Azeem Azhar’s Weekly Wondermissive: Future, Tech & Society
This issue has been supported by our partner: Ocean Protocol.
Four weeks to network launch! Take part in the new data economy.
You are reading the free version of Exponential View.
While this version will stay open for all, please support my work by recommending EV to your network.
Simply retweet this if you’re on Twitter or forward this email to at least three of your friends if you’re not on Twitter. Thanks!
Dept of the near future
🔧 Peak car. We’re approaching peak car, and annual sales are slowing as we near the 100m mark. This is a worthwhile read but makes two errors. It fails to address micromobility, which will be a low-end disruptor more important that autonomous robo-taxi networks in many cities. And it fails to estimate the transition away from ICE vehicle car ownership, which I believe will come closer than even Bloomberg’s estimates. (With bankruptcies aplenty, and questionable unit economics, micromobility still has a ways to go. An on-point economic analysis by Sam Korus here. But the shift from cars is showing up in declining petrol demand.)
💯 Shoshana Zuboff: The real reason why Facebook and Google won’t change. Excellent essay.
🚼 Emily Mullin: We’re already designing babies through preimplantation genetic testing which gives parents an informed choice of what predispositions they want in their child. Superb read.
⏳ Taylor Pearson’s long read on the power of the market and how blockchain may enable the “death of the firm” and a new type of political economy.
Public blockchains make it possible to have ledgers that are trustworthy without requiring a centralized firm to manage them. This would shift the line further in favor of “renting” over “buying” by reducing the transaction cost of trust.
🤔 Sam Altman, who founded the Y Combinator, the world’s leading startup incubator reckons: “So there will be less jobs but the amount of global wealth will just skyrocket [...] but a lack of material abundance will not be a problem.” This arguyment, that abundance solves all, is now so thin. There is a much harder question about the distribution of this wealth and what happens if we get it wrong. We even get hints of what abundance looks like today. US GDP per capita is more than $55k per annum, and the adult American works three hours a day on average (as I discuss with Jeffrey Sachs here), so we already live in an era which is abundant by all historical standards. We still have problems.
✨ 🔟 Bill Gates writes about his top ten breakthrough technologies, which include robot dexterity, nuclear power, preterm prediction, custom cancer vaccines, carbon capture & lab-grown meat. I found myself disagreeing with the importance of a number of these. For example, engineered carbon capture which, even with economic large scale storage solutions, feels like it’ll alway be far too small scale to make a difference, compared to large scale decarbonisation of energy & reforestation, even given current concerns of the real impact of planting trees. (Read a follow-up interview between EV subscriber, Gideon Lichfield & Bill Gates.)
🔥🌎 Burning planet: 412.15ppm
Each week, I’m going to remind us of our level of the CO2 in the atmosphere. We must avoid a level of 450 parts per million.
🇦🇺 Climate disruption? Australia just had its hottest summer on record by a large margin: Dec 2018 – Feb 2019 was 2.14°C warmer than an average Australian summer.
Analysis of the 1983 EPA report, “Can we delay global warming?”
On China’s high-voltage DC smart supergrid. More than 24 TWh moved: “the sheer scale of the new line and the advanced grid technology that’s been developed to support it dwarf anything going on in pretty much any other country.”
Dept of the future of work
Casey Newton’s searing investigation of the teams of contractors moderating Facebook’s worst content brings to life the struggles of contractors and other non-employed labour. It is a must-read, both to reflect what humanity chooses to bubble up (unpleasant enough), and Facebook’s response to it (terrible). But I’m particularly interested in the future of work dimension.
Why does Facebook outsource some content moderation to a staffing firm? Facebook decided that the moderators, who ultimately are the thin blue line between a product with integrity and decency, and one showing the very worst of humanity, should not be Facebook staff.
