I am often reminded of the sage advice from Sir John Templeton: “The four most dangerous words in investing are ‘this time it’s different’.” As investors, I think we need to question whether we are entering a new technological and machine age over the next 10-25 years that could disrupt most businesses and possibly society as we know it. In this regard, the new technological and machine age may be more important than The Industrial Revolution. Quite possibly, this time it is different and whilst heeding Sir Templeton’s advice, as prudent investors we believe it would be neglectful to ignore the technological developments that are almost certain to provide substantial threats and opportunities to businesses.
In a recent TED interview, Charlie Rose asked Larry Page (Co-Founder of Google) what is his most important lesson from business. He said that he has studied why many large businesses fail and he concluded: “They missed the future”. As investors, can we afford to miss the future? In our view, there is mounting evidence that we are approaching a tipping point of exponential technological advancement, particularly through accelerating improvements in artificial intelligence, 3D printing, genomics, computing power, and robotics.
We have numerous recent powerful lessons on the rapid disruption of businesses from technological innovation:
- In 1998, Kodak had 145,000 employees and sold 85% of all photographic film. In 1999, Kodak’s stock price peaked and in January 2012 it filed for bankruptcy. What is surprising about the Kodak story is that it invented the digital camera in the 1970s and yet the company was effectively destroyed by its own invention.
- In 1998, Nokia overtook Motorola to become the world’s largest mobile phone manufacturer. By 2007, Nokia controlled more than 40% of the mobile phone market and was highly profitable. In July 2005, Google bought Android and in January 2007, Apple launched the iPhone. In September 2013, Nokia sold its loss-making mobile phone business to Microsoft.
- Google was founded in September 1998. In 1999, newspapers’ share of global advertising revenue was approximately 35%. In 2015, Google generated advertising revenues of over US$67 billion, or 14% of global advertising. Meanwhile, newspapers’ share of global advertising revenue had fallen to approximately 12%.
Another lesson is that large scale/global disruption from technological advancements appears to be occurring at a faster and faster pace. Uber was founded in March 2009 and is now the world’s largest ‘taxi company’, with operations in 429 cities in 71 countries. Facebook was founded in February 2004 and has more than 1.6 billion monthly active users. The company is expected to generate advertising revenues more than US$20 billion this year. Airbnb was founded in August 2008 and is now the world’s largest accommodation company, with over two million listings in 34,000 cities in over 190 countries.
Exponential versus linear growth
It is difficult to comprehend that we could rapidly face a radically different world from the advancement of technology when our own experience suggests that fundamental change is occurring incrementally and at a gradual pace. A reason we may be underestimating the impact of technological change is that most changes in our life (like ageing, learning, career progression, etc.) occur in a well-established linear trajectory whereas technological progression is exponential.
In exponential growth, a measurement is multiplied by a constant factor for a given unit of time (e.g. computation power doubles every year), whereas for linear growth the measurement is added to incrementally and by a constant factor (i.e. we grow older by one year per year). Early on, it is difficult to feel the difference between linear and exponential growth (i.e. from 1,2,3,4 … to progressions of 1,2,4,8…); however, after 30 iterations the linear sequence is at 30 whereas the exponential sequence is over 500 million. In an exponential world, nothing is perceived to be changing in the early stages and then suddenly change starts occurring at an explosive rate.
There are numerous examples of technology progressing at an exponential rate. Three well-cited examples are:
Computational power - In 1965, Gordon Moore, Co-Founder of Intel, predicted that the number of transistors on an integrated circuit would double every two years (the so-called Moore’s Law). Over the last six decades, computation power has increased over one trillion times per integrated circuit. An iPhone 5 released in 2013 has twice the processing power of the 1985 Cray-2 supercomputer, which at the time was the world’s most powerful computer. At the current rate of progression, a mobile phone is likely to have the processing power of the current largest supercomputer – China’s Tianhe 2 – in around 15 years.
Genome sequencing - When the project to sequence the human genome was started in 1990, given the speed at which the genome could be scanned at that time, it was thought it would take thousands of years to sequence the entire human genome (six billion bases). The full genome was sequenced ten years later. In 2000, the cost to sequence an entire human genome was around US$100 million, and by 2015, the cost had fallen exponentially to US$1,000.
Data - it has been estimated that the amount of digital data in the world is doubling every two years. To put it another way, estimates suggest that more data has been created in the past two years than in the previous history of the human race.
