Greek Complexity

Nuance

Someone on my FB feed shared this post today:

I had two reactions to this. Interestingly, they are contradictory to each other.

My first reaction was a head-slap, face-palm, eye-roll at the lack of nuance in the post. One, none of the incumbents except for Blockbuster is dead. So stop saying they were each "killed" by a tech firm. And two, is lack of customer-centricity the only reason for the troubles faced by the incumbents? Really? And three, is it a foregone conclusion that Uber is a long-term viable business (in its current incarnation)? I'm not so sure.

My second reaction was, "That's just silly. All five arguments as stated here are about price. Not customer-centricity." Which of course replaces the author's single-cause "analysis" with one of my own. Make of that what you will.

What is information?

Here's a question I've been fruitlessly pondering a long time.

Where does information come from and is there even such a thing as "new" information? In other words, when someone writes down a mathematical theorem, have they just created new information or was the theorem always "out there" waiting to be discovered? What about the specific arrangement of molecules that we call the Mona Lisa or the silly limerick I composed yesterday?

Put another way: is information a form of energy and therefore does it obey the laws of thermodynamics, implying that it can neither be created nor destroyed, and can only be changed from one form to another? If it is, then when I composed that limerick, what other information did I change, destroy or use up in composing my limerick? On the other hand, if information isn't a form of energy, does that mean we can create something out of nothing?

There's a related question in maths: is maths invented or discovered? Do numbers have a meaning and existence independent of our own perception of them? I come down on the side of it being discovered but it's a weakly held view. To my knowledge, there isn't a consensus answer to this question.

---

After posing this question to a few people (and not yet having received an answer), I came across John Wheeler's "it from bit" information-theoretic view of physical reality. This might be related to and even answer my question. I'm not sure yet as I'm still digesting what Wheeler was getting at. It's also possible that his characterisation answers a different (opposite) question, namely, where does physical reality come from, and not, where does information come from.

The best book I've read this year

I've read several books so far this year. Spanning science fiction, biology, speculative historical fiction, physics, modern fiction, economics/investments… One book stood out as the best. (The worst – relative to the others, not in an absolute sense – was Breakout Nations by Ruchir Sharma.) That book is Soonish by Kelly and Zach Weinersmith. Kelly is a parasitologist at Rice University and Zach is a cartoonist probably best known for SMBC.

 

Soonish is an exploration, or perhaps an imagination, of technologies that humanity may come up with in the relatively near future. Not so near as predicting the next smartphone, not so far as to be completely speculative (AKA useless). Let's say, in our lifetimes.

 

I'm not going to give away anything in their book or tell you what my favourite chapters were. It's full of great science and really funny jokes. I actually laughed out loud several times. Some real groaners too. All of which make it a wonderful read.

 

Suffice it to say that the book easily lives up to its aim. The Weinersmiths don't try to do too much and stick to ten emerging technologies explained in a good amount of detail, instead of putting out a book-length listicle of two hundred items. They interview a number of scientists and others for their material. They give you the perfect amount of historical context and information about commercial considerations that will guide these technology developments.

 

You can follow along and enjoy the book even if you have no prior scientific knowledge. The evidence for that comes from the chapter on gene-editing. This is literally the first time ever that I've actually understood what a gene is (well, kinda, good enough understanding). Other sources I've read tend to quickly get bogged down into the terminological detail and I lose the woods for the trees. Or they stay high-level (which is ok), only naming the concepts without explaining them – but labelling things is not the same as explaining them.

 

So if you read this for no other reason, just read it to understand what genes are.

 

I also loved the book because… of the authors' surname. I absolutely love it. Not only is it a compound name made up of both of their pre-marriage names – yay, no prioritising the man's name over the woman's – they're also names that originate from two different cultures. In fact, I don't know which is which. Was Zach originally Smith and Kelly originally Weiner? Or was Kelly originally Smith and Zach Weiner? Don't know, don't care, and that's the way it should be. To be clear, they don't make this a political statement in the book. If I recall correctly, they don't even mention it. 

 

---

 

I've always been kinda puzzled by the idea of surnames. 

 

Yes, I know it helps to have a second name in addition to your first to help distinguish Yuri Suzuki from Yuri Uchikata, but surely the information content of that is not REALLY that helpful in the modern age? If you lived in a village of a few hundred people, sure, it would help. But in today's globally interconnected village, there surely are several folks called Yuri Uchikata? (I haven't checked. I somehow like the sense of not knowing. Besides, I'd need to do a web search in Japanese.)

 

I know it also helps with genealogy traces. This seems to me to have greater utility than the unique identifier point above. Though here too I imagine things fall apart fairly quickly as you move up through prior generations and women's surnames drop away or as you move laterally to cousins.

 

It also seems to me to be unnecessarily disruptive for one person (the woman) to suddenly drop one of their names in adulthood in order to adopt a new one. Just thinking about the admin and the record-keeping and the now-confused digital repositories – why go through all that hassle!

 

What puzzles me more is when people make – what seem to me to be quite silly – statements like, "He's a real Donovan." Um? That person has elements of so many different ancestors, who presumably weren't all in-bred, that this statement doesn't make any sense to me. Even more, it might even be that, although the surname is inherited from a male line, that at least one person along the line was not actually conceived by a Donovan. Given the chances of this, and I guess the chances vary based on time and place, if you really want to retain some connection to one's origins via a surname, it makes more sense for women's surnames to be passed down, not men's.

