This essay focuses on various economies of the digital divide, and the related consequences from a macroeconomic point of view. The main issue is to identify the best rules for global competitiveness across the global economy, to mitigate the risks and potential negative outcomes. The main reason for the development of AI is productivity.

_____________________________________________________________________

Article source: Human Decisions – Thoughts on AI. Published in 2018 by UNESCO and Netexplo© 2018 ISBN 978-92-3-100263-2.

This publication is available in Open Access under the Attribution-ShareAlike 3.0 IGO (CC-BY-SA 3.0 IGO) license . 

AI as a global game changer: three key challenges.

By Anne Miroux.

 

 (p.115-121)

In early 2017, PwC predicted that, by year 2030, artificial intelligence (AI) would boost global GDP by $16trn, and that half of this would go to one single national economy, namely China43.

Clearly, some major shifts have started to affect the world economy, which will modify the way we look at production, trade, employment and geo-politics altogether. The present paper will try to identify some of the directions that such changes could take, consider their possible implications regarding the performance of various types of economies, and anticipate some of the critical issues that should be considered most urgently if we want to maximize the global benefits of AI and keep its possible negative implications under control.

As the CEO of IBM, Ginni Rometty, recently reminded us, ‘the term “artificial intelligence” was coined in 1955 to convey the concept of general intelligence: the notion that all human cognition stems from one or more underlying algorithms, and that by programming computers to think in the same way, we could create autonomous systems modelled on the human brain.’44. In the report mentioned earlier, PwC used a slightly different definition, considering that ‘AI is a collective term for computer systems that can sense their environment, think, learn, and take action in response to what they’re sensing and their objectives’. Although such definitions vary, the general consensus would currently converge to defining AI as including automated intelligence (automation of manual/cognitive and routine/non routine tasks, typically found in equipment like robots), assisted intelligence (helping people to perform tasks faster and better, for instance through adaptive software, imbedded in equipment such as cars, or included in search engines), augmented intelligence (helping people to make better decisions, for instance through algorithms that will add contextual data to that already gathered by the user, compare complex situations to similar ones – stored in large memories, generally cloud-based, and turn this additional knowledge into proposals for action/decision), and finally autonomous intelligence (going all the way to automating decision making processes without human intervention, which is becoming more and more common in ‘smart management’ of airline and other transport traffic, or energy grids for example).

________________________________________

43 See PWC (2017).

44 See Rometty (2017). In the same article, she also mentioned that ‘At the same time, other researchers were taking a different approach. Their method-which worked bottom up to find patterns in growing volumes of data-was called IA, short for “intelligence augmentation.” Ironically, the methodology not modelled on the human brain has led to the systems we now describe as cognitive. IA is behind real-world applications such as language processing, machine learning and human-computer interaction. The term “AI” won out in the end, despite being a misnomer.’

Why and how is AI changing the game?

At the risk of oversimplifying a complex equation, one could consider that, although AI is not a new phenomenon, it has recently reached a different level of visibility and impact because of what happens around it, both upstream and downstream. Upstream, ‘big data’ has benefitted from the exponential growth of computing power, and the dramatic reduction of costs regarding the transmission and storage of information. Downstream, the advent and democratization of virtual and augmented reality (virtual reality – VR – and augmented reality – AR-) is de facto offering natural outlets to practically every innovation in AI.

But why are businesses and governments so eager to invest billions of dollars into AI, and why do markets value AI firms so much?45. The key word here is productivity. AI holds the promise not only of significant productivity gains in manufacturing (through a better allocation of resources between capital and labor) but also of true ‘quantum leaps’ in a wide range of industries and value chains that are likely to be transformed, displaced or replaced.

The reason why the anticipated impact of AI is so significant results from the fact that it allows three kinds of changes to happen simultaneously, namely:

  • Doing things faster: large amounts of data can be collected, analyzed and synthetized on a routine basis.

Big data applies massively to areas like law (millions of cases can be read and compared), scientific research (multiple configurations and analyses are performed through simulated lab experiments), consumer behavior (through the continuous gathering and analysis of buying patterns, and its application to customized/targeted advertising), filtering of job applications (analysis of keywords in applications and resumes allow algorithms do perform initial selections before any human eye has even seen the application).

  • Doing things better: greater precision and safety levels can be reached in operations such as self-driven cars (insurance companies have started offering discounts for equipped vehicles), and surveillance (drones, movements and patterns analyses, face recognition, including emotion identification) for instance. Risk evaluations (eg to grant loans) are already performed better by algorithms than by humans. 46
  • Doing things differently: AI is rapidly moving into creative areas. Deep thinking (i.e. the use of self-improving algorithms) has opened the door to adaptive robotics (i.e. machines endowed with the power to improve constantly through software ameliorations, and – progressively – reconfiguration. In this regard, advances provided by IBM (Deep Blue, Watson) and Google (AlphaGo) have pushed the frontier beyond what was considered possible a decade or two ago.

