#167 | Innateness in AI; the automation advantage | FWD to a friend. 

Exponential View

Azeem Azhar's Wondermissive: Future, Tech & Society
Dept of the near future

🏆  Firms which are early adopters of AI achieve an insurmountable advantage suggests an excellent new report on automation and its impact on work from the McKinsey Global Institute. (Full PDF, which I’ve not yet read, here.) Read also: Irving Wladawsky-Berger on what human-machine work teams of the future might look like.

💯 Superb podcast on network effects, complexity theory and the nature of technology with W. Brian Arthur, the economist who identified the importance of network effects. THOUGHT-PROVOKING

🔥 How has AI affected the demand for compute? The amount of compute used to train the most advanced neural networks has increased 300,000-fold since 2012 (much faster than a Moore’s Law style of relationship). Analyst James Wang reckons this is mostly due to firms being willing to spend more money (and thus more chips) to train neural nets and chip manufacturers, mostly Nvidia, improving chip architectures.

🧠 How much innate knowledge or embodiment do we need to develop AI systems? And what are the limits of current data & machine learning approaches? MUST READ overview of the debate.

🇮🇳  Mobile data amongst India’s 700m phone subscribers is exploding.
Monthly data usage per subscriber has hit 1.7Gb, about 20 times higher than three years ago, on a par with users in the UK or France. It’s been driven by the success of operator, Jio, which aggressively sashayed into the market with rock-bottom prices for data. Indian data prices are about 3% of what they were four years ago. India is now the second biggest market for app stores.

🤑  China’s mobile payments market is heading for the moon. $2.9 trillion dollars was sent across the AliPay and WeChat networks last year, equivalent to about half of all Chinese consumer spending. By comparison, US mobile payments were estimated at $100 billion or about one-thirtieth lower in the same year. AliPay has used this momentum to build the world’s largest money market fund. (Do slow, fragmented US banks stand much of a chance? Even Starbucks seems to be doing better. It’s now the largest mobile payment service by users in the US.)

Viability #2 🚀

A huge thank you to those of you who recommended Exponential View to your peers last week. (Great examples include this from Jim Flint & one of our readers at a large industrial conglomerate who recommended us to several dozen peers on an internal mailing list.)

Your recommendations will go some way to make Exponential View viable over the long term. As a reminder, the community really needs to triple in size for sustainability.

The results last week were meaningful, we added about 500 more readers than usual (or about 1% to the target).

Our models and analytics suggest about 2% of our readers supported us via LinkedIn, Twitter or email. Thank you!  We still need your support.  Please take a moment, especially if you didn’t get a chance to help last week.   

The best way you can do this is to forward this email to seven or eight friends/peers/contacts with a personal recommendation they subscribe and point them to the link

The second best way is to write a short review on a blogging platform like Medium or LinkedIn and share it.

Dept of corporate innovation

I spent a fascinating couple of days at EV reader James Mawson’s Global Corporate Venture Capital conference in London this week. Venture capital of all hues is booming. Last year, $163bn was invested into founders globally, compared to a $25-56bn range in the 10-year period to 2013. Corporate venture capital (CVC), whereby large firms invest in interesting startups, is on an upswing. CVC, traditionally a second-tier option for enterpreneurs, now represents about 18% of all venture deals globally and about a third of all dollars invested in venture. CVC of 2017 is bigger than the entire VC industry of 2013.

How do you make sense of it?

Corporate capital is now at substantial scale and could fit the model outlined by Venezuelan economist, Carlota Perez. As many readers know, I’m a fan of Carlota Perez’s thesis in “Technological Revolutions & Financial Capital” which I believe can shed some light on what is going on.

In Carlota’s thesis, innovation goes through a number of stages. A first stage, an installation stage, involves novel technologies backed by financial technologies. The middle stage takes the form of a bubble, collapse and crisis. The third stage, the deployment stage, sees the widespread distribution of the benefits of a now mature set of technologies, funded increasingly by what she terms “production capital”. This is different to financial capital because production capital’s purpose not solely to secure speculative gains.

What we seem to be seeing today is a transition to the third stage of a Perez cycle, the deployment stage, of the information and communication revolution which started with the Intel 4004 in 1971. And we do seem to be seeing a resurgence of the corporate capital in investing. Many of the target investments are classically in “deployment” opportunities, that is products and services higher up the stack rather than in lower-level enabling technologies or infrastructure.

Take the Softbank Vision fund which is quite a category-defying fund. But while much of the capital has come from non-corporate investors, it is in some sense tied to the vision of Masayoshi Son, Softbank’s CEO. Softbank is a corporate, a conglomerate that operates a mobile operator. It is not a traditional CVC, to be sure, but it isn't exactly a stand alone financial VC either.

Equally, a firm like South African internet conglomerate, Naspers, has started to cash in on its holding in Tencent, recently harvesting $12bn of those gains, to put towards a war chest for investments.

Intel, the grand-daddy of the corporate venture, has invested more than $15bn in around 1,500 companies in the past couple of decades. (By scale, that would make it one of the most generous deployers of venture capital ever. The largest traditional VC house, NEA, has raised $8.2bn in the past decade.)