 

In any case, I find the usage of surnames strange full-stop, irrespective of whether they are passed down through the male or female line. This is partly for the information content reason I mentioned above but also because I grew up in India, which has many different naming conventions.

 

Many or possibly most Indians don't traditionally have the concept of a family name. Contrast this with Europe or China which have had family names, surnames or patronymics for between a couple of hundred years (e.g., the Netherlands) to well over 2000 years (China).

 

Surname usage and conventions now differ dramatically across India, with some people still continuing to go without surnames today, some people having adopted certain names as their surname conventions fairly recently, and others now having had surnames for several generations.

 

The ones who have established surnames tend to follow the same sorts of surnaming-forming conventions that have been followed in other parts of the world, such as being named after an occupation or ancestral place of origin.

 

Many official forms now require surnames, so over the last several decades, people have adopted various name conventions.

• For instance, in the south of India, one fairly common name convention used to be Initial1-Initial2-GivenName. Initial1 would refer to one's place of origin, which isn't necessarily where one actually grew up, but some ancestral town or village one might never have been to. Initial2 would be one's father's given name or husband's given name for a married woman. This made sense in a bygone era, where it provided the most context to identify someone as "Greg, son of David, of such-and-such village". Some people still follow this system. 
• Sometimes Initial1 would be dropped as in Srinivasa Ramanujan's name. 
• Others, to avoid confusing people unfamiliar with the convention (including Indians from other parts of the country), have switched things around and adopted the patronymic as one's surname. This can still be confusing – I have sometimes been asked by Indians whether my given name is Murli or Ravi.
• Still other South Indians have resorted to adopting one's clan name, which used to historically be more or less invisible in day-to-day life, as their surname. The virtue of this last approach is that one's descendants can continue to use that same surname, which they can't under the patronymic system, and one's cousins similarly would be easily identifiable as belonging to the same extended family. That last bit may not be obvious at first sight so let me mention that I even have a couple of cousins who are brother and sister, yet have two different surnames. For whatever reason, one's surname is patronymic, and the other's is based on her caste identification!

And all that is just about (some) South Indian names. Other parts of the world, such as Russia, Iceland and Spain, have other conventions. If it's so confusing and arbitrary in just one segment of the population belonging to one region of one country – well.

Zombie ideas

I said in my previous post about knowledge that path dependence can result in certain outcomes that would not have occurred if the path initially chosen had been just slightly different. I want to look at another side of this: zombie ideas.

This post follows up on my previous one and is more question than opinion, let alone an answer to the question.

Martin Wolf wrote a piece in the FT a few days ago, where he listed a number of bad ideas and untruths about Brexit that simply refuse to die. (Ergo, zombie ideas.)

I tweeted a link to his piece thusly:

Zombie ideas about Brexit that refuse to die https://t.co/odhPZaer0S #economics #negotiation #policy #idiocy

— Murli Ravi (@murli184) October 26, 2017

See that last hashtag? I put it down a bit facetiously as a word that rhymed with "policy". But I've actually been seriously thinking about this for a fair bit and thought today that it was worth a proper post.

---

We all know that drinking orange juice or ingesting vitamin C in other forms is good for preventing and curing the common cold. We do know this, right?

Actually, no.

This was a theory that was brought to mass attention in the 1970s by Linus Pauling, a Nobel laureate in chemistry and a superstar of the field of molecular biology. Pauling also believed in the use of vitamin C in cancer therapy. He continues to be regarded as one of the most important scientists of all time for work done earlier in his life. However, his claims about vitamin C have been tested by other scientists since the early 1980s and been found wanting. My understanding is that the medical community at large does not support his claims about vitamin C.

Nevertheless, this idea refuses to die. Post on Facebook that you're coming down with a bad cold and you will invariably receive a few well-meaning exhortations to "Take vitamin C!"

---

So my question is: why? There is some connection to the idea of path dependence, of course, but that doesn't explain it fully.

Daniel Levitin mentions in this video in an unrelated context that:

[…] by some estimates the amount of scientific information in the world in the last 20 years equals all of the scientific information generated up to that point in history [around the 21:52 mark]

Forget about whether the 20-year estimate is accurate. It seems at least directionally true. The rate of generation of information has sped up dramatically (and there's no reason to suppose that it does not continue to speed up even further). If this is so, why continue to rely on old claims that have subsequently been shown to be incorrect?

Why do things that are known to be false continue to live on as "truths" in people's minds?

• Tradition?
• Comfort in the familiar?
• Some other kind of status quo bias?
• Reliance on one's favourite authority?
• Lazy thinking and inappropriate reliance on heuristics?
• Insufficient dissemination of newer or correct information?
• Confusion when faced with claims that contradict each other?
• Siding with the tribe that one identifies with?
• Identification of oneself with a certain pet theory and therefore the unwillingness to call that theory into question because of the harm to one's own ego?
• Lack of critical thinking training being provided in schools?

It is also true that many other claims that have been shown to be untrue have fallen out of favour with the world at large. Such as, say, communism, the idea that an ether permeates the entire universe, that women must not ride bicycles because it would give them "bicycle face", and on and on.

What separates the ideas that do die from zombie ideas?

I'd love to know.

Path dependence

I've been thinking for some years about how knowledge gets created and used. Where does knowledge come from? Why do things that have been known by some people for a long time sometimes remain unknown to the world at large? What could be done to make the process of knowledge creation more efficient?

