OECD Digital Economy Outlook 2024 (Volume 1): Embracing the Technology Frontier

Table of Contents

• Foreword 5
• Table of Contents 7
• Reader’s guide 9
• Acronyms 9
• Country groupings 11
• Abbreviations 12
• Executive Summary 13
• Key findings 13
• The ICT sector continues to outperform the total economy 13
• AI actors must collaborate to unleash innovation responsibly and ensure its benefits are widely shared 13
• Next generation wireless networks are key to unlimited connectivity everywhere 13
• Digital technology diffusion and the skills to use them effectively are uneven, limiting equal opportunity and inclusion 14
• Immersive technologies like VR enable extraordinary experiences, but safety on and off line is key 14
• Negative behaviours in digital environments are rising and disproportionally affect girls 14
• Chapter 1. The growth outlook of the ICT sector 17
• Key findings 19
• An overview of the efforts to measure the digital transformation 20
• DSUTs will improve measurement in the medium term 20
• Box 1.1. The SNA and the digital economy 20
• Data on the uptake of digital technologies provides complementary information 20
• Other approaches to measuring digitalisation of the economy 21
• Despite these efforts, there is a lack of timely monetary measures of the digitalisation of the economy comparable across countries 21
• Methodology for measuring the growth of the ICT sector in real time 22
• A nowcasting approach that is parsimonious with data 22
• Table 1.1. An indicative matching of Google Trends categories related to the ICT sector 22
• Extracting useful information from Google Trends data 23
• Figure 1.1. Sampling noise distribution before and after correction 23
• Figure 1.2. Raw and cleaned SVI time series for ten Google Trends categories in the Netherlands, 2004-23 24
• A neural network to measure ICT sector growth in real time 25
• An agnostic modelling approach 25
• Evaluating nowcasting performance 25
• Figure 1.3. The nowcasting model performs strongly 26
• The growth outlook for the ICT sector across countries 26
• Timely estimates of ICT sector growth 26
• Figure 1.4. The ICT sector is resilient in the face of economic headwinds 27
• Figure 1.5. The ICT sector shows remarkable dynamism 32
• In the past decade, the ICT sector grew in most OECD countries, but not equally 32
• Figure 1.6. ICT growth rates vary markedly across countries 32
• ICT sector growth rates in OECD countries are nevertheless converging 33
• Figure 1.7. Six percentage points separate the top and bottom ICT sector performers 33
• Figure 1.8. ICT sector growth rates are converging across countries 33
• The ICT sector performed strongly in all OECD countries in 2023 34
• Figure 1.9. ICT sector growth is strong across countries 34
• Measuring the ICT sector is key to evaluating its performance and designing sound policies 34
• References 35
• Notes 36
• Chapter 2. The future of artificial intelligence 39
• Key findings 41
• Figure 2.1. Examples of interrelated factors that will likely shape AI governance in future decades 42
• The AI technological landscape today and tomorrow 42
• Advances in neural networks and deep learning are resulting in larger, more advanced and more compute-intensive AI models and systems 42
• Box 2.1. “A” is for artificial intelligence 43
• Box 2.2. Increasingly general AI systems 44
• Foundation models are enabling increasing AI generality across application domains, industries and tasks 44
• AI development and use is expected to continue to depend on access to computing infrastructure 45
• The availability of vast amounts of data for training AI has greatly enhanced systems’ capabilities, including their ability to generate realistic content 46
• Figure 2.2. More than half of open-source AI training datasets are in English 46
• Privacy-enhancing technologies are emerging, including confidential computing methods and federated learning 46
• Generative AI could lead to more representative datasets but also raises concerns about manipulation 47
• Open-source resources can make progress more broadly accessible but introduce other challenges 47
• Computer vision capabilities continue to develop, but applications like facial recognition raise concerns 48
• Robots are getting better and smarter thanks to AI 48
• Will scaling-up current AI models continue to drive advancements in AI capabilities? 49
• Experts offer varying predictions for the future trajectories and implications of AI 49
• AI is expected to yield significant future benefits 49
• Potential benefits of AI come with risks and uncertain future trajectories 50