These are big numbers and meaningful in the relatively parsimonious world of venture capital, which makes up less than 1% of global assets under management (which were estimated at $84.9 trillion in 2016.)

But is the current penchant for corporate venture capital just the usual part of the cycle? That cycle of corporates getting excited by waves of innovation: the internet, then social; and jumping in to venture, a field they don’t know well, just as the market reaches new highs; only to be burnt in a downturn and shut down their programmes? We’ve seen that pattern many times before. But I’ll suggests three reasons why this time it’s different.

  1. The Perez argument is highly relevant. No company can deny that every business model and way of operating seems up in the air. Detroit’s grizzled veterans say the car industry is dead. Traditional retailers across the world are being eviscerated and urged to do something. Newspapers in shock from the violent 1-2 from Google and Facebook. Pharma pipelines are running into dead ends, even for promising areas like Alzheimer’s treatments. And so on...
  2. Corporate research and development is getting progressively more expensive and yielding less and less, except in a handful of superstar firms. And these firms (Google, Microsoft, Amazon, etc) are aggressive in their use of corporate venturing & acquisitions.
  3. The nature of investments in the information age doesn’t suit the internal capabilities, financial controls, strategic planning and risk perception of ‘old economy firms’. The reason is the intangible nature of value created in information businesses (such as marketplaces or software), as described by EV reader, Stian Westlake, is his tasty economics book, Capitalism without Capital. These internet unicorns like Deliveroo or Lyft require capital to build real businesses and aren’t worth much until they do. And if an incumbent can’t build them organically, it'll need to buy them. I have argued that CVC provides a useful tool for large firms to make smarter acquisition decisions.

In the noughties, corporate venture capital was often been seen as an investor of last resort. Knackered (or naïve) entrepreneurs might turn to CVC if the long stroll up and down Sandhill Road hadn’t yielded a term sheet.

CVC funds are getting more professional at doing what they do, and their corporate parents are being less burdensome and annoying towards founders. Not perfect, by any means, but getting there. And CVC is slowly learning to play alongside dedicated VC funds which maintain their expertise in helping founders to build companies, access the right talent and attenuate the existential stress of being a founder.

It’s likely this trend will continue and an increasing number of founders will find a corporate venture capital investor on their shareholder roster.

Dept of GDPR effects

Like many of you, I'm delighted to be off the marketing lists of dozens of firms, all courtesy the GDPR directive, a new EU regulation. This does mark an important turning point. The EU is demonstrating it is the most powerful force in trying to figure out what regulation in an information economy needs to look like. (See a reasonable argument on this point in the Washington Post.)
But naturally there are some second-order effects emerging.

  • Microsoft announced it is expanding GDPR protection to all customers globally (something which Facebook, for example, has not done.) Microsoft is demonstrating increasing ethical leadership under the stewardship of its President, Brad Smith. It is one firm to watch more broadly in the emerging domain of techno-politics.
  • EV reader, John Battelle, argues the legislation threatens the position of sub-scale players, including innovation publishers and app makers. It is true that the very largest firms with the most modern platforms (Facebook and Google) were most well equipped from a technology and organisational perspective to handle the GDPR change. Their business models were also very resistant, because they deal directly with consumer data and don't have much reliance on third-party data. 
  • The Washington Post has launched a premium, ad-free service for European users. This is an interesting test-bed to see whether people are willing to pay to avoid invasive tracking. 

Short morsels to appear smart at dinner parties

Bad blood at Theranos. 😱

Review of David Runciman’s How Democracy Ends. Tl;dr: coups, catastrophes or technological takeover.

 📞 Great profile of Patterson and Hennessy who devised the first RISC architectures back in the 1980s and were recently awarded the Turing Prize. (RISC architectures power pretty much every mobile phone and computer today.)

The number of female chief executives is falling.

A new synthetic biology project to make virus-resistant super cells. 

⛵ The future of transport in waterway-rich cities may be autonomous boats. 

China's social credit system has blocked people from taking 11 million flights and 4 million train trips.

What Arctic ice can tell us about the Roman history.

☀️ Belgium is using Tesla Powerpacks for frequency response to provide a more reliable electrical grid. See also: good analysis on the impact of growing wind and solar energy on overall energy costs.

🎼 Visualisation showing how Billboard hits increasingly sound alike. (For better, more distinct music, try the Exponential View playlist compiled by our readers Marko Ahtisaari and Nadya Peek.)
Azeem's end note 

It's a lovely summer weekend in the southern part of the UK. I'm lucky enough to have a few days in the countryside with some friends. I hope you also have a wonderful weekend, whether you are enjoying a balmy summer, the heat of the tropics or a cooler Southern hemisphere autumn!

Thanks again for reading, and thanks again for spreading the word.



P.S. Here is a tweet you can use to spread the love. 💕

This week's Exponential View has been supported by our partner: 
Copyright © 2018 ExponentialView, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list
Having trouble seeing this email in your inbox? View it in your browser.
For commercial enquiries, please email Marija Gavrilov.