😀 Apple & AI; how to stay human; hacking the DAO; decisions under poverty; Uber, grapes & teenagers++ #67

Apple’s move in artificial intelligence. Rushkoff on staying human in the machine age. Improved norms mean safer societies. Why the poor make bad decisions. The attack on the DAO. More climate records shatter. Google’s dominance in advertising. Robots and grapes.

It’s Father’s Day! Hope this stimulates some great conversation this week with your dads ;)

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

Father’s Day Special: Please help Exponential View to reach more interested readers by forwarding it to colleagues or at the minimum  like us on Facebook & recommend on Twitter.

🌟 What can we do to better align today’s tools with the world we’d like to create. MUST READ interview with Douglas Rushkoff

💡Why do young Americans have a negative view of capitalism? GOOD READ

👍🏽 The ‘evolution of social and cultural norms [is] making society more tolerant and safer’ but bad actors will always find a way, says Marian Tupy. 

😮 Why do the poor make such poor decisions? “Poverty is not a lack of character, it is a lack of cash [and the effects of poverty] correspond to [a loss of] between 13 and 14 IQ points.” THOUGHT PROVOKING

💰 Matt Levine of Bloomberg: On the hack of the DAO which resulted in the theft of $53m. The DAO is a blockchain-based decentralised organisation, something that might have been an experiment in rethinking how organisations can arise (and how their behaviour can be directed). Decentralised organisations might be a complementary structure to top-down institutions like firms or social-contract collaborations like societies. ACCESSIBLE READ. (More technical is this one.)

Does Apple understand artificial intelligence? Well … it depends… there were announcements at Apple’s developer event this week which suggest Apple is going to continue its investments in improving UX through technologies in the AI stack.

Three things stood out.

1. Apple announced support for basic neural networks in iOS. In the short-term this means that Apple can offer accelerated inferencing (predicting) on iOS devices, including accessing the phone GPU,  saving a round-trip to the cloud; or allowing certain models to run when there isn’t internet access. Both of these things will improve user experience. It’s unlikely at this stage to meaningfully be useful for local training. And there will be limitations, due to the memory and GPU footprint on the device, to the kind of inferencing that can be done - don’t expect a full automated speech recognition library to fit on a phone just yet. Of course, Google is pursuing a similar route to move TensorFlow onto Android devices and that team has at least one Apple alumnus who previously woked on image processing applications using early GPUs. 

2. The firm continued to use privacy as a unique selling point. It introduced ‘differential privacy’, a cryptographically secure method for training over user data without compromising individual privacy. It “looks like Apple is honestly trying to do something to improve user privacy” says Matthew Green. GOOD READ. This is an important move - and let’s see whether consumers can understand it enough for it to be a product differentiator vs Google, Facebook or Amazon, or pressure higher standards of personal privacy elsewhere on the Internet. (Tom Simonite on Apple & differential privacy is also accessible.)

3. Siri is open to third-party developers, allowing them to plug into the voice interface. This is similar to the approach taken by Viv and Amazon’s Alexa. It seems like this will be the way we will get seamless voice interfaces that give us access to a broad, general range of underlying resources.

None of this really encouraged the meme that “Apple doesn’t get AI” to die. But  Apple is a firm that cares about the UX almost to distraction. And the realms of artificial general intelligence and artificial super intelligence are not at the point where they do improve the UX, so it’s unsurprising that Apple isn’t announcing (or even investing in) systems that beat Go.

🔮 If you haven’t watched it already, look at the 1987 Apple vision of the future the Knowledge Navigator. INCREDIBLY FARSIGHTED.  Apple also had a much-storied R&D group called the Advanced Technology Group, which created Hypercard and Quicktime. It was also responsible for breakthroughs in natural language interfaces including speech recognition & synthesises, V-Twin (a search index)  and the handwriting recognition in the Newton. Jobs shuttered ATG when he returned as CEO in 1997 because he had the stabilise the firm.

More practically, Apple has been buying companies that deliver the AI stack for the past few years. These include Polar Rose (machine vision), Chomp (search), Novauris (speech recognition), Cue (virtual assistant), Topsy (stream processing), Acunu (big data analytics), FoundationDB (databases), Metaio (augmented reality), Perceptio (machine vision) & Emotient (machine vision).

Expect more AI in Apple’s products. But I would be surprised to see large-scale open source efforts, of the kind we have seen from Google or Facebook. Open source has rarely been Apple’s bag.

Elsewhere: Asymco on the state of Apple’s Ecosystem: $40bn per annum across the platform, growing at 20% y-on-y. Healthy.

I thought the Apple announcements mattered enough, given its reach to spend a bit of time on it. But here is some non-Apple AI stuff.

Very good, clear overview of machine learning by Oren Etizoni. ACCESSIBLE

Exceptionally clear, brief extract by Baidu’s chief AI lead, Andrew Ng: “data is like fuel for a rocket, and deep neural network models are the rocket engine. Because we have more data than ever, we are able to build bigger rockets, with bigger engines that can take us to new places.”

The FBI doesn’t correctly test the accuracy of its face recognition technology. More evidence of the kind of real downside risks we’ll run with the proliferation of these technologies, poor design meeting systemic abuse.

An 11-layer deep convolution network turns hand-drawn sketches into photorealistic images. Rather impressive.

🔬 Most impressive: reconstructing images from thoughts. Here researchers analysed fMRI scans of key brain regions (angular gyrus and occipitotemporal cortex) and recreated the faces being thought of. The original paper describes the technique. Will be interesting to see how far this goes with a larger set of underlying eigenfaces & more sophisticated regression techniques. 

😀 Renewables on track to be the cheapest way to generate electricity in many countries by 2020 and in most by 2030, says Bloomberg New Energy Finance.

🔥🔥 Seven climate records have fallen in 2016. Extremely grim reading.

Flying for their lives. Photo-essay on the plight of Australasian shorebirds. (Resist the temptation to ignore this story!)

Volkswagen aims for redemption through electric vehicles. 

The scale of Google and Facebook’s domination of the ad & data pie is staggering.

Is Google manipulating autocomplete in favour of Hillary Clinton? It is complicated, says Robyn Caplan.

Uber is now profitable in all its developed markets.

Why teenagers like likes.

Watch a robot stitch a grape back together.

A robot escaped a lab and made a dash for freedom. Was it a publicity stunt?

☎️ Five things you can do in India with a missed call. Genius.

Interest rates are at a 5,000 year low. (And no end in sight for Japanese bonds, even 40-year bonds approach a 0% yield)

Any British reader over 40 will remember ‘The Young Ones’. Turns out there was a creepy fifth housemate lurking in many scenes.

Is the Standard Model going down? Hints that we have found an unexpected new particle.

I know there is a referendum in the UK this week. I am giving you a break from it all. But please vote on Thursday, if you can.

Have a great Sunday!