One sub-theme in particular that keeps cropping up in my head is that of path dependence. I remember as a teen reading a novel by either Arthur C Clarke or Isaac Asimov where the author made a throwaway comment about how such-and-such imaginary civilisation never discovered fossil fuels and went directly from pre-Industrial Age technologies (as humans would think of them) to nuclear energy. That then led to consequences of its own, in terms of how much further that civilisation developed and in how much shorter a time period (both because of the compounding effects of powerful technologies and because of not needing to undo prior technologies and their associated consequences).

That's just an example of the theme.

Another completely different example could be in the area of politics and law. For instance, today's culture wars in America around gun ownership. If a couple of sentences in the US constitution had been written in a slightly different way, we would be seeing a very different environment today.

It's about how ideas that seem desirable, uncontroversial or even negligible in the current time sometimes end up becoming major influences on the future when examined in hindsight.

This may sound nebulous and irrelevant to the nuts-and-bolts of daily life so let me cite just two examples from the recent past which may illustrate the point better.

Singapore's train system has a reputation for being efficient and reliable. That reputation has lost some of its shine in recent years as this and other pieces have stated. What's interesting to me is the company's CEO's comments that

part of the operator’s troubles since 2011 had to do with “some deep-seated cultural issues”.

Let's take the comment at face value. This comment means that the culture that gave the company its good reputation in the past is the same culture that is now responsible for the company's current turmoils. Keep that thought in memory.

A better known example is that of Facebook, which has come under fire since last November and increasingly so in the last several weeks for its role in last year's US presidential elections. This piece brings up the path-dependent consequences that were invisible at the time certain policies and decisions were put in place:

What the public sees as Facebook’s failure to recognize the extent to which it could be manipulated for untoward ends, employees view as a flawed hindsight justification for circumstances that mostly fell well beyond their control. […] Today, the engineer’s anecdote reads as a missed opportunity — a warning of an impending storm of misinformation blithely dismissed. But inside Facebook in July 2015, it seemed a rational response.

You'll have to read the source article for further context. Notice that the conclusion I drew earlier is the same one popping up here, i.e., what worked in the past doesn't only not work any more, it actively impedes the company today. 

The implications are relevant not just for businesses but also for areas as wide-ranging as parenting, energy policy, fashion, medical research, how felons are treated by the authorities, and many others. Pretty much any endeavour where early decisions set the direction of future developments and where effects of decisions compound over time. The compounding is key.

What is the implication for businesses in particular? Let's say you want to optimise for longevity and relevance. How would you do this? Is there even one magic formula? Here's what I suggest:

• Set the parameters of your culture (otherwise known as your values) early in the game. Make sure you as founders, managers and employees – but especially founders – scale these values across the company as it grows. I recognise that this is not a particularly controversial or even original idea. (How to do this has been covered by many thinkers I respect, including Patty McCord, Andy Grove, Brian Chesky and Lars Dalgaard. These are just a few names off the top of my head as I quickly bang my thoughts out, aren't in any particular order and certainly shouldn't be considered a comprehensive list.)
• Although that first suggestion isn't particularly novel, what might be somewhat new is the explicit recognition of the effects of path dependence. This is especially especially true for fast-growing startups. As your company grows from two founders and a pet octopus to 50 people, and again to 150, and then to 1,000 and beyond, you will need to re-evaluate whether the values and culture you laid out at your inception are still helping you with your eventual mission. And if they aren't, you may need to start from first principles as though you are starting a new company that happens to have 1,000 staff at its inception. 

That second idea is a bit radical. "You're saying we should change our core values as we grow? Really?"

Hang on.

Everyone recognises that processes change with growth. The casual expense reimbursement process you followed when you were a 10-person firm no longer works when you are at 50. The algorithm (or gut feel!) on which you based your marketing expenditure when you were a 50-person firm certainly doesn't work when you are a serious 500-person company. It seems obvious that processes need to change. But do we need to re-evaluate values too?

If things around you are changing radically, yes, you do. That can mean a fast-growing startup as I said earlier. It can also mean a stable company that is having to cope with large changes impinging upon it from the outside, whether those be changes in customer behaviour, regulations, competition, technology, or anything else.

It can even mean simply that the character of the company has fundamentally changed and therefore its values must change too.

Facebook used to be known to tell its staff to "move fast and break things". That motto has long since been dropped. As a giant company that fairly quickly became an infrastructure provider and left behind its origins as a disruptive startup, it could not live by its old motto any more.

In fact, here is a counter-factual which makes the notion of sticking to one's founding values even seem slightly ridiculous: how can anyone prove that the folks who came up with those founding values were omniscients who knew exactly how the future would play out and designed those founding values to deal with all possible eventualities? When some Americans venerate their Founding Fathers (or some Indians venerate their ancient texts, or many other such examples), I always think to myself, "But even the people who came up with those ideas didn't think of themselves as omniscients. So why do we?"

Once you explicitly recognise that the choices you made early in the game and the path that those choices led you down to where you are today may not be relevant to your future (because they set the direction of future developments) and may even be harmful (because of compounding), then you have no choice but to start again.

---

I wrote a follow-up to this post here.

Headphones

Is anyone else as worried about headphones, headsets, earpods, callemwhatyouwill as I am?

I use these things fairly often. More than I should, probably.