• Figure 2.3. Generative AI-related incidents and hazards reported by reputable news outlets have increased steeply since 2022 50
• Many solutions are being proposed to help yield AI’s benefits and mitigate its challenges 52
• Demystifying debates on maintaining human control of AI systems and on the alignment of AI systems and human values 52
• Policy considerations for a trustworthy AI future 53
• Where are countries in the race to implement trustworthy AI? 53
• AI research and development priorities feature prominently on national policy agendas, with investments poised to continue in the years ahead 53
• Figure 2.4. China, the European Union and the United States lead in the number of AI research publications, with India recently making strides 53
• Figure 2.5. China’s share of “high-impact” AI publications has steadily risen since 2000, notably overtaking the United States and European Union in 2019 54
• While global venture capital investments in AI and overall have declined since 2021, investments in generative AI start-ups have boomed 55
• Figure 2.6. VC investments in generative AI start-ups have boomed since 2022, while VC investments overall and in AI start-ups reached a peak in 2021 56
• Governments are building human capacity for AI as demand for AI skills grows 56
• With industry outpacing academia in developing advanced AI, countries increasingly see AI compute capacity as a crucial resource to be managed 56
• Figure 2.7. Advanced economies are competing for AI talent 57
• Countries are embedding values-based principles into AI legislation, regulation and standards, moving towards future-fit policies for trustworthy AI 58
• Technical standards will play a key role in the implementation of trustworthy AI 59
• Initiatives supporting international co-operation for trustworthy AI continue to grow 59
• Challenges lie ahead in crafting future-fit AI policies that also spur innovation 60
• The future of AI is uncertain and complex, posing great opportunities but also risks for society and economies 61
• References 62
• Notes 68
• Spotlight. Next generation wireless networks and the connectivity ecosystem 71
• Terrestrial connectivity: Beyond 5G technologies 73
• Figure 1.S.1. The road to 6G: Potential timeline 74
• Figure 1.S.2. Interrelated factors that will likely shape 6G during the next decade 75
• Non-terrestrial connectivity: Developments in satellites and other non-terrestrial wireless technologies 77
• High-throughput geostationary satellites and non-geostationary orbit satellite constellations 77
• High-altitude platform stations 79
• Hybrid topologies for connectivity: Towards the integration of terrestrial and non-terrestrial network technologies 79
• Figure 1.S.3. Towards vertical coverage: Convergence of terrestrial and non-terrestrial networks 80
• Box 1.S.1. Flying taxis with electric vertical take-off and landing aircrafts for future transportation 80
• References 82
• Notes 85
• Chapter 3. Digital technology diffusion and data 89
• Key findings 91
• The ability to use the Internet effectively is key for equal opportunity and inclusion 91
• The incidence and frequency of Internet use have increased, but gaps remain 92
• Figure 3.1. Internet adoption has increased 92
• Figure 3.2. Divides in Internet use are pronounced by age, education and income 93
• Figure 3.3. Gaps in Internet use are narrowing but remain pronounced among the elderly 94
• Online activities differ in the extent to which they require education and ICT skills 94
• Figure 3.4. Uptake of Internet banking and online government services varies across countries 95
• Figure 3.5. Online activities such as Internet banking and online purchases are correlated with formal education and ICT skills 95
• Younger and more educated Internet users engage in a larger variety of online activities 96
• Figure 3.6. Younger and more educated Internet users engage in a larger variety of online activities 96
• COVID-19 led people to rely more on online services... at least temporarily 97
• Figure 3.7. Uptake of online services increased during the pandemic 98
• While more telework seems to be here to stay... 98
• ...the large uptick in e-commerce seems to be fading 99
• Figure 3.8. Much of the initial increase in e-commerce during COVID-19 has dissipated 99
• During the pandemic, those with the requisite skills were in a better position to use online services to their advantage 100
• Figure 3.9. COVID-19 was often associated with slowing convergence in uptake of online services 100
• Data-dependent technologies are diffusing at a slow pace 101
