The productivity paradox

Tech progress
In a July 2015 interview with Edge, an online magazine, Pulitzer Prize-winning veteran New York Times journalist John Markoff articulated a widespread idea when he lamented the deceleration of technological progress. In fact he claimed that it has come to a halt.i He reported that Moore’s Law stopped reducing the price of computer components in 2013, and pointed to the disappointing performance of the robots entered into the DARPA Robotics Challenge in June 2015 (which we reviewed in chapter 2.3).

He claimed that there has been no profound technological innovation since the invention of the smartphone in 2007, and complained that basic science research has essentially died, with no modern equivalent of Xerox’s Palo Alto Research Centre (PARC), which was responsible for many of the fundamental features of computers which we take for granted today, like graphical user interfaces (GUIs) and indeed the PC.

Markoff grew up in Silicon Valley and began writing about the internet in the 1970s. He fears that the spirit of innovation and enterprise has gone out of the place, and bemoans the absence of technologists or entrepreneurs today with the stature of past greats like Doug Engelbart (inventor of the computer mouse and much more), Bill Gates and Steve Jobs. He argues that today’s entrepreneurs are mere copycats, trying to peddle the next “Uber for X”.

He admits that the pace of technological development might pick up again, perhaps thanks to research into meta-materials, whose structure absorbs, bends or enhances electromagnetic waves in exotic ways. He is dismissive of artificial intelligence because it has not yet produced a conscious mind, but he thinks that augmented reality might turn out to be a new platform for innovation, just as the smartphone did a decade ago. But in conclusion he believes that “2045… is going to look more like it looks today than you think.”

It is tempting to think that Markoff was to some extent playing to the gallery, wallowing self-indulgently in sexagenarian nostalgia about the passing of old glories. His critique blithely ignores the arrival of deep learning, social media and much else, and dismisses the basic research that goes on at the tech giants and at universities around the world.

Early washing machine

Nevertheless, Markoff does articulate a fairly widespread point of view. Many people believe that the industrial revolution had a far greater impact on everyday life than anything produced by the information revolution. Before the arrival of railroads and then cars, most people never travelled outside their town or village, much less to a foreign country. Before the arrival of electricity and central heating, human activity was governed by the sun: even if you were privileged enough to be able to read, it was expensive and tedious to do so by candlelight, and everything slowed down during the cold of the winter months.

But it is facile to ignore the revolutions brought about by the information age. Television and the internet have shown us how people live all around the world, and thanks to Google and Wikipedia, etc., we now have something close to omniscience. We have machines which rival us in their ability to read, recognise images, and process natural language. And the thing to remember is that the information revolution is very young. What is coming will make the industrial revolution, profound as it was, seem pale by comparison.

Part of the difficulty here is that there is a serious problem with economists’ measurement of productivity. The Nobel laureate economist Robert Solow famously remarked in 1987 that “you can see the computer age everywhere but in the productivity statistics.” Economists complain that productivity has stagnated in recent decades. Another eminent economist, Robert Gordon, argues in his 2016 book “The Rise and Fall of American Growth” that productivity growth was high between 1920 and 1970 and nothing much has happened since then.

1970s car breakdown

Anyone who was alive in the 1970s knows this is nonsense. Back then, cars broke down all the time and were also unsafe. Television was still often black and white, it was broadcast on a tiny number of channels, and it was shut down completely for many hours a day. Foreign travel was rare and very expensive. And we didn’t have the omniscience of the internet. Much of the dramatic improvements that have improved this pretty appalling state of affairs is simply not captured in the productivity or GDP statistics.

Measuring these things has always been a problem. A divorce lawyer deliberately aggravating the animosity between her clients because it will boost her fees is contributing to GDP because she gets paid, but she is only detracting from the sum of human happiness. The Encyclopedia Britannica contributed to GDP but Wikipedia does not. The computer you use today probably costs around the same as the one you used a decade ago, and thus contributes the same to GDP, even though today’s version is a marvel compared to the older one. It seems that the improvement in human life is becoming increasingly divorced from the things that economists can measure. It may well be that automation will deepen and accelerate this phenomenon.