But not as much as a teenaged kid who lives in my apartment building. He has them on all the time. And of course, he's not the only one. When I'm out walking the dog, a large proportion of the folks we pass jump out of the way at the last minute because they're plugged in and can't hear us coming. On the train and in the lift, there's always at least one person whose headset leaks so that I can hear the tinny sounds from a metre or two away.

Which I suppose is fine if that's only occasional usage and if the volume isn't terrifically high.

But I suspect, based on my own usage of not headphones but a handheld Bluetooth speaker I carry and listen to podcasts on, that we set our gadgets at a fairly high volume so that we can drown out ambient sounds and hear everything there is to hear in the audio we're listening to.

That's not great for young ears or old.

My father has always maintained that my grandfather – and some years later, my grandmother – went deaf in his later years because his apartment was right next to train tracks. Stick your head out the window and you have a train right there. My father says that the constant clickety-clack all day made my grandfather go deaf.

And that's clickety-clack outside your head. What happens when you move the sound to virtually inside your head with a set of earpods?

Like everyone, I have my own morbid thoughts sometimes. "What if I were to suddenly go blind tomorrow? What would that be like? How about if I lost the use of an arm? What if I lost all my teeth? What if I went deaf?" The idea of deafness is the one that scares me the most.

Not that the others aren't scary. But I've always enjoyed listening to things. Whether it be a bit of music that I rewind and listen to more than once so that I catch every instrument. Or that (real or imagined) brief hesitation in someone's speech that I imbue with meaning. Or the sound of insects and blades of grass rustling against each other as I sit quietly in a grassy field and just keep my ears open.

Headsets are just great. I love them. But sometimes I put them away and listen to the breeze.

Fast fashion and "fast fashion"

Yesterday, I posted to my social media accounts a link to this piece, which explains the troubles American casual fashion brands are going through. A friend responded, confused and distressed. I thought I would post her question and my response as I thought it would help other readers too.

Here we go:

 

Things that make me puke

Yesterday I read, The World’s Longest Recruiting Process: Josh McFarland is Joining Greylock as a Partner. My pukoradar began to boot up at the sight of that headline but I ignored it. Then I came to:

TellApart applied cutting edge data analytics and predictive modeling to the field of marketing. In the time it takes you to blink — literally 100 milliseconds — TellApart’s technology can respond to an impression in an ad exchange, crunch thousands of model features and serve the right ad at the right place and the right time. And it does this hundreds of thousands of times per second, with great insight and accuracy.

Really? This is the best you can do? I'm by no means the first person to point out the inanity and insanity of absurdly smart people reduced to doing this. I also freely admit that I used to be one of the flock myself, being fascinated by the potential for algorithmic ad placement, trading, optimisation, and so on (and I still am fascinated in some ways). The parallels with, say, high-frequency trading are obvious and have already been seized upon.

Then today I read, Methbot: a $3M-$5M/day video ad-tech fraud:

White Ops, a security firm, has published a detailed report on a crime-ring they call "Methbot" that generated $3M-$5M by creating 6,000 fake websites to embed videos in, then generating convincing bots that that appeared to watch 300,000,000 videos/day -- running virtual instances of various browsers (mostly Chrome) on virtual machines running MacOS X, from a huge pool of IP addresses that they fraudulently had assigned to US locations, deploying clever grace-notes like limiting access to "daylight" hours in their notional locations; simulating mouse-movements and clicks and more.

It's bad enough that prodigious brains and huge amounts of capital are being deployed towards optimising digital ads. But what stories like this second one tell me is that even the fundamental premise of digital ads is invalid (let alone trying to optimise it).

Page views? That's so 20th century.

Clicks? Meh.

Click farms? Old hat.

Automated click farms? Yawn.

Measurement "errors" and over-billing? Surely. Not.

And so we arrive at purely simulated, never-even-existed views and clicks. Ah, yes, let's do that.

A race to the bottom.

Surely we can come up with something better than this? A system that actually creates real value? We're missing the woods for the trees.

---

I'm sure Josh is a perfectly wonderful human being – and the Greylock blog post goes on to tell you so in nauseating detail – but that's not really what set my pukoradar off.

Why RedMart's sale to Lazada/Alibaba matters

I'm cheering Lazada's acquisition of RedMart – but probably not for the reason you think.

A couple of years ago, I wrote about the need for more soft landings. Why? To enable more risk-taking. Set up as many incubators as you want but that isn't going to incentivise new venture creation as much as a culture that, if even if it doesn't quite applaud risk-taking, at least provides the chance of a safety net for those who don't quite make it to the big time.

The South-East Asian entrepreneurship scene has grown leaps and bounds over the last few years. The number of entrepreneurs and people aspiring to join start-ups has definitely hit critical mass. Whereas I might have been at most two degrees of separation away from every single person in tech start-up circles a few years ago, I'm not sure I can make that claim now. There is also a vastly greater amount of sectoral and functional depth in the market now.

The one thing we don't yet have enough of is soft landings. And that's where the Lazada-RedMart news is so encouraging.

No one said it would be easy.

By all accounts, RedMart had been struggling for a while. Singapore is not an easy market to operate in. Consumers are sophisticated (read: informed and hard to please) while also being price-sensitive (read: informed and hard to please). On the back-end, costs are high. Large rounds of later-stage start-up financing are still not easy to come by in South-East Asia and RedMart was certainly one of our few trailblazers in this regard. Which is why, when the company's investors threw up their hands this year, it was imperative for the health of the community overall that RedMart find a buyer.