In order to predict what will happen in the future through technological change, you need to extrapolate and think exponentially.
Ray Kurzweil, a natural language processing pioneer and entrepreneur, a renowned futurist and currently Director of Engineering at Google, wrote in a March 2001 paper titled “The Law of Accelerating Returns”:
“An analysis of the history of technology shows that technological change is exponential, contrary to the common-sense intuitive linear view. So, we won’t experience 100 years of progress in the 21st century; it will be more like 20,000 years of progress (at today’s rate).”
“It is important to ponder the nature of exponential growth. Toward this end, I am fond of telling the tale of the inventor of chess and his patron, the Emperor of China. In response to the Emperor’s offer of a reward for his new beloved game, the inventor asked for a single grain of rice on the first square, two on the second square, four on the third and so on. The Emperor quickly granted this seemingly benign and humble request. One version of the story has the Emperor going bankrupt as the 63 doublings ultimately totalled 18 million trillion grains of rice.”
“As exponential growth continues to accelerate into 4 5 the first half of the 21st century, it will appear to explode into infinity, at least from the limited and linear perspective of contemporary humans. The progress will ultimately become so fast that it will rupture our ability to follow it. It will literally get out of control.”
Bill Gates has commented that “we always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next 10.”
This tendency to overestimate change in the short-term and underestimate the long-term creates an interesting (and possibly dangerous) paradigm for an investor – acting too early by selling or short selling businesses that are most likely to be disrupted may well be detrimental to short-term returns, whereas waiting too long could be very costly, as in the end disruption may occur very rapidly. Judging where we are on the exponential path of technological development is becoming critical for any longer-term investor.
In thinking about the investment impact of exponential growth, it is instructive that five of the world’s ten largest companies by market capitalisation are currently technology companies (Apple, Alphabet, Microsoft, Facebook, and Amazon) and three of these companies did not exist less than 25 years ago. Tellingly, all of these companies are making significant investments into artificial intelligence.
Artificial intelligence is the field of computer science involved with the development of computer systems capable of performing tasks normally requiring human intelligence.
According to Nick Bostrom, an Oxford philosopher and leading artificial intelligence thinker, the progression of artificial intelligence can be split into three categories:
- Artificial Narrow Intelligence (Narrow A.I.) is artificial intelligence that specialises in one area, such as a computer beating the world chess champion or winning the quiz game, Jeopardy. Narrow A.I. is responsible for Facebook automatically labelling your friends in photos, for Amazon and Netflix making personalised product and video recommendations, and for airline reservation systems automatically setting prices. Google search is likely the world’s largest Narrow A.I.; ranking, sorting, and retrieving relevant information from across the internet. Computer engineers are rapidly advancing Narrow A.I. with many uses today such as search, translation, voice recognition, and image recognition.
- Artificial General Intelligence (AGI) is a computer system that is as smart as a human across any intellectual task (including complex reasoning, thinking abstractly, and learning from experiences). It is said that an AGI will be unpredictably creative. This will require a computer to have the mental capacity to solve problems, think creatively, understand language, interpret images, think abstractly, learn quickly, and learn from experience. While it is likely to take many years to develop a computer system that has AGI, it would appear that the building blocks for AGI are being rapidly developed, with material advances in machine learning, voice and image recognition, computational power, and the development of advanced neural networks.
- Artificial Superintelligence or Singularity is when computer intelligence surpasses human intelligence and then rapidly advances to say a billion times more powerful than all human intelligence. Leading thinkers, including Stephen Hawking, Bill Gates, and Elon Musk, have warned of Artificial Superintelligence posing an existential threat to humanity. Others, including Google’s Eric Schmidt and IBM’s Gini Rometty, have argued that artificial intelligence is a positive force that will augment human abilities.
In thinking about whether AGI remains in the realms of science fiction and whether it’s likely within a reasonable timeframe (maybe 10-15 years), it is worth reflecting on some recent comments by Mark Zuckerberg (Founder and CEO of Facebook):
“So, the biggest thing that we’re focused on with artificial intelligence is building computer services that have better perception than people. So, the basic human senses like seeing, hearing, language, kind of core things that we do, I think it’s possible to get to the point in the next five to 10 years where we have computer systems that are better than people at each of those things.”
