This newsletter now goes to 140,000 people. Feel free to spread it around - sign up here.
✏️ My essays
Not even wrong - ways to predict tech. Link
The VR Winter. Link
COVID and cascading collapses. Link
Social voice: my old shop a16z beat off stiff competition to invest $10m at a $100m (!) valuation in Houseparty, a pre-launch social app. Yes, this is just a VC deal, but it's also interesting. Some thoughts:
- Houseparty is live voice chat rooms around people you choose to follow - a little like twitter. Open the app, see what rooms are live and who's around, go in and and just listen or join in. Is it great? It could be, certainly. Consumer social is pop culture, and you have to find some way to express or capture something at the top of Maslow's pyramid. Do you have lightning in a bottle? Can the founders capture it? Maybe.
- The dollar figure gets attention, but social apps generally have binary outcomes: they either go to zero or end up worth a lot more than this. The ones that work can return a whole fund several times over. And VCs plan for half of their portfolio to lose money and the 5% that really work to pay for the rest - you're deliberately right out at one end of the risk-reward curve. This is the business. (Also - a16z has never done one of these this early before)
- On the other hand, it is still in private beta, with only a few thousand users, mostly in Silicon Valley. All early-stage is about future potential, but you'd often expect to see an impressive usage graph to get to a number like that. This really is all about the vision, the demo and the early tests.
- So, this feels like a 2019 deal, before everything came crashing down. But it also reflects the fact that right now we really do have to do all of our social online, and so all sorts of new patterns will emerge.
- Finally, it's funny how many of the same people who've spent the last few years saying Facebook is a threat to democracy and desperately needs competition now think this is a silly deal and a trivial product.
FB buys Giphy: talking of 'trivial' products, FB paid $400m to buy Giphy, which makes the 'add a GIF with a joke/cultural reference here' across half the world's social apps (Slack is a customer, for example). This also reminds me of 2019 - it's an actual normal deal. Does raise questions of access for FB competitors that integrate it, and data flows (Slack already says no user date is involved). Link
High fashion on Amazon, hell freezes over. A bunch of small high-fashion labels are partnering with Amazon to implement ecommerce in lockdown (many deliberately are normally off-line only). Amazon's commodity retailing model has always been structurally unsuited to higher-end fashion, despite various efforts it's made over time (on the other hand it's a top 3 apparel retailer in the USA). Now that we're in the foxholes that might be on hold, but I don't know how long this will last afterwards. Link ($)
Facebook moderation issues: in the last few years FB has had to rush to add a lot more content moderation, hiring at least 30k human moderators (no, AI can't do all this). They had to look at a lot of bad stuff and it appears that in the rush FB didn't build enough process to help, and a bunch of them developed PTSD. Now there's a $52m settlement. Link
Digital legislatures: the UK Parliament has moved to remote, digital voting during the lock-down. Link
Apple VR: Apple confirmed it bought NextVR, which records sports events to distribute in VR. The price paid was apparently $100m, which is less than it raised (which points to the VR winter we're in now - I have heard elsewhere that VR sport has not worked well). It's possible Apple is planning a VR headset, but seems more likely they're sitting it out and will only do AR, which would make it harder to see why they want this. Underlying tech? Merging VR and AR experiences? Link ($)
- TSMC (Taiwanese, one of world's biggest chip makers) plans to build a plant in the USA, as part of today's geopolitical shifts. I have no idea if this is more real than Foxconn's Potemkin plant. Link
- More rumblings of an imminent US case against Google for abusing dominance in online advertising. Link ($)
- Uber may be trying to buy Grubhub to consolidate food delivery. Link
- Temasek (Singapore sovereign wealth fund) is joining Facebook's Libra bitcoiny project. Link
Politico on how Apple and Google rolled right over a bunch of other more centralised COVID contact-tracing projects. There is a real tension between how far Apple/Google are making engineering and platform decisions here and how far they're making public policy decisions (such as not collecting location). Link
However, the Washington Post has a rather odd story covering the Apple/Google virus tracking project from an entire negative slant, in which most of the people quoted genuinely do not appear to have read the technical proposals or understand what's involved at all - there are a lot of flatly false assertions made. John Gruber dismantles. Link
James Dyson in the Times on how his electric car project cost £500m and didn't work out (the car would have cost £150k). With a photo of the car. Link ($)
Ben Thomson on the continuing absurdity of newspapers demanding to be paid for links. The sad thing is that this argument was being made a decade ago, and it has never made any sense. If you want particular companies to be taxed and the money used to subsidise you, you should be honest and ask for that, instead of inventing fake business models with no economic logic. Link
Simon Andrews makes the bull case for Quibi. I am skeptical of the 'big launch with no user contact' model but am also cautious of the rather knee-jerk assumption that it makes no sense at all. Link
Shared everywhere so you might as well see it - a visual stunning demo of the next version of the Unreal game engine. Link
Yes, websites really are starting to look more similar (maybe this is the same as the way wind tunnel data made all cars look the same). Link
The impact on Manhattan if lots more people work from home. Link
How the biggest consumer apps got their first 1,000 users. Link
The Passion Economy. Link
McKinsey on fashion’s digital transformation. Link
From a decade ago: "Facebook is not worth $33bn'. We're wrong all the time, but thinking about why we're wrong has value. It's rarely just hindsight bias. Link
😮 Interesting things
This word does not exist: an ML system for generating new words that might mean something. Link
A Vitra documentary on designer chairs. Link
A blog with 79 pages of Sankey diagrams. Running since 2007 and still being updated. (I want to make a Sankey diagram for advertising since 2000) Link
IAB UK Video on Demand Trends. Link
Survey: most Uber/Lyft drivers don't want to be employees. Link
The 2020 Digital Attitudes Report. Link