Exploring the new political economy. The algorithmic enterprise and the future of work. China redux.
Enough to inspire hours of conversations!

**✉️ ** Would your colleagues enjoy this issue? Forward it!

🤟 Please share *|SHARE:twitter|* *|SHARE:linkedin|*

🙏** Supported by our partner OnePlus.**

DEPT OF THE NEAR FUTURE

💼 Tom Friedman on the future of work: "Always think like an immigrant, because we’re all new immigrants to the age of accelerations."

⏏️ Frank Pasquale on the "shift from territorial to functional sovereignty is creating a new digital political economy [...] only political organization can stop its functional sovereignties from further undermining the territorial governance at the heart of democracy." (Fascinating insights with Amazon as a case study.)
See also, giant companies look invulnerable until they don’t.

📄 New metrics for the algorithmic enterprise. What is the correct way to measure digital firms if, by 2020, non-tangible asset measures should account for 75% of company value? THOUGHT PROVOKING

🇪🇪 Estonia, the digital republic. Its government is virtual, borderless, blockchained, and secure.

🏁 China and AI: China is poised to dominate multiple sectors of AI (including consumer applications, autonomous vehicles and perception/vision) while lagging the US on business applications, argue Kai-Fu Lee and Ian Bremmer. Chinese firms are also moving from copycats to leapfrogging American competitors, to support the State's vision of $1.5trn in AI-related revenues by 2030. MUST READ

🛢️ Decarbonisation: the World Bank will stop supporting oil and gas projects from 2019. French insurer, AXA, will also bow out from insuring oil sands projects. More than a third of cars in Norway are EVs. And Saudi Aramco sees ridesharing currently as a bigger threat to oil demand than EVs.

DEPT OF AI & HUMANS

Augmenting human intelligence: cognitive transformation is "about changing the operations and representations we use to think. GREAT ESSAY (takes 20 minutes).

Large firms are starting to apply artificial intelligence to workforce management and optimise work allocation and performance management. One citation:

"What managers do mostly is identify potential, build teams, assign tasks, measure performance and provide feedback. Generally speaking, humans aren’t very good at these tasks,” said Tomas Chamorro-Premuzic, a professor of business psychology at University College London.

An alternate on the algorithmic management of workers is in this interview with sociologist, Antonio Casilli:

The problem is not that robots are stealing our work, but that we continue to work more and more, and that the platforms are fragmenting and rendering invisible the labor that is necessary to make the algorithms work.

Voleon, an early hedge fund to deploy machine learning, enjoyed some strong early results (as much as 46.3% annual gain), but then found it a bumpy ride

It was three times as hard, and it took three times as long, as anticipated.
The idea that we could just take the machine-learning techniques in speech recognition and computer vision to generate better forecasts just didn’t work.

See also: DNNs fed with decades of fundamental corporate information might be the best new way to automate stock picking.

Elsewhere:

AI: What is working, what is not? Frank Chen’s 30-minute presentation worth watching.
How BBC News thinks about and implements fair and unbiased AI systems.
90% of people think AI will take half of all jobs. 91% of people don’t think theirs will be affected.
Two myths of AI and job replacement: productivity growth is still sluggish & job tasks will change, rather than jobs will disappear, argues Barry Eichengreen.

DEPT OF CHINA

China is on a roll after the Party Congress with its “Made in China 2025” plans focussing on making itself dominant in tech hardware (like communications, EVs and aerospace), as well as a significant programme of financial reforms that should have Chinese bond, banking and equity markets exceeding the size of the US by the same date.

Chinese firms like Yitu Technology are leading the world in combining machine vision with CCTV in real applications. Yitu’s systems “can very easily recognise you among at least 2 billion people in a matter of seconds” and contributed crime fighting.

[P]ickpocketing on Xiamen’s city buses has fallen by 30 per cent; 500 criminal cases have been resolved by AI in Suzhou since June 2015; and police arrested nine suspects identified by algorithms during the 2016 G20 summit in Hangzhou.

Interesting BBC video of public safety CCTV cameras in action.

Machine vision has more to do than monitoring citizens, of course. So it's interesting that Baidu’s former head of AI, Andrew Ng, has launched Landing.Ai to use machine vision to improve manufacturing. First client, Foxconn. (Andrew’s video has an interesting tidbit - that the machine vision system learnt from only five instances rather than the millions of images deep learning systems normally need.)

On China’s Alipay Zhima social credit score:

_Because I lacked a transaction history, I immediately faced what felt like an embarrassing judgment: a score of 550….A scan of a [rental] bike’s QR code revealed a four-digit number that unlocked the back wheel, and a ride across town cost roughly 15 cents. Because of my middling [social] score, however, I had to pay a $30 deposit before I could scan my first bike. Nor could I get deposit-free hotel stays or GoPro rentals, or free umbrella rentals. I belonged to the digital underclass. [...]
For those with good behavior, Zhima Credit offers perks through cooperation agreements that Ant Financial has signed with hundreds of companies [...] a car rental company, allows people with credit scores over 650 to rent a car without a deposit [...] For a while people with scores over 750 could even skip the security check line at Beijing Capital Airport._

Elsewhere:

EV reader, Vivek Wadhwa, argues that China, certainly, doesn’t need Silicon Valley. Its top firms are amongst the most valuable in the world, and are now innovating at greater rate than US firms. One example, WeChat which is “so far ahead in innovation that Facebook may be copying its features.”

Chinese online retailer to open hundreds of “unmanned” shops.
Didi, Uber’s Chinese rival, plans to launch in Mexico

SHORT MORSELS TO APPEAR SMART AT DINNER PARTIES

🌈 Grindr may be killing off gay and lesbian bars in London.

🚑 Uber reduces amublance use by about 7%.

Should we dismiss IQ tests or not?

Full Grown is a British company growing trees into chairs.

🤯 Gut bacteria conspires with kidneys to regulate blood pressure.

🎁 Three-quarters of Christmas jumpers are never worn a second year.

The first super-rich arose from the power of draft animals.

Bot re-writes a chapter of Harry Potter.

🤒 Men experience worse flu symptoms than women.

Japanese retail investors account for 30-50% in bitcoin trading.

If more nerve cells mean more smarts, then dogs beat cats. WOOF! 🐕

Researchers document bird species evolving through hybridization.

WHAT YOU ARE UP TO

EV supporter John Havens and his team at IEEE have published the second version of their Ethically Aligned Design: A Vision for Prioritizing Human Well-Being with Autonomous and Intelligent Systems. It's open for public commentary and suggestions until March 2018. I know John would value input from EV readers, so please do contribute.

END NOTE

Hey!
There was a flurry of snow in London this week and I was able to build an amazing snow-lady with my kids. It reminded me that we're nearly at the end of 2017.
The newsletter schedule for the rest of the year is reasonably straight forward. We'll just publish as normal :)
That is correct. No beats will be missed!

Cheers!
A

🎅🏾 P.S. And if you think EV has been nice this year, please Tweet a recommendation. And if EV has been naughty, please share on LinkedIn!

This week's issue is brought to you with the support of our partner, OnePlus.

"The OnePlus smartphone is spectacular on every level." - Mashable

Exponential View by Azeem Azhar

🔮 Augmenting human intelligence; two views…

Comments on this post are for paid subscribers