Zuckerberg continues, describing how artificial intelligence can improve the experience of Facebook users: “That’s because today our systems can’t actually understand what the content means. We don’t actually look at the photo and deeply understand what’s in it or look at the videos and understand what’s in it, or read the links that people share and understand what’s in them. But in the future, we’ll be able to, I think on a five or 10-year period.”
Jeff Bezos, Founder and CEO of Amazon, said in an interview at the 2016 Code Conference: “I think it’s gigantic. I do. I think natural language understanding, I think machine learning in general, artificial intelligence…it’s probably hard to overstate how big of an impact it is going to have on society over the next 20 years. It is big.”
He also spoke of the growing synergy between data, computation power and language, stating that “The combination of newer and better algorithms, vastly superior compute power and the ability to harness huge amounts of training data – those three things are coming together to solve some previously unsolvable problems, and they are going to drive a tremendous amount of utility for customers and customers are going to adopt those things.”
Are we nearing a tipping point?
We believe there is evidence that technology may be nearing a tipping point – technology is now advancing at such a rate that a breakthrough in AGI may be rapidly approaching.
Firstly, we believe that the world’s major technology companies are collectively assembling the equivalent of the “Manhattan Project” that led to the development of the atomic bomb in World War II. Companies such as Alphabet (Google), Facebook, Microsoft, IBM, Alibaba, Baidu, Amazon, and Apple are investing unprecedented amounts of money in artificial intelligence research and development, expansion of computational power, collation of the world’s data and knowledge and assembling the world’s leading intellectual capital by hiring leading graduates and researchers/scientists in fields of artificial intelligence and computer engineering from the world’s leading universities.
Secondly, over the last few years, there have been dramatic advances in machine learning, voice and image recognition, machine understanding of language (machines can now read and understand documents) and the early development of quantum computers. Each of these areas appears important in the development of AGI, and it seems reasonable to expect accelerating advances in the years ahead.
Finally, March 2016 may well be remembered as a seminal moment in the advancement of artificial intelligence, when AlphaGo (a computer program developed by Google DeepMind) beat the Go world champion, Lee Sedol, in four out of five games. Experts had predicted that a computer program would not master Go, an ancient Chinese board game still played today, for another decade given the complexity of the game. There are apparently more possible moves in a game of Go than there are atoms in the universe. The breakthrough with AlphaGo is that it is a self-learning algorithm that learns from raw data. AlphaGo taught itself to play by playing itself 30 million times. Google DeepMind’s website states: “The algorithms we build are capable of learning for themselves directly from raw experience or data, and are general in that they can perform well across a wide variety of tasks straight out of the box.” An algorithm that learns for itself is a fundamental building block of developing AGI.
The potential commercial advantages for companies that create a lead in the development of AGI are enormous and may well be definitive. The winners in the AGI arms race are likely to have access to:
- the best intellectual capital;
- massive computing power; and
- vast data across all areas (personal, written documents, image/video).
In making any predictions, and at the risk of appearing naïve in hindsight, I am reminded of the famous quote from one of the most prominent names in American Baseball, the late Yogi Berra: “It’s tough to make predictions, especially about the future.” With this health warning, it appears possible/likely over the next 10-20 years that the following could occur:
(1) Development of an intelligent virtual personal assistant that knows who you are, understands natural language, anticipates what you want, reads, and understands your email, is able to answer most questions and organises your life. The major global digital platforms all have early virtual personal assistants including Google Assistant, Apple’s Siri, Facebook’s M, Amazon’s Alexa, and Microsoft’s Cortana.
(2) Development of augmented and virtual reality through widespread adaptation of new computing interfaces.
Augmented reality is a live direct or indirect view of a physical, real-world environment whose elements are augmented (or supplemented) by computer-generated sensory input such as sound, video, graphics, or GPS data. Augmented reality interfaces have been used by the military where data is delivered and viewed as a heads-up display on the inside of a helmet’s visor. Microsoft is developing HoloLens which is an augmented reality interface that can project images to supplement the wearer’s physical vision. It is likely that HoloLens and other devices will be initially developed for use in commercial settings, such as to interact with architectural blueprints or to display schematics to a surgeon or to a mechanic. Over time, augmented reality could become an integrated communication interface, where a video or holographic representation of people you are communicating with, or information you are reviewing, are projected into your world.
Virtual reality replaces the real world with an immersive simulated experience. In the future, a virtual reality experience may be indistinguishable from a real-life experience. Initial applications of augmented virtual reality are focussed on entertainment, such as Pokémon Go. In the longer term, virtual reality could replace any real-world experience such as shopping or an overseas excursion for a group of school children. Facebook’s Oculus Rift, HTC Vive and Google Cardboard have already launched their virtual reality devices and software.