The particulars of the future are always unknown, and all predictions are perilous. But the idea that the world will be largely unchanged three decades hence is little short of preposterous.

Productivity Paradox

Advertisements

Reflections on Dubai’s World Government Summit

IMG_0449

I lived in Dubai for three years in the early 1980s. The United Arab Emirates (UAE) was a very young country then, but its ambition was clear. The tallest building was the Dubai Trade Centre, at 30 stories. In fact, it was the tallest building in the whole of the Middle East at the time, and many people thought it was a folly: why build a skyscraper in the desert? It was a fair question: the road leading south from the Trade Centre towards Abu Dhabi was flanked on both sides by desert.

Now the Trade Centre is a smallish shrub in a forest of skyscrapers, most of them much taller, that has sprung up to accompany the traveller all the way to Jumeirah, 20 kilometers to the south.

Dubai is still run by the man who was effectively in charge when I worked there, and despite Emirate’s immense achievements, the ambition has not dimmed. The World Government Summit held there this week was a bold statement to the world: Dubai is looking to the future; it is visionary and optimistic.

IMG_0465
In particular, Dubai is looking forward to the opportunities presented by our most powerful technology: artificial intelligence. It has appointed the world’s first Minister for AI, and in the Summit’s opening address, the Minister for Cabinet Affairs spoke knowledgeably about many of the opportunities and challenges that AI presents.

No country is perfect, and Dubai is not without its problems, both of reality and of image. This isn’t an article about those, but one thing that struck me was that the ratio of men to women was far more equal among the local delegates than among the foreign contingent.

Dubai’s optimism is sailing into headwinds. The wealthy Russians who used to spend freely in the emirate’s luxury malls have had their wings clipped by Western sanctions against Putin’s aggressive kleptocracy. The wealthy Saudis are similarly chastened by the revolution taking place the other side of the Liwa desert, and the rift with Qatar does not help. Local businesses are chafing under new restrictions and fees that all this is causing.

No-one knows whether these will prove temporary setbacks, or whether they will cause lasting damage. For now at least, But Dubai is looking confidently towards the future.  

Landing back in the UK, the comparison is stark.

The metropolitan elite of Murdoch, Dacre, Farrage, Banks, Johnson, Rees-Mogg, Redwood, Hannan, Jenkin, and the rest have a vision of the UK as a buccaneering, free-trading nation, cutting deals with the rising powers of China and India as well as the existing superpower. Their claim that these are the fastest-growing countries in the world is obviously true, but it is far from obvious that Britain will be able to cut favourable deals with these thrusting new economies if it weakens itself by flouncing out of the powerful trading bloc in which it enjoyed such a privileged position, and churlishly insulting its former colleagues in the process. (Thanks for that, Mr Farrage.)

In any case, the metropolitan Brexiteer elite didn’t win the 2016 referendum with this vision. They knew that a population which feels wounded by globalism would not vote for the laissez-faire, low-tax, small-state Britain which would enrich the elite. Instead they appealed to our most backward-looking, resentful instincts, to our fear of the immigrant, our fear of the foreigner; they promised we could “take back control”, and head back in time to a halcyon bygone era.

Take back control

Swallowed whole by Brexit, the UK government is largely ignoring the AI-fuelled future which is bearing down on us. When they do mention it, they talk blithely about the UK being a leader in the field, not least because we are home to DeepMind (ignoring its US ownership). Most of our politicians* seem genuinely unaware that AI is currently a two-horse race between the US and China, with everyone else jostling for a distant third place.

In particular, our political leaders are mostly either ignorant of the possibility of technological unemployment, or dismissive of it, eagerly lapping up the bromides of tech giant CEOs, who mumble soothingly that automation will not cause unemployment in the future because it didn’t in the past.

Britain sparked the industrial revolution, but its leadership appears largely clueless about the information revolution. Instead it is tiny Dubai, part of a country founded as recently as 1971, which faces forward to the information revolution, and seems eager to grasp the opportunities afforded by AI, and tackle its challenges.

* The All-Party Group on AI strives valiantly to cast light into the darkness, but attendance by parliamentarians is thin. [Disclosure: I’m one of its many advisers.]