________________________________________________

45 From January 2009 to March 2017, the MSCI index (worldwide) appreciated by 40 %; the growth was 200 % for the S&P Global BMI (Information Technology), and 280% for the Robo Global Index (Robotics and Automation). See : Li, H. (2017).

46 Kai Fu Lee, a former executive at Microsoft and CEO of Google China, who has created his own venture capital company (Sinovation Ventures) predicts that ‘big banks will fall first to artificial intelligence’. See Kai Fu Lee (2017a).

The convergence of these three trends has massive implications for:

  • Productivity: as flagged earlier, labor productivity will soar because of modified capital/labor ratios; this will also be the case for total productivity, because of continuous (24/7) production, faster delivery, and improved quality control.
  • Security/safety: VR, AR will help a global relinquishing of repetitive and dangerous jobs.
  • Competitiveness: both enterprises and nations will benefit, especially if they are among first-comers.
  • Social organization: the future of work will be strongly influenced by the availability of life-long learning tools; individual strategies (how to position oneself not to be competing with robots/AI) will rapidly improve as a result.
  • New goods and services will be developed: self-learning robots, operator-coaching/teaching equipment, algorithm-based decision-making goods and services (self-driven cars, predictive software for retail, advertising, entertainment, e.g.) will drive significant segments of the global economy.

Altogether, massive changes will affect how value is distributed along production and delivery chains, locally and globally. Entire industries will be relocated, transformed or obliterated altogether while others will emerge as critical and strategic.

The current geo-economics of AI – a growing divide

The world is currently divided between a handful of major players and potential players on one hand, and a majority of countries with little or no ambition of capabilities in the field of AI. The threat of a ‘global AI divide’ is hence already a reality.

  • The incumbents: The United States, Germany, Japan and Korea. The United States is spearheading the development of autonomous vehicles, led by companies like Google, Tesla and Uber. In consumer markets, Google, Facebook, Microsoft, Amazon are making extensive use of AI and expanding operations to other countries. Korea is still a leader in many semi-conductors segments, and companies like Samsung or Hyundai are clearly benefitting from this situation in developing their own AI efforts. But this situation may be under threat from China, since the Chinese current five-year plan (2016 – 2020) has identified semi-conductors as a priority area. More tensions should be expected in this sector, as many AI experts predict that an increasing proportion of basic AI algorithms will be hardwired inside chips, rather than programmed ex-post.47 With leaders like Toyota, Japan has focused very much on robotics, and in particular ‘human looking robots’ able to replace employees in hotels and various other sectors. It appears to be a market with significant potential to rapidly adopt AI as a regular part of its consumer goods and domestic services. Germany has also been spearheading AI efforts in areas such as on-board systems for cars and machine-tools, with a high concentration of AI start-ups in the Berlin area. Those countries are expected to remain among the leaders (and main beneficiaries) of the anticipated growth of AI, at least in the coming decade.

_____________________________________________________

47 Illustrated for example by the recent release of Googles new ‘neural network’ chip (TPU 2.0), which will not be sold to other companies. See Metz (2016 and 2017).

 

  • Possible contenders: a handful of advanced economies should be able to build respectable market shares in AI goods and services. Such countries include France (where a company like Atos is granting priority to the development of quantum computing, which may prove critical to the acceleration of deep thinking and deep learning), Nordic countries and Switzerland (where relatively small companies – eg in Norway – and a dense tissue of top universities and research centers – eg around Switzerland’s EPFL ‘Brain Project’ – could combine into a dynamic AI ecosystem), and finally some visionary fast growing economies (such as that of the UAE) where significant funding could accelerate the adoption of ambitious AI-related strategies.
  • The upcoming giant: China. China is already a world leader in several key AI markets. For example, the Chinese speech-recognition company iFlytek and several Chinese face-recognition companies such as Megvii and SenseTime have the highest market capitalizations in their respective field. China is also moving aggressively in AI for consumer markets, through companies like Baidu, Alibaba and Tencent. Recently, Baidu (which can be regarded as the equivalent of Google in China) issued an academic paper proposing to use a combination of AI and Baidu Maps to predict when and where dangerous crowds are forming, and alert users in the area, as well as local authorities.48

__________________________________________________

48 China Daily (2016).

The foreseeable future: China casts a long shadow

In the race to develop AI goods and services, five key ingredients will matter, namely: market size, data volumes, technology (which is largely a function of R&D funding), talent and ambition.