(3) Digitisation of goods via the mass commercialisation of 3D printing (also known as additive manufacturing). The mass commercialisation (with exponential reduction in cost and improvement in performance) of 3D printing could ultimately result in most hard (and possibly soft) goods becoming virtual goods, similar to what has happened in the music and book publishing industries. While 3D printing is primarily currently used for industrial prototyping, early consumer applications are appearing, like Adidas’ Futurecraft 3D, a running shoe made with a 3D-printed midsole tailored to the consumer’s foot. With the commercialisation of 3D printing, many goods we consume could effectively be digitalised with a source code that could be downloaded from the internet and printed at home or at a local 3D printing facility (UPS is already setting up a network of local 3D printing facilities in the United States).
(4) Digitalisation/automation of white-collar tasks. It is likely that advancements in artificial intelligence (machine learning, voice and language understanding, image recognition) will lead to the development of software that could one day replace many white collar professionals (lawyers, accountants, journalists, doctors, dentists, and fund managers). A 2013 paper concluded that “according to our estimate, 47 percent of total US employment is in the high-risk category, meaning that associated occupations are potentially automatable over some unspecified number of years, perhaps a decade or two.”
(5) Development of advanced fully autonomous specialised robots, which could replace most highly specialised manual labour such as surgeons. In May 2016, the Smart Tissue Autonomous Robot (STAR) undertook complex soft tissue surgery on pigs with a lower error rate than human surgeons.
(6) Commercialisation of driverless cars and expansion of car sharing. Google’s self-driving car fleet has driven more than 1.5 million miles with only one minor accident, demonstrably safer than humans. Should regulatory hurdles be cleared, it would appear likely that within the next 10-20 years, fully autonomous driverless cars will become widely available. The social implications for driverless cars are enormous. Driverless cars have the potential to dramatically reduce road accidents and the 1.2 million global road deaths annually. Through networked intelligence, autonomous cars are likely to dramatically reduce road congestion, bringing down transportation costs. They also unlock peoples’ travel time for productivity or to engage in entertainment. A car is the second largest capital expenditure item in most developed market households, yet they are utilised only approximately 5% of the time. Gabe Klein, futurist and former Head of the Chicago and Washington D.C. transportation departments, has estimated that driverless cars and expanded car sharing could reduce the number of cars by 85% and dramatically increase urban space – approximately 25% of the land area of Los Angeles is taken up with roads and car parks.
(7) Breakthrough advances in medical technology and longevity. It appears likely with continued advancements in genome sequencing epigenetics, gene editing, image recognition, machine learning, big data, and nanotechnology/ implantable devices that there will be major breakthroughs in medical science over the next 10- 20 years. Improvements in data collection, sharing and analysis will further accelerate breakthroughs and shift healthcare from ad hoc care to one that is personalised and primarily preventive and proactive. These advancements are likely to improve health, reduce disease and possibly radically extend human lifespan. Development of AGI (assuming it occurs) and genomics is likely to advance medical science to the point that we will eventually solve the causes of ageing and virtually all known medical problems. The speed of advances in medical technology is likely to be slower than in other areas, and meaningful advances in longevity may well be outside the 10-20-year timeframe due to the need to pass regulatory approval to ensure full diligence checks are completed, and the many complex ethical issues are addressed, particularly around reprogramming the human genome.
(8) Development of humanoid intelligent robots (i.e. robots that look and sound and react exactly like humans). A humanoid robot could undertake many tasks that require human interface, such as a receptionist, a store assistant, a waiter, or bartender. A humanoid robot could eventually become a personal/home assistant, or even a companion.
It is likely that advancements in technology will disrupt many industries and economies over the next 10-20 years. Issues that spring to mind include:
- Will personal digital assistants begin to erode the value of brands and advertising as a means of product discovery? What will happen to the media companies that rely on advertising?
- Will advancements in 3D printing and robotics lead to the re-localisation of manufacturing, effectively reversing globalisation? Will this lead to the future loss of hundreds of millions of manufacturing and transportation jobs in emerging markets? Will 3D printing and advanced robotics change the paradigm in manufacturing economies of scale?