Don’t panic!

Roosevelt

Franklin D Roosevelt was inaugurated as US President in March 1933, in the depth of the Great Depression. His famous comment that “The only thing we have to fear is fear itself” was reassuring to his troubled countrymen, and has resonated down the years. If and when it turns out that machines will make it impossible for many people to earn a living, fear will not be our only problem. But it may well turn out to be our first very serious problem.

Fully autonomous, self-driving vehicles will start to be sold during the coming decade – perhaps within five years. Because of the substantial cost saving to the operators of commercial fleets, the humans driving taxis, lorries, vans and buses will be laid off quickly during the decade which follows. Within fifteen or twenty years from now, it is likely there will be very few professional drivers left.

Well before this process is complete, though, people will understand that it is happening, and that it is inevitable. Most of us will have a friend, acquaintance or family member who used to be a professional driver. And the technology that destroyed their job will be very evident. One of the interesting and important things about self-driving cars is they are not invisible, like Google Translate, or Facebook’s facial recognition AI systems. They are tangible, physical things which cannot be ignored.

Most people are not thinking about the possibility of technological unemployment today. They see reports about it in the media, and they hear some people saying it is coming and others saying it cannot happen. They shrug – perhaps shudder – and get on with their lives. This response will no longer be possible when robots are driving around freely, and human drivers are losing their jobs. This cannot fail to strike people as remarkable. Learning to drive is a difficult process, a rite of passage which humans are only allowed to undertake on public roads when they are virtually adult. The fact that robots can suddenly do it better than humans is not something you can ignore.

No doubt some will try to dismiss the phenomenon by explaining that driving wasn’t evidence of intelligence after all: like chess, it is mere computation. Tesler’s Theorem – the definition of AI as whatever we cannot yet do – will persist. But most people will not be fooled. Self-driving vehicles will probably be the canary in the coal mine, making it impossible to ignore the impact of cognitive automation. People will realise that machines have indeed become highly competent, and they will realise that their own job may also be vulnerable.

canary_art22

If we have a Franklin D. Roosevelt in charge at the time – perhaps one in every country – this may not be a problem. If there is a plausible plan for how to navigate the economic singularity, and a safe pair of hands to implement the plan, then we may be OK.

Unfortunately we do not currently have a plan. There is no consensus about what kind of economy could cope with a majority of the population being permanently unemployed, nor how to get from here to there. Neither are all the top jobs in safe pairs of hands.

In the absence of a solid plan explained by a reassuringly competent leadership, the reaction of large numbers of people realising that their livelihoods are in jeopardy is not hard to predict: there will be panic.

When will this panic occur? Within a few years of self-driving vehicles starting to lay off human drivers. In other words, in a decade or so.

The election of Trump and the Brexit referendum result were political earthquakes. Politics has not been so “interesting” since at least the fall of the Berlin Wall and the end of the Cold War at the end of the 1980s. But compared with what could happen if a majority of the population believes that they are about to become unemployable, they were relatively minor. The possible impacts of a panic about impending widespread joblessness could be enormous, and they are worth expending considerable effort to avoid.

Don't panic with Marvin

Reviewing last year’s AI-related forecasts

Robodamus 3

This time last year I made some forecasts about how AI would change, and how it would change us. It’s time to look back and see how those forecasts for 2017 panned out.

A bit rubbish, to be honest – five out of 12 by my reckoning.  Must do better.

  1. Machines will equal or surpass human performance in more cognitive and motor skills. For instance, speech recognition in noisy environments, and aspects of NLP – Natural Language Processing. Google subsidiary DeepMind will be involved in several of the breakthroughs.

A machine called Libratus beat some of the best human players of poker, but speech recognition in noisy environments is not yet at human standards. DeepMind made several breakthroughs. I’ll award myself a half-point.

  1. Unsupervised neural networks will be the source of some of the most impressive results.

Yup.

  1. In silico models of the brains of some very small animals will be demonstrated. Some prominent AI researchers will predict the arrival of strong AI – Artificial General Intelligence, or AGI – in just a few decades.

