🔮 Jobs or no jobs; AI’s past guides its future; against metrics; pineapples, bacteria & influencers++ #163
|Apr 29, 2018||Public post|
Dept of the near future
⏳ “The past is prologue.” Rodney Brooks aims to clarify the origins of AI, and identify just how much more exciting work there is to do. Must read.
📐 Against metrics: has the obsession with metrics-driven the short-termism and productivity slowdown afflicting many Western economies?
💳 How did China leapfrog the rest of the world in payments? A clear case of necessity as the mother of invention. Great analysis.
💈 Emerging markets can’t quit Facebook. More than 550m people in those markets use it as a commerce platform. (About a third of e-commerce happens on the platform across multiple markets.)
💯 Should central banks issue cryptocurrencies? (Tl;dr “No”, but accessible and interesting argument.)
Dept of automation & the future of work
At current employment rates, that puts 210m jobs at risk across the 32 countries included in the study;
The variance in automatability across countries is large: 33% of all jobs in Slovakia are highly automatable, while this is only the case with 6% of the jobs in Norway;
The occupations with the highest estimated automatability typically only require basic to low levels of education. At the other end of the spectrum, almost all of the least automatable occupations require professional training and/or tertiary education. (This conclusion is in common with most studies of this subject.)
The risk of automation is the highest among teenage jobs. "The relationship between automation and age is U-shaped, but the peak in automatability among youth jobs is far more pronounced than the peak among older workers. In this sense, automation is much more likely to result in youth unemployment, than in early retirements." (Anyone for angry, unemployed youth? Quick! Better distract them with video games or narcotics!)
Individuals in fully automatable jobs spend 29 hours less in job-related training annually than those in non-automatable jobs. (Key point on social stratification, class and the most vulnerable.)
🇺🇸 The US is very poorly prepared for an automation wave due to the structure of its education, suggests a new report. The main reason is the emphasis on long-college degrees rather than apprentice-oriented on-the-job training found in some European countries. This latter (closer to the push-and-pull of the market) may be more adaptable to employer needs.
Another great workplace challenge is the 100-year career. (One college degree won’t be enough.)
Automakers are embracing collaborative robots that work alongside and complement factory workers (in the same way that a clamp or screwdriver does);
Extremely interesting benchmarking between Google TPUs and AWS GPUs on image recognition tasks;
Researchers at Northwestern University used machine learning to figure out how to make new metal-glass hybrids 200 times faster than traditional techniques.
Short morsels to appear smart at dinner parties
WeWork lost nearly a billion dollars last year, and costs are rising faster than revenues. Hadn’t realised that Adam Neumann essentially controls the firm and has even leased buildings he or his family have financial interests in.
Ride-share services to make owning a car more expensive by 2027.
😎 We can do it! Beautiful graph showing how the UK reduced its dependence on coal.
😨 The twisted world of computer-generated Instagram celebrities and their real-world influence.
🚧 The threat of the loss of status--both within society and by America on the global stage—not economic hardship, or being left behind, may have been the key driver behind the election of Donald Trump.
Researchers keeping pigs brains alive outside the body.
Scientists demonstrate quantum entanglement between thousands of particles in two separate locations.
Bacteria communicate in groups to avoid antibiotics. 💉
Improving Crispr-CAS9 accuracy 10000-fold.
I was taught never to do this. Science seems to say I was wrong.
What you are up to
EV reader, Sougwen Chung, is opening a show this Sunday evening in NYC, showcasing her exploration of human-robot collaboration through art. The show is a culmination of her residency at Bell Labs, where she researched collaborative robotics, biometrics and computer vision. Stop by to say hi if you’re in New York.
EV reader, Peter Yuen, reflects on how quants in finance rapidly automated themselves out of jobs. Might the same happen with AI researchers?
Sorry, this issue is a bit late and I hope its late arrival hasn’t upset your Sunday routine. I’ve been extremely busy over the past few weeks. 🚀