How will AI change design? Can Facebook cope with its new role? Are curators the new black? What does VR hold in the next few years? Why is AI forcing a reorganisation of the world’s largest firms? Just how reliable are Teslas? How does your smartphone get its battery? What is the rudest swear word?

Hope this stimulates some great conversations this week.

Please recommend this: Twitter  | Facebook

SUPPORT EV: Donate via PayPal. (You choose the level.)

Referred by a friend? Sign up

_For generalist readers, this week is marginally more technical than usual for which I apologise. Bear with it, there is good stuff in here for you. _

Dept of the near future

💡 Design in the artificial intelligence economy: “we might someday talk about ‘discovering a design’ through the joint efforts of human and computer neurons, rather than 'creating a design.'” THOUGHT PROVOKING

🌟 Monopoly in the algorithmic economy: "the invisible hand that we all rely upon has been pushed aside by … the “digitized hand.” This hand is controlled by corporations and can be manipulated.” EXCELLENT READ ht @kerryritz

🔮 Facebook is being taken somewhere it never wanted to go, argues Emily Bell. “Like the BBC, which also started out with an engineering mission, Facebook cannot see publishing decisions—whoever takes them—made on its platform as separate from its corporate health and reputation. They are the same thing.” **INSIGHTFUL. **

🙋🏽 In the age of the algorithm, the human curator is back. OUROBORIC

🚀 Elon Musk wants to take us to Mars. Tim Urban’s essay on the subject is the one to read. LONG READ

🔥 We have galloped beyond the 400ppm threshold for CO2 permanently and we aren’t stopping there. We need more coordinated collective action. AY CARAMBA

👾 What do the next few years promise for virtual reality, asks Kyle Russell of A16Z. (See also Robert Scoble’s portrait of Apple’s augmented reality future & this breakthrough in high-fidelity haptic gloves.) GOOD READ

Dept of artificial intelligence

Why deep learning is changing your life. GREAT introduction for the general reader into what’s behind the current deep learning boom (and why it’s real).

Microsoft joined the other major firms announcing a significant strategic realignment spurred on by artificial intelligence.

Key amongst them:

• 💥 Fantastic  story from Cade Metz describing Microsoft’s custom AI hardware,  field-programmable gate arrays which can be optimised for machine learning tasks. These are being used not just in Bing but across Azure. Remember, Google, Facebook and Apple are all making their own custom hardware (or even silicon) to accelerate machine learning performance.

• A reorganisation of the company around a 5,000 person AI division. The primary goal appears to be to reduce the latency from new research innovations making their way into products. AI is a special-case technology which doesn’t require a huge amount of end-user retooling for tech companies to deliver value to users. The growth of the Internet required consumers to buy PCs, get ISP subscriptions, acquire modems. And they arrival of mobile as a platform required us to buy smartphones. A consumer AI enhancement can work with your buying new hardware because the AI improvement can be delivered as a lightweight runtime via an app upgrade or managed on the back end in the server farm.

I’ve made a rough estimate of the sort of scale of human investment into AI and machine learning by major internet companies. It is pretty mind-boggling. (Several of these firms also combined to form Partnership for AI, an interesting group to address ethics issues.)

Indeed, the cost of more powerful models is rising. Eliot Turner tweets that in some recent papers the computational cost to train some Google models ran to $13k. This is significant as you may have to run multiple experiments before you get to a model you like and on-going learning will be expensive. Once trained models are relatively cheap to run and execute in their ‘inferencing’ mode.

But it demonstrates the increasing barriers to entry for smaller firms, and perhaps a challenge for financial controllers handling budgeting. I’m not an accountant but one thought is that you should be able to amortise the upfront investment in building a machine learning model. But you could make a case for depreciating it over time (and treating ongoing training as 'maintenance’). I suspect in many smaller firms this isn’t the case and it is all treated as an operating expense. Any accountants want to weigh in, lmk.

Elsewhere:

• Fascinating write-up of a paper where researchers were able to reverse engineer various prediction algorithms. Essentially, you can extract the salient characteristics of a training model by using it and training your own model with the outputs of the one you want to train.

• Nvidia demonstrated a rather cool video of a self-driving car trained by observing human drivers. Impressive demo. The chip maker also announced it’s new Xavier system-on-a-chip for self-driving which will manage 20Tflops while consuming only 20w of power.

• Google released a cloud platform for or'nery folk to do machine learning. This appears to be a massive upgrade for developers (or businesses) wanting to undertake simple machine learning tasks.

• The company also showed off vastly improved machine translation using neural nets. By their own metrics, these are substantially closer to human quality translations.

• Google also open-sourced a labelled image dataset for anyone wanting to explore machine vision applications. (Limitations: only 6,000 image categories so probably not granular enough for specialist applications.)

Short morsels to appear smart at dinner parties

I am a huge fan of Eric Beinhocker’s work in complexity economics. Here is an interview with him celebrating the ten-year anniversary of his brilliant book, _The Origin of Wealth. _

World’s first three-parent baby born.

🎮 Are American men rejecting paid-for work in order to play video games?

🇨🇩 The Cobalt Pipeline: from the heart of the Congo. Meet the barefoot minders who find the raw ingredients for your iPhone’s  battery. MUST READ

A Tesla Model S has completed 200,000 miles of driving with virtually no maintenance and just 6% battery degradation. (Think of what this means for car dealerships and repair yards, not to mention the car upgrade cycle.)

😳 Every British swear word ranked in order of offensiveness. The things our regulators get up to. (Read in the context of Facebook’s acceptable content rules.)

What will English sound like in 100 years? Weird.

🌽 What would happen if the world went vegetarian? Longer lifespans & fewer greenhouse gases, amongst other things.

What you are up to

Profile of EV reader Khemaridh Hy and his spiritual journey.

EV subscriber Obi Felten is promoting a cool game designed to help kids with their mental health.

Creative skills are more important than STEM, argues EV reader John Hagel GOOD READ

End note

Welcome to Autumn!

I updated the testimonials page for Exponential View. Take a look and retweet the one you like.