Now, does this mean that this was a profitable sale for investors? Not likely. The earliest angels may have made money if all investors had pari passu status (very unlikely) or the latest investors may have got their money back if they had LIFO liquidation preference rights (but that doesn't mean they would make a profit). Either way, few if any are going home happy.

But again, that's not the point: the entrepreneurial community, the broader matter of consumer choice and investors all reap the benefits when entrepreneurs try new things, staff join them, and the market funds them. Even if the company doesn't eventually get to the promised land.

---

The twist in all this, of course, is that Lazada was itself the beneficiary of a white knight in the form of Alibaba swooping in to rescue the company. I didn't comment on that deal when it happened because Lazada wasn't a traditional start-up in the sense we really think of one. The spillover benefits to the overall community in terms of providing encouragement to other founders weren't great, because Lazada (and other Rocket Group companies) although run by people with stellar backgrounds don't really fit the image of traditional founder-driven start-ups.

A version of this post first appeared on e27.

AT&T-Time Warner and competition policy

1/4 Proposed AT&T-Time Warner deal illustrates once again how oligopolistic US business tends to become. Not a problem IMHO as long as…

— Murli Ravi (@murli184) October 23, 2016

2/4 …real competition possible. In A) networks-effects-based firms which B) become dominant over time, A & B both vital, hard for new rivals

— Murli Ravi (@murli184) October 23, 2016

3/4 to fight incumbents. That's why, say, retail, energy, real estate and on-demand taxis don't merit the same level of anti-competitive

— Murli Ravi (@murli184) October 23, 2016

4/4 scrutiny as say, AT&T-Time Warner, MS Windows, big airlines and Facebook.

— Murli Ravi (@murli184) October 23, 2016

You read that right. I don't think Uber merits anti-trust scrutiny (yet) and do think Facebook does.  There are plenty of competitors as well as business model reasons that act to keep Uber on the straight and narrow (mostly). Facebook? Nope.

Does this mean Uber, Airbnb and Amazon will never need to be monitored? Of course not. Do I mean to imply that Facebook is in fact acting in an anti-competitive manner? No, because I don't know.

Also, does my highlighting Facebook imply that all other social networks merit the same degree of scrutiny? No. Or at least, not yet. Snapchat, Twitter, Pinterest and arguably even WhatsApp don't qualify as dominant even though they obviously have huge network effects.

Another thought to keep in mind is that competition regulators might be better off cooperating with each other, sharing information and formulating regulations holistically, rather than trying to regulate companies in a patchwork manner. The cost of not coordinating efforts is that some of the most sophisticated, fast-moving companies play off regulators against each other. Companies that fulfil my conditions A and B above have long since become pan-national and regulatory actions taken at the national level cannot possibly be sufficient.

I do acknowledge that what I am saying is much easier said than done because of conflicting objectives such as the desire to build up national champions at the expense of external interlopers or differing cultural values on issues such as consumer privacy and wage policy. That's without even considering the cost, time and expertise involved in coordinating large multi-agency efforts.

Then again, if, say, multiple national police agencies can cooperate on criminal investigations spanning multiple jurisdictions, then that could well offer a template for anti-trust agencies to explore.

Whither 2016? Prognostications on IoT, the venture capital model in Asia and more

Last January, I jumped on the Here Are My Predictions For The Year bandwagon and came up with this set of 8 predictions. Although I was not rigorous enough to attach probabilities to my predictions, you can see that some of my predictions were deliberately bold: in the start-up world, the mantra is, “Go big or go home." You can judge for yourself how well I did.

 

This year, I suddenly find myself in March not yet having put out a similar list for 2016 even though I’d consistently Evernoted down some thoughts all through 2015. (An anti-prediction: “Evernoted” will never become a colloquialism!) No matter — here we go now. This year’s list is somewhat eclectic and doesn’t focus just on start-ups and venture capital.

 

As usual, you’ll find a TL;DR version at the bottom.

 

1. “IoT" is dead, long live IoT.

Just like “on-demand” last year, “big data” a couple of years ago and “daily deals” before that, I think this is the year the term “IoT” will become so widely used as to become meaningless, banal, even uncool. This doesn’t mean that the move towards an Internet of Things will no longer be a truly powerful trend. The trend will, in fact, accelerate. But when anything with a chip in it can be said to belong to the Internet of Things, which could encompass everything from wearables to logistics tracking technologies, home appliances, entertainment gadgets, cars and more, then I argue that the term not only loses its meaning but actually inhibits clarity of thought on how to tackle a given opportunity. (In the same manner, I've heard people declare with a straight face that their little Excel spreadsheet with a few thousand rows of data constitutes Big Data — urgh.)

 

As “IoT” itself becomes meaningless, what I expect to see is more specific terms within the broader IoT universe gaining currency, reflecting specialisation and maturation of yesterday’s sub-niches.

 

2. Fintech touching consumers more directly and in more ways

Today, most ordinary consumers don’t know what “fintech” means. In Asia, most don’t even experience fintech directly except occasionally in the area of payments. This will change. Fintech underpins pretty much every aspect of banks’ internal operations as well as many of the ways institutional/corporate clients interact with banks. This year I expect to see the beginnings of fintech directly touching consumers. Untapped opportunities exist in personal financial planning, term insurance, investment execution, credit, security as well as in payments itself. Each of these is a vast area.