In such a context, it is not difficult to see why China (and to some extent the United States) are enjoying both a headstart and an increasing advantage. Kai Fu Lee summarizes it as follows: ‘AI is an industry in which strength begets strength: The more data you have, the better your product; the better your product, the more data you can collect; the more data you can collect, the more talent you can attract; the more talent you can attract, the better your product. It’s a virtuous circle, and the United States and China have already amassed the talent, market share and data to set it in motion’.49

China’s efforts are helped by a massive funding effort, which is not limited to support from its central government agencies. As noted by J. Mozur and P. Markoff, ‘Quantifying China’s spending push is difficult, because authorities there disclose little. But experts say it looks to be considerable’. 50 It is also interesting to note that, more and more, local entities (such as municipalities) are taking a visible role in attracting talents51 and financing local innovative ventures, especially in the field of AI, spending billions on developing robotics for example. Cities like Xiangtan (Hunan province) has pledged $2 billion toward developing robots and artificial intelligence. In Suzhou (close to Shanghai), leading artificial intelligence companies can get about $800,000 in subsidies for setting up shop locally. In Shenzhen (a megacity close to Hong Kong, and host of Huawei’s headquarters) $1 million is offered to any AI project established there. It is also interesting to note that Chinese tech giants like Baidu, Tencent and Didi Chuxing have opened artificial intelligence labs in the United States, as have some Chinese startups. Over the past six years, Chinese investors helped finance 51 American artificial intelligence companies, contributing to the $700 million raised.52

_______________________________________________

49 Kai Fu Lee (2017b).

50 Mozur, P. and Markoff, J. (2017).

51 Policies designed and implemented by cities to grow, attract and retain talents are becoming critically important in shaping the geo-economy of work in fields like AI. See for example Lanvin, B. (2017).

Recent progress made by China in innovation has attracted international attention: in 2016, China moved into the top 25 of the Global Innovation Index, and in 2017, its ranking improved even further (it is now 22nd)53. Growing talent should reinforce this trend, especially in areas like AI. As The Economist recently noted, ‘As well as strong skills in maths, the country has a tradition in language and translation research, says Harry Shum, who leads Microsoft’s AI efforts. Finding top-notch AI experts is harder in China than in America, says Wanli Min, who oversees 150 data scientists at Alibaba. But this will change over the next couple of years, he predicts, because most big universities have launched AI programmes. According to some estimates, China has more than two-fifths of the world’s trained AI scientists.’54. Since China is now the largest market for data collection and handling, its clearly has a growing advantage in terms of ability to develop deep-learning, i.e. the process by which the quality of algorithms (i.e. the core of AI) improve through the continuous identification of patterns across huge amounts of data.

Conclusion – AI’s three key challenges

Experts have long been warning us about the opportunities and challenges raised by AI. In his 1999 book ‘The age of spiritual machines’, and even more so in ‘The singularity is near’ (2005)55, Ray Kurzweil spelt out the fundamentals of today’s debates. In the movie version of that book (released in 2012) 56, Kurzweil – playing his own role – discussed those with nineteen ‘big thinkers’ of the time. If that movie were produced today, the list would most probably include business leaders such as Bill Gates, Elon Musk, Jeff Bezos or Mark Zuckerberg. 57

There is no doubt that AI will continue to develop fast, attract significant investment and R&D spending, and Infrastructure and equipment, and fundamentally alter the current bases of production, trade and value creation.

_____________________________________________________

52 At his point in time, it is difficult to know what the future of such cooperation will be. The rapid development of China’s spending on AI is in stark contrast with the trend currently observed in the US. Mozur, P. and Markoff, J. (2017) also note that ‘President Trump’s proposed budget, meanwhile, would reduce the National Science Foundation’s spending on so-called intelligent systems by 10 percent, to about $175 million. Research and development in other areas would also be cut, though the proposed budget does call for more spending on defence research and some supercomputing. The cuts would essentially shift more research and development to private American companies like Google and Facebook.’

53 Global Innovation Index (2017).

54 The Economist (2017).

55 Kurzweil, R. (1999, 2005).

56 See http://www.kurzweilai.net/the-singularity-is-near-movie-available-today .

57 Stephen Hawking, Bill Gates and Elon Musk have rightly raised concerns about the risks inherent with AI capable of equaling, or even surpassing, human intelligence. Anticipating the emergence of even more powerful and increasingly autonomous AI reinforced by quantum computing, they have been are asking for a collective reflection upon what could constitute a challenge to mankind, a technology that could dominate its creator. Elon Musk recently launched his own ‘AI Initiative’. See https://openai.com/about/

In this process, obstacles may lay in AI’s path, and in that of its champions. International cooperation and policy attention will be required in at least three areas, namely:

  • The design, implementation and surveillance of ‘fair competition rules’, preventing the emergence of monopoly or dominance situations, which would create unsurmountable barriers to entry for those not already in the AI race.
  • The formulation of the innovative mechanisms of concertation, regulation and governance that the world of AI urgently requires. Such regulation will also be critically important for the governments and companies that are ahead in the field, as they could mitigate some of the unavoidable backlash that the looming of an AI-driven global economy will generate across public opinions because of its perceived and/or actual impact on civil liberties, individual freedom, privacy and employment for instance.
  • The immediate reform of education systems and life-long learning mechanisms to allow a constant re-skilling of the workforce at all levels (from vocational to global knowledge skills, to use GTCI terminology).