Not as far as I’m aware. No points.

  1. Speech will become an increasingly common way for humans to interact with computers. Amazon’s early lead with Alexa will be fiercely challenged by Google, Microsoft, Facebook and Apple.

Yup: Alexa is very popular, and the competition is indeed heating up.

  1. Some impressive case studies of AI systems saving significant costs and raising revenues will cause CEOs to “get” AI, and start demanding that their businesses use it. Companies will start to appoint CAIOs – Chief AI Officers.

There have been fewer case studies than I expected, but they do exist, and it is a rare CEO who is not building capability in AI. CAIOs are not yet common, but Dubai has a minister for AI, and the UK’s All-Party Parliamentary Group on AI has called for the UK to have one. (Disclosure: I’m an adviser to the APPG AI, and I support that call.) Half a point.

  1. Self-driving vehicles (Autos) will continue to demonstrate that they are ready for prime time. They will operate successfully in a wide range of weather conditions. Countries will start to jockey for the privilege of being the first jurisdiction to permit fully autonomous vehicles throughout their territory. There will be some accidents, and controversy over their causes.

Investment in Autos is galloping ahead, and they are clocking up huge numbers of safe miles, and generating huge amounts of data. States and countries are competing to declare themselves Auto-friendly.

  1. Some multi-national organisations will replace their translators with AIs.

Not as far as I’m aware. No points.

  1. Some economists will cling to the Reverse Luddite Fallacy, continuing to deny that cognitive automation will displace humans from employment. Others will demand that governments implement drastic changes in the education system so that people can be re-trained when they lose their jobs. But more and more people will come to accept that many if not most people are going to be unemployed and unemployable within a generation or so.

The Reverse Luddite Fallacy is proving tenacious. If anything, there seems to be a backlash against acceptance that widespread mass unemployment is a possibility that must be addressed. No points.

  1. As a result, the debate about Universal Basic Income – UBI – will become more realistic, as people realise that subsistence incomes will not suffice. Think tanks will be established to study the problem and suggest solutions.

Nope. No points.

  1. Machine language will greatly reduce the incidence of fake news.

Sadly not yet. No points.

  1. There will be further security scares about the Internet of Things, and some proposed consumer applications will be scaled back. But careful attention to security issues will enable successful IoT implementations in high-value infrastructural contexts like railways and large chemical processing plants. The term “fourth industrial revolution” will continue to be applied – unhelpfully – to the IoT.

There was less news about the IoT this year than I expected. It was all blockchain instead, thanks to the Bitcoin bubble. But there was plenty of fourth industrial revolution nonsense.

  1. 2016 was supposed to be the year when VR finally came of age. It wasn’t, partly because the killer app is games, and hardcore gamers like to spend hours on a session, and the best VR gear is too heavy for that. Going out on a limb, that problem won’t be solved in 2017.

Yup

Putting your money where your mouth is

Long Bet image

Robert Atkinson and I have made the 749th Long Bet shown above (and online here). Robert is president and founder of the Information Technology and Innovation Foundation, a Washington-based think tank.

Robert’s claim

With the rise of AI and robotics many now claim that these technologies will improve exponentially and in so doing destroy tens of millions of jobs, leading to mass unemployment and the need for Universal Basic Income. I argue that these technologies are no different than past technology waves and to the extent they boost productivity that will create offsetting spending and investment, leading to offsetting job creation, with no appreciable increase in joblessness.

My response

AI and robotics are different to past technology waves. Past rounds of automation have mostly been mechanisation; now we will see cognitive automation. Machines can already drive cars better than humans, and their story is just beginning: they will increasingly do many of the tasks we do in our jobs cheaper, better and faster than we can. Unlike us, they are improving at an exponential rate, so that in ten years they will be 128 times more powerful, in 20 years 8,000 times, and in 30 years (if the exponential growth holds that long) a million times. We are unlikely to see the full impact of technological unemployment by 2035, but it should be appreciable. Our job now, of course, is to make sure that an economy which is post-jobs for many or most people is a great economy, and that everyone thrives. The way to do that may well be the Star Trek economy.

