On the automation, jobs and work. The bright future after Moore’s Law. Understanding decentralisation. The crisis in democracy. The crisis in capitalism. Precision farming. Oxygen forests. Origins of morality.
The 100th Issue of Exponential View. Have super conversations!
Dept of the near future
🌟 “The end of Moore’s Law … will bring about a golden new era of computer architecture” STELLAR essay by MIT’s Rodney Brooks.
💡 Vitalik Buterin: The meaning of decentralization GREAT READ
💰 The crisis of capitalism by Simon Tilford. “Globalisation did not force governments to adopt policies that divided their countries, exacerbated inequality and hit social mobility.” (See also: workers are losing out as monopolies corner the market by Robert Colville.)
🚨 On the decline of support for democracy in the West: what is causing it and how to address it. THOUGHT PROVOKING (Also, the income share of the bottom 50% of Americans is collapsing. And also Deepmind explores multi-agent cooperative games which shows some analyses on how cooperation or social aggression emerges.)
🔮 What Magic Leap is building. Big demo this week, apparently.
Dept of automation, jobs & work
I’m exploring factory automation and what it means for meaningful work. There are a whole bunch of issues that I’m wrestling with right now so I thought I’d share some early, incomplete, thinking.
I tweeted a story about a Chinese factory which replaced a 90% of its workers with robots and saw output rise and defects drop. It struck a nerve with more than 1,000 retweets. This is part of a long-term trend of automating work that dates back to flint heads and the plough.
Even in the context of modern factories, the desire to go manless goes back decades. Here is the New York Times from 1981 writing about Japanese manless factories:
Robots dramatically improve industrial productivity and quality control. They are expected to widen Japan’s current trade advantage over the United States. They are also expected to compensate for a severe labor shortage in Japan and to rescue the small-business man, a significant economic and political figure in Japanese life.
Lights-out manufacturing (lights out because automated production lines don’t require people) are pretty common, even in the heartland of middle America. Two such firms, glad of their lights out capabilities include Staub and Makuta. (Not to single out those firms, I just found them by searching the web.)
The New York Times published on Siemens Energy plans to build a factory in North Carolina with this punchy realisation:
People on the plant floor need to be much more skilled than they were in the past. There are no jobs for high school graduates at Siemens today.
So the importance of the right kind of education is necessarily paramount. And this is against the backdrop of what seems to be a continuing enfeeblement of education. For example, that state of Tennessee admitted that a third of its high-school graduates didn’t meet the right standard (which wouldn’t get you a job in a modern factory, anyway.)
It is also worth noting that industrialisation during the 18th and 19th centuries led to a massive increase in hours worked. The typical hunter-gatherer worked a handful of hours per day during their short lives. That spiked up towards 10-12 hours at day at the peak of industrial age Britain.(Data for that here and also here.)
👉🏼 Through the gains of modern economy (automation, trade, workers rights) we’ve reduced that workload towards 7-8 hours per day. The average working Brit puts in about 37 hours a week. According to Jeff Sachs, America’s $55k per capita GDP is produced by adults working only a few hours a day:
Basically our society is so sophisticated that we can support an unimaginably high level of living with 3 hours and 10 mins.
Where are we seeing job growth? One argument is that coding is the place to go - that it will be the new blue-collar job.
My inclination is also to think of coders today as the US auto factory worker on the 1950s. Swelling into a highly-automatable sector because the technology hasn’t caught up, getting better than average salaries because of skills shortages. And just not many auto workers were Henry Ford, very few coders will be Bill Gates. Equally, block-and-tackle coders may soon be caught flat-foot by the automation of the software creation process. (This is why I have argued for promoting more flexible thinking skills than coding alone.)
- 🏢 How China’s robot revolution could change the US commercial real estate market
- 🤖 The AI threat isn’t Skynet it is the end of the middle class
- The rise of the data engineer. New type of job, I’ve found quite hard to recruit people with expertise in this area.
- Traders are out computer engineers are in at Goldman Sachs
- On Sentient Technologies AI-driven hedge fund
- Ocado testing robot hands to pick fruit. (Fascinating video of Ocado’s current warehouse robots.)
- 💡 Profile of Philippe Chabot who runs a fascinating twitter feed looking at automation. (Recommended)
Dept of renewables & climate change
🌞 Solar power in India has become the cheapest energy source with three recent bids for a 750MW contract running to 2.9c per kWh. This represents a 35% decrease on the previous cheapest rate. Fully loaded without subsidies it runs to less than 5c per kWh - and so competitive to unsubsidized coal.
Indeed, Goldman Sachs argues that “peak coal” is coming sooner than expected, which will have an impact on the US govt desire to bring back coal jobs.
- Electric battery cost declines
- Tesla will face genuine competitors in the EV sector in the coming years
- 🔥 It is 25°C higher than normal in the Arctic
- On the “horrifying heatwave” in Australia
Short morsels to appear smart at dinner parties
Tren Griffin: A dozen lessons on growing startups
👨🏽🌾 Introduction to precision farming. (A farmer, an engineer and a computer scientist walk into a field…)
⚓️ The ocean freight shipping market clear & interesting overview by EV subscriber, Brian Laung Aoaeh
🐩🐒 On the origins of morality: pets and monkeys show a preference for people who help others
Brain scanners. The future big brother?
Replacing the engine of a triple-7 that made an emergency stop on a remote Arctic island
Very clever introductory lecture series to deep learning. (I have only read lecture 1a so far.)
One hundred issues!
Wow. When I kicked this off, I wasn’t sure where we’d get to. I have learnt a lot in the last 100 weeks, and slowly been able to build a better understanding of the coming years. I’ve also been lucky to make some great friends and meet lots of fantastic supporters of Exponential View.
P.S. I have been trying Syncbot, which creates playlists to help you work. It’s made by EV reader Marko Ahtisaari. I really like it.