 

Anne Miroux. Lecturer, Cornell University. Former Director, United Nations Conference on Trade and Development (UNCTAD)

References

Chandran, N. (2016), ‘China’s tech ambitions puts South Korea on alert’, CNBC, 9 Mar 2016. See https://www.cnbc.com/2016/03/09/south-korea-tech-at-risk-as-china-steps-up-ambitions.

China Daily (2016), Baidu develops AI algorithm to predict crowds, avoid stampedes, chinadaily. com.cn, 25 Marc 2016 http://www.chinadaily.com.cn/china/2016-03/25/content_24093618.htm

Fay, P. (2017), La Bourse passe à l’ère de la robolution, Les Echos, 1er Juin 2017, https://www.lesechos.fr/finance-marches/marches-financiers/030361740719-la-bourse-passe-a-lere-de-la-robolution-2090888.php

Global Innovation Index (2017), Cornell-INSEAD-WIPO. https://global-indices.insead.edu/gii.

Global Talent Competitiveness Index (2017), Adecco-INSEAD-HCLI. https://global-indices.insead.edu/gtci.

Gossett, D. (2016), Artificial intelligence and geopolitics, China Daily European Weekly 09/16/2016, page 10. http://www.huffingtonpost.com/david-gosset/artificial-intelligence-a_2_b_10710612.html

Kai Fu Lee (2017a), Let Machines Invest your Money, interview to Quartz (16 Mai 2017). See https://qz.com/984709/kai-fu-lee-a-former-google-china-ceo-predicts-big-banks-will-fall-first-toartificial-intelligence/.

Kai Fu Lee (2017b), The real threat of artificial intelligence, New York Times, 24 June 2017. https://www.nytimes.com/2017/06/24/opinion/sunday/artificial-intelligence-economic-inequality.html

Kurzweil, R. (1999), The age of spiritual machines – when computers exceed human intelligence, Viking https://en.wikipedia.org/wiki/The_Age_of_Spiritual_Machines

Kurzweil, R. (2005), The singularity is near – When Humans Transcend Biology, Viking. https:// en.wikipedia.org/wiki/The_Singularity_Is_Near

Lanvin, B. (2017), Benchmarking Cities as Key Players on the Global Talent Scene (in Global Talent Competitiveness Index 2017), INSEAD, Adecco, HCLI 2017. http://global-indices.insead.edu/gtci .

Lanvin, B. and Evans, P. ed. (2017), Talent and Technology, (in Global Talent Competitiveness Index 2017), INSEAD, Adecco, HCLI 2017. http://global-indices.insead.edu/gtci .

Li, H. (2017), ‘Long-term Growth Investing – Robotics, Automation, Cyber Security’, ETF Securities, 23 March 2017. http://www.etfexpress.com/2017/02/27/249065/etf-securities-reports-growth-cyber-security-and-robotics-etps.

Metz, C. (2016), ‘How AI Is Shaking Up the Chip Market’, Wired, 28 October 2016. See https://www.wired.com/2016/10/ai-changing-market-computer-chips.

Metz, C. (2017), ‘Google Rattles the Tech World With a New AI Chip for All’, Wired, 24 April 2017. See https://www.wired.com/2017/04/race-make-ai-chips-everything-heating-fast.

Mims, C. (2017), A Hardware Upgrade for the Human Brain, Wall Street Journal, 5 June 2017. https://www.wsj.com/articles/a-hardware-update-for-the-human-brain-1496660400.

Mozur, P. and Markoff, J. (2017), China gains on America in artificial intelligence, New York Times, 29 May 2017. https://www.nytimes.com/2017/05/27/technology/china-us-ai-artificial-intelligence.html.

PWC (2017), Sizing the prize – PwC’s Global Artificial Intelligence Study: Exploiting the AI Revolution. http://www.pwc.com/ai.

Rometty, G. (2016), The natural side of AI, The Wall Street Journal, 18 October 2016, https://www.wsj.com/articles/the-natural-side-of-a-i-1476799723.

Rousseau, Y. (2017), Le japonais SoftBank s’offre les spectaculaires robots de Google, Les Echos, 12 Juin 2017. https://www.lesechos.fr/tech-medias/hightech/030375701278-apres-pepper-le-japonaissoftbank-soffre-les-robots-spectaculaires-de-google-2093143.php.

The Economist (2017), China may match or beat America in AI, 15 June 2017. https://www.economist.com/news/business/21725018-its-deep-pool-data-may-let-it-lead-artificial-intelligencechina-may-match-or-beat-america.