- Will the advancement of 3D printing technology lead to the digitisation of many goods we purchase? What will the digitisation of goods mean for manufacturers and brand owners? Will they own the source code for goods of the future or will many goods be openly sourced and be essentially free of brand and intellectual property costs? Who owns the intellectual property in a digital world? Will many/most manufacturing industries go the way of newspapers, books, and music?
- Will 3D printing and drones shrink existing supply chains and distribution systems? What will be the role of wholesalers and retailers? Will many existing transportation/logistics companies cease to exist?
- Will integration of virtual/augmented reality and 3D printing lead to the decline of the retailing industry as we know it? Will consumers ever need to visit a shop when they can experience shopping in virtual reality, including trying on clothes, and then downloading the digital source code for the good and having it printed on their 3D printer?
- How will driverless cars and expanded car sharing impact car manufacturers and insurance companies? With vast amounts of real estate being freed up, what will happen to land prices? What will be the impact on employment and economies with the mass displacement of truck and taxi drivers around the world (there are 3.5 million truck drivers in the United States alone)?
- Will the falling cost of renewable energy and battery technology lead to an all-electric transport future? What will renewably-fed electric vehicles do to the demand for oil and gas? Will energy companies (oil, gas, and coal) cease to exist in the future?
- Will advances in language, speech, and image/ video recognition lead to the automation of many knowledge-based human tasks over the next 10- 20 years? Will this displace millions of white-collar workers such as accountants, lawyers, doctors and even fund managers in the years ahead?
- Will advances in machine learning and robotics lead to the displacement of highly specialised white-collar jobs like surgeons?
- How will people spend their leisure time in the future? Will we travel or experience things in virtual reality? What will this mean for hotels, airlines, and airports? Will people still commute to a (remaining) job if they can experience it virtually? What will this mean for car manufacturers and toll roads?
- Will people leading extremely long and healthy lives still require pharmaceuticals? Will hospitals, retirement homes, funeral homes and crematories face massive overcapacity?
As investors, we need to assess many issues that could arise from rapidly advancing technology, including:
- Over what timeframe will technology develop and what will be the pace of disruption? How will regulation mute the pace of innovation and disruption?
- Are there industries that will be safe or immune in this rapidly advancing world? It appears likely that people will still go to KFC and drink coffee!
- What industries/companies are best positioned to win from advancing technology? Technology platform companies appear well positioned, although picking the winners may prove difficult.
- Who will control the distribution of digital goods?
- Will there be fundamental shifts in widely held investment beliefs, such that:
- As an investor, can you continue to rely on a reversion to the mean in this new world? Is this consistent with an exponentially changing world? Evolutionary processes do not appear to mean revert; rather, they adapt and improve.
- What should we be assuming about the growth of emerging markets over the next 10-20 years? Can we rely on a rise in the middle class in emerging markets if manufacturing jobs are massively displaced?
- Is it right to assume rising demand for health care services if there are radical breakthroughs in medicine that dramatically reduce the incidence of age-related illnesses?
- Could 3D printing, driverless cars, machine learning and automation lead to massive labour displacement and persistent structural deflation? How will this affect interest rates and savings in the longer term?
- What assumptions should we make about longevity and the impact of ageing? What would this mean for retirement ages and adequacy of pensions?
- How do you value businesses in this rapidly disruptive world?
As a society, it is likely that we will be facing a rapidly changing landscape which will raise many issues, including:
- What are the jobs of the future?
- Will inequality continue to widen as the number of well-paying jobs shrink?
- Will there be increasing social instability, leading to a rise in radical political parties?
- Where are the boundaries for privacy?
- Will certain companies become too powerful?
- What will be the geopolitical implications of the reducing demand for fossil fuels?
- Will there need to be a universal basic income to support people who have been permanently displaced from the workforce?
- Can policy and regulation keep pace with a rapidly advancing world? Will regulation stifle innovation?
- Where are the ethical boundaries for genomics?
- Will advances in solar and battery technologies and electrification of transportation solve problems associated with climate change?
In our view, disruptive and profound changes to businesses, industries, and economies from exponential advances in technology appear to be ever closer to our doorstep. As investors, we need to carefully weigh up nearer-term investment opportunities against the likelihood of exponential progress and be prepared and positioned for fundamental and disruptive change over the longer term. The risk is that we will fail as investors if we fail to see the future. This time it may well be different.
Written by Hamish Douglass, CEO, CIO, and Lead Portfolio Manager of Magellan Financial Group: (VIEW LINK)