Remember that Facebook staff are treated like prize ponies, with a median salary of $240,000 per annum and perks which typically include free bike repairs, an egg-freezing allowance, free food & onsite barbers. But not so the contractors. These contractors were paid $28,000 a year and were allowed nine minutes of wellness time a day (to recover from the murders and bestiality they would have had to watch.)
Facebook has 15,000 contractor moderators around the world. Some in India, who are responsible for policing Arabic content, make $1,404 annually. Now, Facebook is profitable enough to employ, with usual benefits, all its moderators. Even at median rates of pay, 15,000 employed moderators would only cost the firm $3.6bn per year. To give moderators 2-3 hours of wellness or recovery time a day would cost the firm much less. Facebook’s 2018 EBITDA was $25.4bn.
The question is, how do you make the future of work, one where we have advanced automation and sophisticated marketplaces, good work?
Dept of artificial intelligence
Causality in machine learning, as pioneered by Judea Pearl, is going to be an exciting area to watch. Here is an excellent overview of why it matters by Seth Hendrickson:
A machine learning model that captures causal relationships of data is one way to ensure that the model will generalize to new settings, one of the most difficult aspects of machine learning.
An amazing new feature in Microsoft Excel: snap a picture of a printed data table to convert it into editable spreadsheet.
🏁 May Mobility’s Edwin Olson argues that exponential improvements in disengagements will be self-driving cars’ Moore’s Law.
How to scale an AI startup by Matt Turck.
US Cyber Command launched attacks on Russian’s Internet Research Agency around the time of the 2018 Midterms.
🗯️ How recommendation algorithms favour conspiracy theories and outrage over fact.
🛏️ This Airbnb rental does not exist: GANs strike again.
Chinese farmers are using biometrics to track pigs. Swine flu outbreaks are a significant risk to Chinese food security.
Nice summary on Artificial Intelligence and human accountability by Joanna Bryson.
Short morsels to appear smart at dinner parties
Understanding the impact of austerity on young Britons. (Excellent NYT report.)
👏 The University of California cancels its Elsevier academic journal subscriptions, in a brave move tackling the terrible business practices of the domimant academic publisher.
🐀 Human-rat cyborgs.
Chinese movies are winning Oscars for the first time in history.
Nike’s sneakers need a software update.
🐙 💤 Do octopuses dream?
🗺️ The average civilisation lasted 336 years.
A beautiful paean to the periodic table. 150 years old this week. Luv ya! 🤩
I am going to be at South by South West this year. I am speaking on AI and Democracy on Monday and on AI and artisanship on Sunday. If you want to meet up informally, then drop me a mail. Looks like we have enough interest to have a small group.
I’m also putting together the last few guests for season three of the Exponential View podcast and I’m interested in getting a few of the following guests on the podcast. If you can make a personal introduction, I’d appreciate an intro to Judea Pearl, Fei Fei Li or Yann LeCun.
Have the bestest of weeks,
💌 P.S. If you enjoy reading Exponential View, and you’re on Twitter, please show your support by retweeting this. As usual, scroll down for notes from our readers.
This week’s Exponential View has been supported by our partner: Ocean Protocol.
Ocean partnered with AI Singapore to develop AI research talent and accelerate AI innovation. Find out more
What you’re up to — notes from EV readers
Lawrence Lundy-Bryan: "We need a crypto trias politica."
Peadar Coyle has launched an online course on probabilistic programming.
Michael Froomkin invites you to join him and a group of policy-makers and academics at We Robot conference in Miami to discuss the issues and opportunities technology will face over the next 5-10 years.
Spencer Bailey spoke to a number of AI experts about how they're shaping the world using this technology.
Nick Burnett is asking the Mindhive: What terms best describe the skills and mindsets needed to thrive in a digital and automated future? Chime in.
Congrats to Archie Norman for securing a £1.8bn tie up for Marks & Spencer with robotically-enabled food delivery service Ocado.
Rowland Manthorpe's excellent overview of predictive policing in the UK (with screenshots).
Have something to share? Send it to email@example.com.