• Uneven diffusion of data-dependent digital technologies may undermine productivity growth 101
• Uptake of data-dependent technologies such as big data analytics and AI remains low 101
• Figure 3.10. Adoption of data-driven technologies remains low 102
• Cloud computing has been diffusing three times more rapidly than big data analytics 103
• Figure 3.11. Cloud computing has been diffusing three times more rapidly than big data analytics 103
• AI adoption is concentrated in the ICT sector 104
• Figure 3.12. Adoption of cloud computing and IoT technologies is distributed evenly across sectors 104
• Firm size is more important for adoption of data-dependent technologies and software than for IoT or cloud computing 104
• Slower diffusion of data-dependent technologies might be linked to scale economies, financial frictions or lack of access to data 105
• Figure 3.13. Firm size is a more important predictor of adoption for data-dependent technologies and software than for IoT technologies or cloud computing 105
• Far-ranging adoption of AI might require yet more experimentation and co-invention 106
• Boosting equitable uptake and diffusion of digital technologies are vital to bridging digital divides and fostering productivity growth 107
• Annex 3.A. Regression tables 108
• Annex Table 3.A.1. Effect of education, ICT skills and income on uptake of online services 108
• Annex Table 3.A.2. Effect of COVID-19 on uptake of online services 108
• Annex Table 3.A.3. Effect of COVID-19 on uptake of online services by level of education attainment 109
• Annex Table 3.A.4. Effect of the semiconductor shortage on M2M SIM cards 109
• Annex Table 3.A.5. Diffusion of big data analytics and cloud computing 109
• References 110
• Notes 114
• Chapter 4. Virtual reality and its opportunities and risks 119
• Key findings 121
• Understanding VR 122
• A continuous cycle of tracking, rendering and display enable VR 122
• The technological features of VR 123
• Touch and smell are complex to render in VR 124
• The benefits and opportunities of VR 125
• Teaching empathy 125
• Medical rehabilitation 126
• Mental health 126
• Training 127
• VR digital twins 128
• The downsides and risks of VR 128
• Privacy risks 128
• Box 4.1. Tracking data: Actions speak louder than words 129
• Simulator sickness 130
• Cognitive development and behaviour of children 130
• Distracted driving 131
• Overuse and addiction 131
• Governing VR 132
• Policy action on VR and immersive technologies is focused on promoting the domestic VR industry 132
• Towards the development of “rights” or “principles” for VR and immersive technologies 133
• Data from VR applications bring new challenges to existing privacy frameworks 133
• VR mental and physical safety must be carefully considered 134
• VR businesses have been self-regulating 135
• References 136
• Notes 141
• Spotlight. Mental health and digital environments 143
• Anonymity, disembodiment and disinhibition help explain why people communicate and interact differently on line 145
• Cyberbullying, PIU and PSMU are associated with mental health problems 146
• Immersive technologies bring new opportunities for mental health but can also exacerbate the risks 147
• Evidence suggests that negative behaviours in digital environments are on the rise and they disproportionally affect girls 147
• Cyberbullying is becoming more prevalent across countries, with girls experiencing higher rates than boys 148
• Figure 2.S.1. Cyberbullying rates have increased in nearly all countries 148
• Girls are more likely than boys to be problematic users of social media and the gap is widening 149
• Figure 2.S.2. Girls are problematic social media users more often than boys 149
• Moderate use of digital technologies tends to be beneficial, but “overuse” may be detrimental 149
• Box 2.S.1. Is time spent on line associated with problematic behaviour? 150
• Figure 2.S.3. Evidence suggests that intensive use of online communication is associated with PSMU 150
• Towards a policy agenda for fostering mental health in the digital age 151
• Raise awareness about negative behaviours in digital environments and promote media literacy 151
• Promote safety by design 151
• Identify specificities of immersive environments that present risks for mental health 151
• Improve the evidence base on mental health and digital environments 151
• Engage in partnerships with a range of stakeholders to prevent and address negative behaviours in digital environments 152
• References 153
• Notes 157
• List of Figures 159
• List of Tables 160
• List of Boxes 160