 

3. Rationalisation and specialisation of VC firms in Asia

Even with recent market turmoil, there is plenty of growth capital available to start-ups in Asia from non-VCs. These include wealthy individuals and families, corporations, government agencies, crowd-funding platforms and more. Most of these actors don’t invest for a living and therefore may have lower return thresholds and/or different measures of success from traditional VC firms. This means that VCs who fail to recognise this new reality (where they are not the sole source of capital) will get blown away in negotiations and fail to win deals. Over time, these are the VCs who will end up as has-beens. Other, smarter VCs will focus on building their strengths in specific industry sectors, stages of development or other areas of specialisation, which will help them make a compelling and differentiated case for winning the right to invest in companies.

 

Note that phrase: "the right to invest". This has not historically been the way capitalists in Asia have thought.

 

To be clear, I expect this to be a multi-year trend, not just in 2016: the venture capital business has very long lead times and poorer performing firms don't die overnight, instead fading away over several years.

 

4. Privacy and personal control renaissance

The proliferation of smartphones and Facebook’s continued insinuation of itself into every aspects of many people’s private lives will, I hope, finally have the effect of provoking a substantial backlash and precipitating a move towards greater personal control for users over their data, phone features, notifications, etc. (I call out Facebook because they are the biggest culprit but they are by no means the sole culprit in this regard; LinkedIn may be worse and several others deserve blame too.)

 

I think there will continue to be a number of people for whom privacy does not matter, but I think others will demand and gravitate towards services that respect users’ desire for full control over their own information and devices. How could this come about? At multiple levels of abstraction, ranging from changes to mobile operating systems — I hope my next phone will allow me to set app permissions at a much more granular level instead of presenting me with a take-it-leave-it "choice" when I install a new app — through to businesses built around VRM-like principles, and government regulation and standards-body frameworks concerning what’s permissible.

 

5. Push back against always-on towards rediscovering mindfulness and focus

This is a strange suggestion for me to be making given that I’m writing primarily about tech. I think this is the year that many of the earliest adopters of tech will say, “Enough.” I see this in my own habits — and I've been online since the early 90s — but this isn’t (purely) a self-reflective prediction. It also comes from my observation of the behaviour and expressed desires of hipsters (I know, I know…), very busy and tech-savvy people, and other early adopters of new trends.

 

Ok, so that may strike you as anecdotal — but take a look at how often people have searched for the term "mindfulness" on Google. As a comparison, you can add other terms like “stress” or “meditation” that you might reasonably expect to be correlated with mindfulness. The growth in searches for "mindfulness" has been much greater (off a smaller base) than these other terms.

 

The implication? Perhaps users choosing not to use apps that incorporate infinite scroll and create a state of permanent distraction. Or more single-button apps, i.e., press a button, get what you want, move on. More prominence given to Do-Not-Disturb modes. More services, apps and OS features that actually help with focus and productivity rather than always-on and interruptions.

 

All of these will change the way we work, play, learn and live.

 

TL;DR

1. “IoT" is dead, long live IoT.

2. Fintech touching consumers more directly and in more ways

3. Rationalisation and specialisation of VC firms in Asia

4. Privacy and personal control renaissance

5. Push back against always-on towards rediscovering mindfulness and focus

6. “Evernoted” will never become a colloquialism ;-)

 

8 predictions for Asia’s startup scene in 2015

It’s that time of year when tech pundits everywhere emerge with prognoses for the year ahead. I’m jumping on the bandwagon too.

TL;DR

1. Early-stage valuations in India will continue to build towards a breaking point
2. Startup valuations in Singapore will continue rising unabated
3. Facebook will make its first big Asian acquisition
4. Google will continue to NOT make acquisitions in Asia
5. Sovereign investors will follow GIC’s and Temasek’s lead into funding late-stage tech firms
6. Large Chinese tech firms will invest outside their home market
7. Southeast Asia will see its first US$500 million tech startup exit
8. Japanese firms’ investment activity in Emerging Asia will remain steady

Let's get into the details.

1. Early-stage startup valuations in India will continue to build towards a breaking point.

India is currently seeing quite a bit of excitement in the early-stage funding scene. By “early-stage” I mean companies raising (much) less than US$5 million in financing in their last round. There are sound reasons for the excitement — accelerating growth in smartphone usage, positive developments on the government and policy front, increased availability of early-stage capital, and a lot else. Yet at the same time, it is fair to say that valuations may be getting a little ahead of themselves.

I think there is still some room for valuations to build further, driven mostly by capital chasing high-quality entrepreneurs in attractive market segments, a combination that still remains scarce today. That is the bull case.

The bear case is that valuations will over-shoot and the trend will break — and this is what I think will happen. All it will then take is for a couple of leading venture capitalists, who have a lot of available capital in comparison to the size of these small early-stage rounds, to take a breather. I anticipate that this will happen in the second half of the year. These bigger funds will then retreat to their original focus on growth-stage investments and leave the early-stage scene to specialists and smaller funds.

2. Startup valuations in Singapore will continue rising unabated.

Although Singapore is very small and hasn’t produced big successes on a consistent basis, government support in one form or another — grants, equity funding, smart promotion of Singapore to entrepreneurs and financiers outside Singapore, assisting Singapore-based firms with commercialization and internationalization, educational resources, and more — has resulted in Singapore punching well above its weight in terms of the number of early-stage ventures at the top of the funnel. Even more so than in India, early-stage startups in Singapore have lots of sources of capital to choose from (and not just from the government). This implies that any decent early-stage startup can raise initial rounds of funding on attractive terms.

