing a renaissance. It follows the Dark Forest Theory of the internet:

In response to the ads, the tracking, the trolling, the hype, and other predatory behaviors, we’re retreating to our dark forests of the internet, and away from the mainstream.

I have modest ambitions for this. I hope it serves as an effective public record of my understanding of what’s going on. There’s no format. It’s not a blog post and it’s certainly (hopefully) not just a link flittering away before you declare email bankruptcy. It’s a blog-ish newsletter that can go from anywhere from 100–1000 words, and reach anywhere from 1–100 people. It’s liberating to not fully know the constraints of the medium you’re working with every once in a while.

Let’s begin.

John Herman of The New York Times looks at the paradox of metrics on Instagram, addressing a world without like counts:

Likes and retweets used to be translated into signals for people. Now they just provide signals for the machines. In the newer, less immediately comprehensible versions of our feeds, where what we see is chosen according to processes that we can only guess at, Twitter and Instagram suggest the existence of dozens of metrics that are hidden by default. If our feeds are already assembled based on cues from countless secret metrics, why not hide a couple more?

The response to Instagram’s announcement has been muted, aside from some panic among professional influencers. 

This is certainly true for Facebook and Instagram’s real customers, advertisers. With a change in the season, Facebook usurps expectations on metrics. “It’s not Page Likes! It’s an Indicator of Organic Reach!” “It’s not just Shares or Likes! It’s Brand Engagement,” and so on. As engagement plateaus, removing like counters might be the last straw in maintaining the News Feed’s uncompromising position as a brand awareness channel. We’ll see.

Related: A world where everyone’s an Instagram Influencer.

The Mitsubishi TO-ST1-T bread oven costs $270. It only makes one slice at a time.

Communications and transportation technologies have been on the top of my mind lately. A fascinating account of the invention of GDS (the Global Distribution System) which powers the entire OTA industry:

IBM and American began research on the system almost immediately. 7 years and 40 million dollars later ($350M adjusted for inflation), SABRE (Semi-Automated Business Research Environment*)* took flight. At the time, it was the largest real-time data processing system outside of the US government, and was the first major e-commerce system ever, processing many millions of dollars per day. This was all achieved before the Internet. […]

It’s worth noting that the most powerful industry software, even today, is command-driven.

Related: IBM in-house posters.

The growth of podcasts as a sustainable media industry continues unabated in the United States. As a model that started out seeking respite from AdTech, the current state of podcast advertising is reflective of the broader internet. Andreessen Horowitz has a comprehensive investigation of the podcasting market.

Highlights: China leads with monetisation; Spotify has doubled its market share (to 9% since 2015) and is out for more with a dedicated podcasts tab in the west. (Spotify, also, the subject of Liz Pelly’s ire in The Big Mood Machine.)

India Roundup:

With the impending release of macOS Catalina, iTunes is finally making its way to software heaven. iTunes was the first piece of software I felt indebted to. Back when Apple made software it was proud to market, it was a key part of the iPod + iTunes package. It helped lead Apple’s resurgence to a consumer electronics company. The launch video is classic.

100% of all eulogies make the claim that iTunes grew up to do too much, all at once. And soon, it will be sent into software heaven. Craig Federighi’s announcement at WWDC was the perfect set up, but incomplete in paying its dues to the music app that could—and did. The Verge—helpfully—has a timeline of iTunes, and Kevin Roose takes it a step further with his eulogy:

I’ve come to think of iTunes as a core piece of what I call the Middle Internet — the period between the Wild West days of Napster and the hyper-centralized era of Facebook and YouTube.

It was an era of clean, well-lit marketplaces where people could buy things to listen to, without worrying about buffering or corrupted files. It filled an important technological gap in the period when lots of people had internet access, but few people had smartphones with data plans capable of streaming high-quality media on demand.

And it was a time when people actively curated their own online media, rather than having it algorithmically spoon-fed to them.

R.I.P. (mix, burn) iTunes.

This is far belated, but in her annual Internet Trends presentation Mary Meeker packed a plethora of insights, as expected. In her keynote, Meeker bills it as a report meant to be read, not presented, so here’s Recode with the analysis and the slides.

From the Machine Learning coffers at Stanford:

In television and film, actors often flub small bits of otherwise flawless performances. Other times they leave out a critical word. For editors, the only solution so far is to accept the flaws or fix them with expensive reshoots.

Imagine, however, if that editor could modify video using a text transcript. 

Twitter accounts to follow: @visakanv, @aaronzlewis, @naval, @david_perell.
Newsletters to subscribe to: The Margins, Why Is This Interesting, and an all-time favourite in NextDraft.

You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.
—Buckminster Fuller

Loading more posts…