Winter 2019
Resistance is Futile: But is AI for or Against Us?
The race to apply machine learning[1] in nearly every area of life is well under way. We are at an exciting moment when data – the fuel for smart algorithms – combined with sensors, high speed communications, advanced computing and new problem solving tools and techniques, are creating the conditions for Artificial Intelligence (AI) to change the way we live, learn and work.
Those angling for Number 1 and 2 in the AI Superpower race are the United States and China. While the US currently has the edge, it is challenged every day by the Chinese government’s ambitious masterplan to become an AI superpower by 2030. The way the race continues – via investment, changes to school and university curriculums and the fostering of AI-friendly ecosystems, incubators and accelerators — will determine who will shape the next decades.
We must remain mindful of the implications of this race on humanity. All is not bleak, but neither is the future entirely bright.
Do we care who is an AI superpower?
Artificial Intelligence is what economists benignly term a ‘general purpose technology’. It is not an innovation at the scale of a typewriter (which changed how we perform a single task, a calculator (which eliminated the need for one kind of labor) or even the cotton gin (which disrupted an entire industry). It is a technology only comparable with the disruption brought by the steam engine, electricity, computers and the internet.
AI will continue “the great decoupling”[2] of productivity and wages that the information and computer technologies (ICT) revolution began. Productivity will continue to shoot upwards – and wages and jobs will continue to flatten or fall. AI will continue the economic stratification in developed countries like the United States where the economic gains of ICT largely accrue to the top 1%[3].
The ICT and AI revolutions differ from the two previous general purpose technological revolutions in that they favor highly skilled workers. They empower the world’s top knowledge workers and undercut the economic role of many in the middle. As we transition into a world in which the online and offline fuse, who creates the AI we use will increasingly influence how we live our lives.
AI will create some jobs
It is true that AI is already creating jobs that did not exist before, much as every other innovation has. AI-enabled jobs that already exist include Cyber Calamity Forecaster (someone who monitors, detects and forecasts cyber threats, and predicts their impact), Augmented Reality Journey Builder (one who collaborates with engineers and artists to create an augmented reality experience) and Fitness Commitment Counselor (a coach or fitness instructor who uses smart garments, smart machines or fitness tracking). Teenagers can also look forward to future job titles which include Algorithm Bias Auditor (one who can review systems and software for fairness, legality and values) and Smart Home Design Manager (a person who can work with architects, engineers and technology in the home), Virtual Identity Defender (someone who can identify deep fakes and authenticate claims made online) or Virtual Reality Arcade Manager (a person to run an immersive, high definition, massive online multiplayer gaming house)[4].
These teenagers can also look forward to a world in which they may earn a living through six jobs instead of one. This may be by choice as the mobile, connected world allows them to use the gig economy to their advantage. Luke Tang of AI incubator Techcode describes this as freedom – as it offers an ability to use all of your skills and interests – not just one[5]. But these teenagers may not be juggling jobs by choice, it may be that one job simply doesn’t provide enough to earn a living.
Source: Rocketeer Piste of Government
AI will likely destroy more jobs than it creates
It is also true that AI will destroy jobs. The threat to truck drivers is well known. And it isn’t too far a stretch to see that fruit harvesters, restaurant cooks, cashiers and garment factory workers also risk losing their jobs to AI. Less appreciated is the scale of the threat – and the speed with which AI will subtract jobs from white collar jobs. Customer services reps, loan underwriters, lawyers, fund managers, telemarketers, radiologists, journalists[6], translators, insurance adjusters are a few of those jobs at risk. While experts vary from the wildly optimistic (only 9% of jobs will be lost according to the OECD) to the dire (47% of jobs lost to machines said Oxford researchers) – the PWC study that lands around the fact that 38% of jobs will be lost to automation by 2030 still appears the most robust.
And automation is already eliminating jobs. Andrew Yang, underdog in the race to become the Democrats’ choice to take on President Trump in November 2020 continues to campaign on a simple pitch. Automation is destroying American jobs, and we need a president with some kind of answer. (His answer: Universal Basic Income. He calls it a ‘freedom dividend’ because it polled better with Americans. Namely a $1,000 monthly check to every person in America, paid for with a Nordic-style Value Added Tax.)
So if we want a hand in this world in which AI will play a large part in how we’re employed, then yes we do care who creates it.
AI will make life better
Artificial intelligence also has the potential to help tackle some of the world’s most challenging social problems. The McKinsey Global Institute analysed potential applications for social good — from diagnosing cancer to helping blind people navigate their surroundings to identifying victims of online sexual exploitation, and aiding disaster-relief efforts. They suggest that AI could contribute to tackling cases across all 17 of the Sustainable Development Goals, potentially helping hundreds of millions of people in both advanced and emerging countries[7].
As Erik Brynjolfsson, director of the MIT Initiative on the Digital Economy has said, it is “more likely than not that we will use this power to make the world a better place. For instance, we can virtually eliminate global poverty, massively reduce disease and provide better education to almost everyone on the planet.”
What fuels advancement in AI?