There’s nothing to prevent more such startups being founded and many more will continue to be. However, it will remain true for the foreseeable future (not just in 2015) that the amount of early-stage capital available will easily dwarf the number of companies that can absorb that capital. This underpins my prediction.

Whether this story will eventually have a happy ending — and who will be its beneficiaries — is something to think about.

Note however that this is only about early-stage startups as I defined earlier. Among startups that are at series B and beyond, attractive companies will continue to raise money at appropriate valuations while others may find themselves plateauing or worse turning into zombies. At that stage, it is difficult to talk about valuations in broad-brush language.

3. Facebook will make its first big Asian acquisition.

Facebook has made many acquisitions in the US and a few elsewhere. In contrast, their deal activity in Asia has been subdued. In Emerging Asia (loosely defined as India plus Southeast Asia), Facebook has made only two acquisitions – an Indian startup called Little Eye Labs and Malaysian company Octazen.

This will change.

I think Facebook will make its first big acquisition in Emerging Asia in 2015, where I define “big” in this context as anything valued at over US$50 million.

My reasoning is straightforward. The two biggest countries in Emerging Asia are also two of Facebook’s most important markets: India and Indonesia. Facebook’s India userbase is set to surpass its US userbase in the near future. Indonesia is not far behind.

As to what Facebook might buy, I think it’s very unlikely Facebook will purchase a company purely to bulk up its userbase. Facebook’s traction is already so large and has such strong in-built network effects that it doesn’t make sense to do this. They bought Whatsapp not (only) because it had a large userbase — of which a big proportion is in India and Indonesia — but because it brought them exposure to a new, parallel area. The same is true of Instagram, Oculus Rift, Internet.org, Little Eye Labs, and so on.

So what might Facebook buy in Emerging Asia that operates in a parallel domain and warrants a good-sized cheque? Here we move away from prediction and towards speculation, but here are some ideas anyway:

• Technologies that enable Facebook to acquire, service, and retain new kinds of users – the types who don’t fit the profile of the typical Facebook user today. This is a stated mid and long-term ambition for Mark Zuckerberg (Internet.org being one manifestation of this goal) and Emerging Asia would be the ideal place to run a large-scale experiment that could be rolled out to the rest of the world if it succeeds.
• Companies that offer (access to) new kinds of content that Facebook doesn’t have familiarity with today. As an analogy: the biggest app stores in China aren’t Apple’s and Google’s official stores; they’re actually third-party marketplaces. People are trying the same thing in Indonesia and there’s no reason why the trend should stop with Indonesia. Facebook is already a platform in more ways than one and this could be a way for them to become an even more comprehensive cross-OS platform. It would have the added benefit of helping with user retention as they traded up from feature phones to smartphones or from low-end smartphones to higher-end devices.
• Infrastructure. As early as mid-2012, Facebook was far-sighted enough to invest in a submarine network cable in order to secure guaranteed capacity and the ability to better serve users in Emerging Asia. If I were them, I would look closely at doubling down on other similar bets.
• On the customer (B2B) side, smaller companies that have the distribution capability to sign up advertisers or tools to help those advertisers achieve their goals more easily.

4. Google will continue to not make acquisitions in Asia.

Google is perhaps the most global Silicon Valley company I can think of in terms of its userbase. Yet, most of its commercial focus has continued to remain the US and Europe. With the majority of its revenue still coming from its core search advertising business, Google doesn’t gain much by aggressively pursuing users, new technologies, or even advertisers in Emerging Asia. This isn’t to say that they will not grow their business organically in Emerging Asia; just that M&A dollars are far more easily justified in their home market. I’d love to be proven wrong with this prediction.

5. Sovereign investors will follow GIC’s and Temasek’s lead into funding late-stage tech firms.

Singapore’s two government-backed investment companies GIC and Temasek Holdings have raced ahead of their peers in terms of getting exposure to technology titans in the US, China, India and elsewhere, seemingly as a result of putting a formal programme in place rather than merely making opportunistic investments. Temasek-affiliated firms like Vertex Venture Holdings and Pavilion Capital have given Temasek further exposure to tomorrow’s potential titans. I don’t know of too many other sovereign investment firms that have done the same thing, barring the occasional one-off deal such as the Qatar Investment Authority’s recent investment in Uber. I think this will change this year, as other sovereign investors start to pursue large tech companies who continue to stay private well past the US$1 billion valuation mark.

6. Large Chinese tech firms will invest outside their home market.

Many Chinese tech companies are already quickly becoming household names outside China by virtue of organic expansion and some investment activity. We can expect investment and M&A activity by these large technology companies to substantially ramp up in 2015, and specifically in Emerging Asia and the US. The most active players will be ones like Alibaba, Tencent and Baidu, but several other less known players will also appear on the scene.

7. Southeast Asia will see its first US$500 million tech startup exit.

This is perhaps the one prediction I am least comfortable with. I am aware of Southeast Asian companies that have already reached this valuation level. What I am less sure of is whether they will achieve an exit this year.

8. Japanese firms’ investment activity in Emerging Asia will remain steady.

Over the past few years, Japanese investors have, for a variety of reasons — fall-out from Fukushima, contracting home market, desire to diversify away from China — looked at Emerging Asia as an attractive investment destination. Japanese firms generally make decisions in a careful and deliberate manner and thereafter tend to “stick with it” unless and until there are strong countervailing reasons to change course. Investments by Japanese firms in Vietnam, Indonesia, India and other countries in Emerging Asia have generally followed this pattern and I anticipate that this will continue in 2015.

