5. MARCO TEÓRICO
5.3 Escuchar con el corazón o de la educación sentipensante
The potential contribution to economic performance of new digital technologies can hardly be underestimated. Scholars and commentators are now convinced that the “Gordon paradox” (i.e. the fact that we see technology everywhere, but not in measures of productivity) is a thing of the past, and that new digital technologies will finally start massively contributing to productivity worldwide, including in the EU. Recent reports by Accenture/Frontier Economics, McKinsey and PWC conclude that technologies like AI will be game changers for total factor productivity and growth, by gradually rising as a third pillar of production, together with labour and capital. Chen et al. (2016) estimate the cumulative economic impact of AI from 2016 to 2026 as lying between $1.5 trillion and $3 trillion (0.15% to 0.3% of global GDP). Furman and Seamans (2019) review some of the most interesting literature on the impact of AI on the economy, which mostly finds that AI and robotics have the potential to increase productivity growth but may have mixed effects on labour, particularly in the short run. They also conclude that many economists believe that “AI and other forms of advanced automation, including robots and sensors, can be thought of as a general purpose technology (GPT) that enable lots of follow-on innovation that ultimately leads to productivity growth”; the fact that AI has not (yet) translated into large productivity gains,
164|LEADING THE WORLD IN TECHNOLOGY RULES AND STANDARDS
according to Brynjolfsson, Rock and Syverson (2017), is due to a “lag between technological progress and the commercialization of new innovative ideas building on this progress which often rely on complementary investments” (Furman and Seamans, 2019) – a lag, these authors claim, that is particularly notable in the case of GPTs. The prospect of unprecedented productivity gains, and the certainty that “winter is not coming” for AI (Renda, 2019), have led countries around the world to launch dedicated strategies for the new family of digital technologies, and in particular on AI. The US and China are currently trapped into a digital arms race and making dominance in AI and related technologies (notably, quantum computing) the top strategic priority for the next decade. This, in turn, had led investment to skyrocket. Especially in the fields of high-performance computing, 5G/IoT and artificial intelligence, Europe risks being left behind by the scale of investment (including in the defence sector); the less precautionary approach to regulation; and the weaker importance attached to protecting users and citizens in countries such as China.
For example, the EU currently provides about 5% of supercomputing resources worldwide but consumes one third of them, and member states have long been in dire need of expanding their supercomputer capacities. 119 A look at quantum computing
patents and investment is also revealing. As shown in the Figure below, the US and China are in a race to patent quantum-related technologies, although the EU is the region in the world with the highest number of research publications in this field. China is particularly interested in quantum cryptography, which has immense implications for military operations. In June 2019, in an attempt to catch up with these developments, the European High- Performance Computing Joint Undertaking (EuroHPC) selected eight sites for supercomputing centres located in eight different member states to host the new high-performance computing machines,120 for a total investment of €840 million. The Barcelona
supercomputer called MareNostrum 6, in particular, will be faster than the fastest US-based supercomputer (200 petaflops against 140), and may see the light by 2025.
HIDDEN TREASURES |165
Figure 23. Patents on quantum computing technology
In the coming years, the race for developing a full digital technology stack will become even more hectic, as testified by the recent skirmishes between the US and China over 5G connectivity and wireless equipment. The most intensely competitive battlefield is probably artificial intelligence, given its pervasive, dual-use nature and the relevance it will have for all sectors of the economy. In this space, the debate at EU level is increasingly loaded with hopelessness: Europe has lost the B2C race; Europe does not have any of the technology giants; the data train has left the station, etc. There is some element of truth in these statements, but also excessive pessimism, which may become an alibi for policymakers not to take action to reverse the trend.
More specifically, it is true that the B2C domain is currently dominated by large US tech giants such as Google, Facebook, Amazon, Apple, Twitter, Microsoft and Netflix, though this dominance is increasingly challenged by Chinese giants such as Alibaba, Baidu, Tencent and Huawei. The two superpowers hardly fight in each other’s backyard (at least when it comes to platforms); yet they compete at arms’ length in global markets. The US tends to dominate on the software and applications side, but Chinese companies dominate the infrastructure domain, and will increasingly need to serve non-domestic markets once its middle class has consolidated and related markets are saturated.
In all this, it is widely acknowledged that Europe will not be able to compete on an equal footing with the United States and
166|LEADING THE WORLD IN TECHNOLOGY RULES AND STANDARDS
China in terms of the sheer size of investment in new technologies. The Communication on Artificial Intelligence, adopted in April 2018, acknowledges the investment gap and highlights a possible alternative strategy for Europe, mostly based on a combination of competitiveness and ethical rules. In other, related fields such as 5G wireless communications, the internet of things, online platforms, high-performance computing and blockchain, the Commission has shown similar intentions, but so far relatively poor implementation. In yet another set of high-tech areas, such as genetics and genomics, the EU has remained almost silent despite the fact that the policy debate that emerged over the past few years was ethically loaded.
Could Europe play a leading role in the setting of rules and ethical principles for the development and commercialisation of new technologies? As things stand, the answer cannot be positive. If one looks at the overall technology stack that is emerging as the so-called Web 3.0, Europe appears to be lagging behind other regions of the world in many crucial respects: not only the size of investment in R&D&I, but also in terms of fixed and wireless broadband deployment (very high capacity networks, as well as 4G); level of per capita investment in e-communications infrastructure, relative presence in 5G standards; level of investment in AI, blockchain, the IoT; uptake of new technologies among consumers, as well as among firms; relative development of high-tech skills and competencies; ownership of patents in key enabling technologies; readiness for the quantum supremacy age; and even skills available in the public administration. All this weakens Europe’s potential authoritativeness when it comes to credibly proposing (let alone imposing) global standards.
In all this, the recent entry into force (in May 2018) of the General Data Protection Regulation is considered by many as a possible exception, which may chart a new course in EU technology policy. By requiring strict standards of data protection, as well as restrictions on profiling, a right to data portability and to receive a meaningful explanation of how algorithms reached sensitive decisions, the GDPR seeks to establish a global standard in the high- tech world, and to chart a new course in technology policy, making it more user-centric after many years of rather drastic laissez faire vis-à-vis data protection.
HIDDEN TREASURES |167
The GDPR has been implemented only recently, and significant uncertainty still exists as regards its success in terms of improved protection of end users’ right to data protection, as well as in terms of actual levels and modes of compliance. Accordingly, it is probably too soon to draw conclusions, but a few lessons can already be learnt. First, the GDPR has shown that courage pays at the EU level: the decision taken a few years ago on the need for a regulation on data protection has made the GDPR a path-breaking text in a world dominated by relatively lenient data protection rules. Second, Europe has successfully conquered the front pages of the international press, as well as grabbing the attention of top company CEOs through leveraging its ability to draft consistent, comprehensive rules, as well as its large and relatively rich internal market. Third, and relatedly, the EU has discovered that well- conceived, sound technology rules can potentially translate into effective export products, although there is no certainty that this is actually happening with the GDPR. Assuming that the GDPR will eventually be a success, can the EU replicate this experience in other, related fields? One attempt, as already mentioned, is underway in the field of artificial intelligence, and this link is so explicitly made inside the European Commission that the slogan “AI is the new GDPR” is becoming increasingly popular. At the same time, Europe has no consolidated tradition in the ethics of AI, which would help it build credibility in a field in which it has certainly neither research and innovation leadership (with some exceptions), nor a very established infrastructure.