Data availability and data analysis. The most valuable AI companies in the world in computer vision, speech recognition, speech synthesis, machine translation and drones are all Chinese companies. A principle reason for this is the vast availability of data. China has three to four more users than the US. The restrictions on what can be collected, held and analysed are lower. There are 50 times more mobile payments and 10 times more food deliveries in China than there are in the US. The three hundred times more shared bicycle rides have a number of sensors submitting all kinds of information into the cloud.
Overall, there’s about ten times more data in China than the US. Data is the power source fueling the AI grid. The more data, the better the AI works, regardless of how brilliant the researcher analysing that data is.
A question heard multiple times this year at European and American Smart Cities conference is: ‘who owns the data collected by cars and buses?’ This question is already posed differently in China, namely: ‘who should share in that data collection?’ Indeed, the state is enabling the sensors to be installed throughout cities to enable the development of a native autonomous vehicle industry. This is happening in the west as well, but at a slower and uneven pace. Monopolies
One well-acknowledged aspect of the Information and Computer age is the idea that power tends to end up in the hands of a very few companies. As soon as the invention of the web spread from CERN further afield the concern about monopolies dominating the internet developed. When the internet got going, everyone used the Netscape browser, thus Netscape controlled the web. Concern has shifted from monopolistic control over browsers and operating systems (Microsoft) to search engines (Google) to social networks (Facebook) to retail (Amazon). As Sir Tim Berners-Lee said recently, ‘it has been clear from very early on that the web is dominated by monopolies. But which monopoly dominates has shifted, and it may shift again.’[8]
In China, the vast majority of online activity also happens in a handful of companies — Baidu, Alibaba and Tencent.
AI as a technology and an industry also gravitates towards monopolies. Because AI products hoover up data to improve, they create a feedback loop of better products leading to more users. Those users produce more data which leads to better products. This repetitive cycle can turn a company’s early lead into an insurmountable barrier for entry to other companies.
Where does all this leave us?
PwC estimates that AI will add $15.7 trillion to the global economy by 2030. China and the US will take $10.7 trillion of it – accounting for almost 70% of the global economic impact[9]. All of the seven AI giants and an overwhelming portion of the best AI engineers are concentrated in the US and China. There is significant investment in Europe (and acknowledged accrued benefits) but there is a risk for lower income countries who have yet to bring the majority of their populations online. When low-wage factory labor is no longer a competitive edge to kick-start development, and when manufacturing and services can be done by intelligent machines located in AI superpowers, the risk is that lower income countries will find themselves with large numbers of displaced workers.
All is not lost
Prospects may – at present — appear challenging, but we are at least aware of the scale of the challenge, and we have the experience of having faced the aftermath and job losses of previous industrial revolutions.
Countries moving up the income ladder – notably in Africa – are building Artificial Intelligence into their development plans. Rwanda, Nigeria, Sierra Leone and Kenya have attracted some remarkable talent to return home to do so, are investing in skill-based education, and have already produced AI-enabled products (sensors and drones) that solve problems not only for Africans but for people all over the world[10].
We may well need to rethink the entire concept of work in a world where AI does the heavier lifting. Andrew Yang’s freedom dividend / Universal Basic Income is the idea already being tested in Alaska and Finland. Universal Basic Services is a related idea advanced by many, including John McDonnell of the UK’s Labour Party — where everyone is guaranteed a minimum internet contract, amount of mobile data, energy, shelter etc. Others including AI investor Kai-Fu Lee advocate a social investment stipend – a salary for those who invest their time in socially beneficial activities (care work, community service and education) that will always require human warmth.
Much has been written about the human attributes AI will not replace — namely empathy, love and compassion. Just as the future belongs to those who can understand and harness data, so too does it belong to those who can use those very human skills. We are not passive spectators in the race to implement AI – we are the participants in it.
Notes
1. Artificial Intelligence is the broader concept of machines being able to carry out tasks in a way that we would consider “smart”. Machine Learning is a current application of AI based around the idea that we should really just be able to give machines access to data and let them learn for themselves.
2. Recommended further Christmas reading on this subject: The Second Machine Age by Erik Brynjolfsson and Andrew McAfee
3. This elite group in the US has roughly doubled its share of national income between 1980 and 2016. By 2017 the top 1% of Americans possessed almost twice as much wealth as the bottom 90% combined. As ICT has proliferated across the economy, real wages for the median of Americans have remained flat for over 30 years, and they’ve actually fallen for the poorest Americans. Yet more recommended Christmas reading: AI Superpowers: China, Silicon Valley and the New World Order by Kai-Fu Lee.
4. Cognizant Jobs of the Future Index
5. Your correspondent for this article has three jobs by choice and can attest that it does provide freedom, variety, income and indeed uses all of her skills. It is also challenging.
6. See http://unitedrobots.ai/ and https://www.innovators-summit.com/news/detail/article/how-automated-content-is-revolutionising-swedish-publishing-explains-robin-govik-and-soeren-karlsson/
7. Notes from the AI Frontier. Applying AI for Social Good, McKinsey Global Institute
8. The Richard Dimbleby Lecture, Sir Tim Berners-Lee: The World Wide Web – A Mid-Course Correction
10. https://globalgoalscast.org/artificial-intelligence
Edie Lush has been a political analyst for UBS and is an author, journalist and communication trainer. She is Executive Editor of Hub Culture, Co-Host of the Global GoalsCast podcast and is a regular commentator and broadcaster on technology issues.