The one exception to the steady-as-she-goes approach is SoftBank (of course), which has made big bang investments in Emerging Asia recently and is likely to follow these up with more such deals. A related investor is Visionnaire Ventures, a joint fund by Temasek Holdings and Taizo Son, brother of Softbank’s founder Masayoshi Son.

So there you have it.

---

I focus in this post mostly on investment activity, not broader predictions about attractive new opportunities for startups and VCs because:

• such predictions are easier to evaluate within a defined period, and
• other trend predictions will either take longer than a year to play out or likely occur within a year but be so banal as to be meaningless.

---

This piece first appeared on Tech in Asia. 

 

Book review: Sync

Sync: The Emerging Science of Spontaneous Order by Steven Strogatz

Rating: **** ¹/₂

My review on Goodreads:

Superb read! My one and only small quibble is the paucity of illustrations. Strogatz mentions at one point that the Lorenz attractor is fascinating but hard to describe in words. He then goes on to try and describe it! Just show us already! I already knew what the Lorenz attractor looked like but I can imagine other readers' frustration. In another location, it was about how someone is amazed by what he sees on his screen -- again I felt cheated at not being able to experience the same feeling. I can understand that Strogatz's intentions (or perhaps it was the intention of his editor and publisher) was to not scare away the non-mathematically inclined with complicated-looking charts. But tantalising yet nebulous textual descriptions actually detract from that very goal. If it had a few illustrations to complete this otherwise superb book, it would be a 5 for sure.

I posted a link to the review on my Twitter account and was astonished to see a quick and truly humble reply from Strogatz himself!

click to zoom

Humility definitely helps when it comes to exploring new ideas -- something that Strogatz is obviously very, very good at -- yet I wouldn't say humility is all that common in the academic community!

The next big opportunities from the next Silicon Valley

This piece first appeared on Ask(Your)VC.

That title isn't clickbait: I'll tell you upfront that I don't believe that there is any such thing as the Next Silicon Valley. Apples to apples comparisons across regions are mostly meaningless. Now that that's out of the way...

I was invited to speak on this topic at dmexco earlier this month, Europe's largest digital media conference and expo, with 30,000 attendees. The discussion was moderated by Terry Kawaja of Luma Partners and my co-panelists were Marvin Liao of 500 Startups and Stefan Gross-Selbeck of BCG Digital Ventures. The conference organisers have produced a slick recording of the session here.

Although I don't believe in notions like the Next Big Thing, I do strongly believe that there is globally unparalleled potential in what I term Emerging Asia.

South and South-East Asian countries are a few years behind Japan, Korea and China, and still have poor broadband coverage on the whole, low digital media spend, low incomes relative to the developed world, etc etc. Yet these same marketes contain in aggregate close to 2 billion of the world's people. Their median age is a mere 26 and the fastest growing demographic is aged 15-24. The top 5% wealthiest people in India have the same aspirations and spending power as the average Western consumer -- while also easily outweighing their Western compatriots in terms of sheer numbers.

People in Indonesia and Philippines are far and ahead the world's leaders in terms of time spent per day looking at screens. Not US users, not Koreans, not even the Japanese come close.

Attitudes to entrepreneurship are changing and no one bats an eyelid at the idea of launching a company in Emerging Asia that targets Western businesses as clients.

The number of VC firms actively seeking to make Series A and B investments in Emerging Asia of between $1 million and $10 million (i.e., well below Silicon Valley norms) is less than 30, possibly less than 20.

Singapore, although small and far wealthier than its neighbours, sits right in the centre of all this activity and is a natural hub for start-ups to set up regional HQs, access financing and, at a later stage, become a venue for exits.

Finally, not all of this is mere potential: over the past 6 or 7 years, we have seen about a dozen technology companies hit the magic billion-dollar valuation mark.

Isn't all this mind-boggling?

It's clear that these markets are massively attractive though not without their risks and obstacles. We discuss some of this in the video.

This is what a good day feels like

I'm travelling for work in India at the moment and just ended what was pretty much a perfect day. (This isn't so much an announcement to the world as it is a Note To Future Self.)

I flew from Bombay to Bangalore this morning for a day trip, which meant that I had to lurch out of bed at 4 am. That was the one and only thing about today that I would change.

Maybe.

The rest of it was just perfect.

My flight to Bangalore landed five minutes ahead of schedule.

I got through not one or two but six great meetings during the day, each with a smart, passionate and engaging entrepreneur. SIX! I learned a lot today. This is hands down the best part of my job.

There was even spare time in between meetings to make follow-up notes on each meeting and get other work done. I love it when things move to plan.

I met a friend quite by chance whom I haven't seen in ten years.

Old friends passed on some great news about the company they co-founded.

The drive back to the airport from Bangalore city took a mind-bogglingly short 50 minutes in peak hour. The drive is usually an uncomfortable, stop-start two-hour affair.

Checking in and getting through security at Bangalore airport took about seven minutes all in. Smooth as silk. Even with the airport undergoing a renovation, it's really a joy to get through now and even manages to look good. There, I said it.

The seat next to me on my return flight to Bombay the same evening was perhaps the only one on the plane that wasn't occupied. And to top it all off, the flight landed a full 40 minutes ahead of schedule.

Yes, my day lasted over 20 hours but tell me that isn't a good day!