I would like to be able to credit the person who created the excellent image below. If you know who it is (or if it is you!) please do let me know.

Horses and tech unemp

Our wonderful future needs you!

The Future needs you, Uncle Sam

The media today is full of stories about artificial intelligence, and there is universal agreement that it is a very big deal. But ironically, most people are not paying close attention. This is probably because the stories are confused and confusing. Some say that robots will take all our jobs and then turn into murderous Terminators. Others say that is all hype, and there is much less going on than meets the eye.

And so most people shudder slightly, shrug their shoulders and get on with the business of living. And who can blame them?

When you pull back from the headlines, much of the difference between the two camps is about timing. No, robots will not take all our jobs by 2019, but can we be so sanguine about 2039? And while few AI researchers agree with Ray Kurzweil’s confident prediction that we will create the first artificial general intelligence (an AI with all the cognitive abilities of an adult human) by 2029, surveys indicate that most of them think it likely to happen this century.

There are many reasons to be excited about AI and what it will do for us and to us. It is already making the world more intelligible, and making our products and services more capable and more efficient. This progress will continue – at an exponential rate; if we are smart and perhaps a bit lucky, we can make our world a truly wonderful place.

There are also many reasons to be concerned about AI. People worry about privacy, transparency, security, bias, inequality, isolation, killer robots, oligopoly and algocracy. These are all important issues, but none of them is likely to throw our civilisations into reverse gear, or even destroy us completely. There are two issues which could do precisely that: the technological and the economic singularities.

2 sing

The technological singularity is the moment when (and if) we create an artificial general intelligence which continues to improve its cognitive performance and becomes a superintelligence. If we succeed in ensuring that the first superintelligence really, really likes humanity – and understands us better than we understand ourselves – then the future of humanity is glorious almost beyond imagination. The solutions to all our major problems should be within our grasp, including poverty, illness, war and even death. If we don’t manage that … well, the outcome could be a lot less cheerful. Ensuring that we do manage it is probably the single most important task facing us this century – and perhaps ever, along with not blowing ourselves up with nuclear weapons, or unleashing a pathogen which kills everyone.

Before we reach the technological singularity we will probably experience the economic singularity – the point when we have to accept that most people can no longer get jobs, and we need a new type of economy. The stakes here are not so high. If we mis-manage the transition, it is unlikely that every human will die. (Not impossible, though, as in the turmoil, someone might initiate a catastrophic nuclear war.) Civilisation would presumably regress, perhaps drastically, but our species would survive to try again. Trying again is something we are good at.

On the other hand, assuming it is coming at all, the economic singularity is coming sooner than the technological singularity. The technological singularity is more important but less urgent, while the economic singularity is less important but more urgent.

The economic singularity is not here yet. The impact of cognitive automation is being felt in modest ways here and there, but the US, the UK, and many other leading economies are close to full employment because there are still plenty of jobs that humans can do. (Some of it doesn’t pay very well, but there are jobs.) This will not last.

Self-driving cars will be ready for prime time in five years or so. When they arrive, inexorable economic logic dictates that professional drivers will start to be laid off rather quickly. At the same time, most other sectors of the economy will be seeing the effects of advanced AI. The outcome can be wonderful – a world where machines do the boring jobs could be one where humans get on the important parts of life: exploring, learning, playing, socialising, having fun. But it is not obvious how to get from here to there: we need a plan, and we need to communicate that plan to avoid a dangerous panic.

It will probably take at least five years to develop that plan and generate a consensus around it. So we have to start now. We need to set up think tanks and research institutes all over the world, properly funded and staffed full-time by smart people with diverse backgrounds and diverse intellectual training. In the context of the importance of the challenge, the resources required are trivial – probably a few tens of millions of dollars – but they are sufficient to require significant political support.

At the moment, our politicians and policy makers are distracted. The US is understandably mesmerised by the antics of the 45th President, and in the UK, Brexit has swallowed the political class whole. Other countries have their own distractions, and the pain of the recession which started in 2008 endures. Artificial intelligence is poised to create the biggest changes humanity has ever been through, and yet it hardly featured at all in recent elections.

