🔮 Machines vs humans; women leaders; the tech cycle; Facebook's bubble; North Korean startups, chaste men, illusions++ #71

The Exponential View

Has the tech cycle come to an end? Is machine learning isn’t magic? Do machines still need humans? Are we entering an age of women leaders? Is the media filter bubble is getting worse? How do we manage AI monopolies? And much more to keep your Sunday thought-provoking.

PLS DO THIS Recommend us on Twitter, especially if you haven’t done so in the last few weeks. Thank you!


Dept of the near future

🔮 Elad Gil on the end of the tech cycle. “A software-enabled, network-connected, crowd-funded, smart toaster is when all is said and done, still just a toaster.” THOUGHT PROVOKING

🐰 It isn’t magic, it’s just machine learning. How to reduce AI hype and get to reality. GOOD FRAMEWORK by @smerity (See also this amusing tweet.)

⚙ McKinsey & Co: Where machines can replace humans, and where they can’t. **EXCELLENT **research and framework (albeit a not-so-subtle wrapper to promote McKinsey’s organisation & change services)

👩 An age of women leaders: female presidents, prime ministers and rulers have pretty impressive track records argues Matt Ridley GOOD READ

Dept of future of media (aka Facebook)

Emily Bell is EXCELLENT on the dangers of the online filter bubble where algorithms (rather than editors) determine what we read.

The point here is that our news is now heavily curated through Facebook. And Facebook determines what we view based on an algorithm essentially designed to maximise time on site & retention. That is, Facebook figures out what to show us to have us staying longer and coming back to Facebook. (Some good recent data on the firm’s dominance as a news platform.)

That may not be educational, informative or diverse, even if you deliberately try to add diverse sources. Why? Because your own internal bias will reward the maths one way or another. And that slight bias will be seized on and amplified.

💡 WSJ demonstrates this effect in a brilliant interactive showing what news looks like depending on your particular digital echo chamber.

Live video is visceral and raw. As Facebook promotes it more, how will it deal with the grisly and violent? “BuzzFeed News put all these questions to Facebook. But the company declined to answer them.” (Was it the Police who made the Facebook video of the Castile shooting disappear for a while?)

Part of the issue - Silicon Valley’s twin heritages of libertarianism & hippiedom. Layer on a solution-oriented engineering culture that by and large has not studied nearly enough philosophy, political theory or history and you can understand why these platforms take the stance they take.

Chartbeat analysis: What Brexit tells us about how people read news

Dept of artificial intelligence

☀️ Good summary of the White House’s recent confab on AI MUST READ

The challenge of AI monopolies and public datasets. This week “millions of NHS patients’ medical histories are being traded with DeepMind in exchange for some free services. And none of these people have been asked if they agree with the specific trade.” EXCELLENT read from Natasha Lomas

How Microsoft is betting the farm on machine learning and AI.

Robopocalypse now, Asian edition. More than half of all salaried workers in Cambodia, Indonesia, the Philippines, Thailand, and Vietnam could be displaced by automation and advanced technologies, says the ILO.

The human memory champ who beat Alibaba’s face recognition AI.

The new Prisma app which uses neural nets to apply semantic style transfer is kinda mind-blowing. Recommended iOS download.

Short morsels to appear smart at dinner parties

Brilliant: Apollo 11 guidance computer code published on Github. (Link on Github.)

🇰🇵 North Korea’s tech scene: Pyongyang Startup Week.

Up to 40,000 fMRI-based brain studies may be invalid due to a software glitch.

What are the implications of Dallas police using a robot to apprehend and kill a suspect this week?

For black Americans, the US is about as dangerous as Rwanda

Why we voted leave: Compelling video looking at pro-Brexit voters in North of England by Sheena Sumaria.

😃 Gorgeous visualisation: data mining the six major story arcs.

“The out-dated idea that females are chaste and males are promiscuous needs to be thrown away”.

LGBTQIAPK: what do all those letters mean? (See also how intersex issues might be a feature of the Rio Olympics.)

Meet the artificial stingray propelled by rat muscle. #bioengineering

Size distribution of Earth’s lakes? Power laws, natch.

😲 The illusion of the year.

End note

I am on holiday for much of the next four weeks. EV will continue to publish during that time, as we’ll retrospect some of the most important stories of the past year.

Have a great summer & please continue to recommend us.