But the race is far from run. Politicians do respond to the public mood. (The most talented ones anticipate it slightly, although they are careful not to get too far ahead of us, or we sack them.) If we demand they pay attention to the coming impact of AI, they will. It is time to make that demand, and you can help. Talk to your friends and colleagues about this: get the conversation going. Insist that your political representatives pay attention.

A wonderful world can be ours if we rise to the challenges posed by the exponential growth of our most powerful technology, and navigate the two singularities successfully. Let’s grasp that wonderful future!

ASI

Putting the AI in retail: How cognitive reasoning systems could make life easier for consumers

Another guest post by Matt Buskell, head of customer engagement at Rainbird.

There was a time when booking a holiday meant a single trip to the high street travel agent. Nowadays, the process of online research seems to take longer than the holiday itself. The difference, back then, was the travel agent – a human being who could look you up and down, talk to you about your preferences, and make a recommendation based on their judgement.

travel-agent-2
In the world of AI, we like to call this ‘inference’. Travel agents never asked any questions like the filters and features you find on travel websites today – location, price range, number of stars. Nothing. Instead, they inferred what we would like, basing their judgement on factors such as how we were dressed, what we liked to do, and how we spoke to them.

Where does the time go?

The average time spent researching holidays in 1997 was just 45 minutes. Now, it’s over eight hours.

frustrated-600x423-335x236
The pattern is the same with other retail sectors that have moved online – from books, to groceries, music, clothing, and even cars. Hours are whittled away on websites like ASOS, TripAdvisor and Amazon. Imagine walking into a real-life store, asking the assistant for advice, and being handed a mountain of products and reviews to spend the next few hours scouring through. You’d probably just walk straight back out. So why do we settle for it online?

Convenience is one thing: for most of us, the ability to browse during the morning commute or on a lunch break is more appealing than a trip to the high street.

Many of us have also convinced ourselves that spending time looking at different online retailers and social media sites is the best way to ensure we have all the facts we could possibly need to get a ‘good deal’.

Online research
But when the choice of online stores and the availability of information was limited, it was a much simpler task. Now, we’re faced with an overload of choice, and the process of doing thorough research can feel laborious.

So why is it that targeted personal advice is lacking in online stores, whilst it is universally expected in the physical stores of our best retailers?

There are three main problems with online retailers today that limit their ability to provide the most suitable recommendations for individual customers:

1) You search for a product using narrow features, e.g. price, size, or category.

2) The system does not explain or offer a rationale for any recommendations it makes.

3) It’s a one-way interaction. You click, the computer displays.

Back to the future

The good news is that ‘conversational commerce’ and cognitive reasoning are going to bring the human element to online retailers. Ironically, the latest trends in AI are actually sending us back in time to the days of personalised shopping.

Imagine an online store in which an artificially intelligent assistant has been trained by your best human retail assistant, your favorite DJ, an experienced travel agent, or a stylist. You ask for advice or a product recommendation, and the system conducts a conversation with you, just like a real-life shop assistant.

robot_travel_agent500px
Let’s take holiday booking as an example. The cognitive reasoning system, channeled via a chatbot – let’s call it Travel Bot – asks you a range of questions to gauge your priorities and their order of importance. During this interaction, you say that you like the beach, enjoy city breaks, hate long journeys, and indicate that price isn’t your deciding factor. Travel Bot recommends a five-star beach resort in Cannes. You baulk at the price and ask for an explanation, and Travel Bot explains that beach property is fifty percent more expensive than inland. You decide that the beach isn’t that important – to which the Travel Bot responds with an altered recommendation.

You end up with the perfect compromise – a reasonably priced hotel stay in the centre of Nice, ten minutes from the beach.

In this instance, Travel Bot mirrors a human travel agent. It makes inferences, explains its recommendations, and continuously alters its advice to cope with uncertain customer responses.

A computer cannot completely replace a good human adviser- yet. We are just too complex to model. But by bringing back the human element of customer service and combining it with the retail arena of the future, we can take a lot of the stress out of online shopping.

online-research-baby
Rainbird is a leading Cognitive Reasoning engine that is re-defining decision-making with AI in Financial Services and Insurance.