This article builds upon the paper written by Fide after the 2019 Oxford Congress. Click here to consult it
The fourth industrial revolution -also called Industry 4.0- refers to the digital revolution that is characterized by a fusion of technologies that is blurring the lines between the physical, digital, and biological spheres. It relates to the phase in digitalization driven by disruption including human-machine interaction, connectivity, big data and IA. This wave of change began in the mid 2010s and posed several challenges and opportunities, including the positive impact of technological advancements on innovation, productivity, efficiency, impact on jobs and on the labour market or the challenge of how to strike the right balance in regulating emerging technologies. Companies are no longer experimenting with AI; it is now seen as a technology that can deliver tangible business value for customers. According to Accenture, by 2024, the percentage of organizations investing more than 30% of their tech budgets in AI will increase to 49% (Accenture). Researchers and experts are now referring to the new term of the so-called fifth industrial revolution, that adds concepts such as “human-centeredness”, “sustainability and concern for the environment” to the core elements of the fourth industrial revolution of transforming the industrial structure through the use of AI, IoT and big data.
A little more than a year ago, in November 2022, the general public heard for the first time about Generative AI. Since then, Generative AI quickly became not only a hot and buzzy term but a reality that is already transforming businesses. Generative AI refers to the next generation of AI technology that will power many products using cutting-edge machine learning models, such as Large Language Models (LLMs) and image generating models.
Cloud computing: the gateway for GenAI and for energy efficiency
Cloud computing offers a doorway for Generative AI and the combination of cloud and Generative AI already stands as a transformative force, offering cost-effective, innovative and scalable solutions to organizations. For the years to come, we’ll see an increase in cloud computing adoption that will allow public sector and private companies to digitally transform, become more cost and energy efficient and reach their environmental goals while improving security.
Cloud computing platforms bring compute, storage and networking, which are needed for GenAI to scale. For many organizations, cloud adoption starts with a simple objective: handling core computing tasks like storage, networking, and application management more cheaply and efficiently. Getting all of these to play well together requires a network that is secure, reliable, and fast enough to process data in real time, even in areas with low accessibility and high latency. By moving data out of physical servers, organizations can take advantage of virtually unlimited storage while saving money on upkeep. They also unlock cloud computing’s other networking and computing benefits, such as high reliability and minimal downtime.
But the cloud also offers complementary tools to facilitate data processing and analysis. For example, thanks to the cloud’s ability to deliver computing power at relatively low cost, artificial intelligence and machine learning are becoming mainstream. Traditionally, organizations have looked to the public cloud for cost savings and to develop rich, web-based applications. Today, public and private sector organizations around the world recognize they can achieve superior security outcomes via the cloud in comparison to on-premises data centers.
Sustainability issues are growing in political prominence in Europe, with initiatives such as the green deal which aims to make Europe the first climate neutral continent. Data centers from public cloud service providers are more energy efficient as a typical enterprise data center and a clear benefit of the use of cloud computing. European companies and associations launched a Climate Neutral Data Centre Pact. The alliance pledges to improve the sustainability of their operations in the EU, and work with the European Commission to help deliver a Climate Neutral Data Centre industry by 2030. Definitely, one strong reason for companies to opt for the public cloud is cost reduction both in economic and environmental terms.
Opportunities and benefits
A generative model can use what it has learned and create something completely new based on that information. Large language models (LLMs) are an example of generative AI since they create novel combinations of text in the form of natural-sounding language.
Generative AI will augment existing technology tools as well as create new opportunities for enterprises in delivery of services (customer service, manufacturing, research, product development). GenAI is already a reality for business. For example, global beauty company Estée Lauder Companies (ELC) announced back in August 2023 it will start to use generative AI across ELC’s brand sites to help them understand consumer sentiment and power its research and development efforts for product innovation.
Generative AI allows developers to quickly ship new experiences including bots, chat interfaces, custom search engines or digital assistants. Generative AI allows businesses and governments to turn large, complex volumes of data into summaries, interactive multimedia experiences, and human-like conversations. Many organizations are interested in leveraging this technology for not only customer-facing experiences, like in branding, but also more complex data science scenarios. For example, digital assistants can help data analysts and business users up-level their skills, by generating SQL queries, enabling exploration of data through natural language queries, and more.
As we look at what’s next for the cloud, we are currently experiencing one of the most significant technological shifts in history, with the emergence of generative AI. It has the potential to transform entire businesses and industries. If we look back at the virtual machines cloud era, cloud choices were primarily about storage, compute and databases. This is why most startups and enterprise companies turned to the cloud, because it was cheaper and more convenient than building and owning data centers. The cloud, then, evolved to enable the largest-scale use of data and organizations increasingly migrated to the cloud to reduce cost and use collaboration tools to be more efficient. We’re now in an AI transformational era where the cloud is an integral element needed for organizations to progress in their digitalization journey and for embracing the most disruptive technologies.
Challenges and risks
Unless people trust the technology, AI will not be adopted at a huge scale and organizations and individuals won’t be able to get all the benefits. Trust and transparency are sides of the same coin and are directly linked as well to AI explainability. Discrimination, unfair bias, privacy and security breaches, copyright issues, are among the challenges that AI faces.
Responsible choices need to be made around AI since, actually, technology can create both harm and good. Organizations want to know they can trust AI systems to make good, fair recommendations that reflect human values and act in ways that are helpful to individuals and society. AI advancements can create meaningful positive change but there is a clear need to ensure responsible development and ongoing governance in every scenario where AI is deployed. Generative AI is an emerging and rapidly evolving technology with complex challenges and this is why responsibility and accountability plays a key role. AI/ML is a transformational technology and those that are thought-leaders and commercial leaders have a responsibility to define how this tech is used and find ways to build products to enable responsible use.
AI works by learning from diverse information and data. Thus, there is a challenge with ensuring that rightholders protect their creative works, while at the same time, a pro-innovation, balanced approach is followed, that encourages the development of AI by opening access to the data needed to train AI systems.
AI is too important not to regulate but there is a need to get it well. It’s critical that regulations support enterprises and governments of all shapes and sizes seeking to use innovative approaches for delivery of beneficial goods and services, while also ensuring we create clear guardrails to mitigate against potential harms. But making AI work for the people is not just about addressing potential threats, it also means that promoting innovation needs to be at the core of the regulation. For AI regulations to be future-proof, they need to be technology neutral and risk-based. An important legislative milestone was recently reached in the EU with the political agreement of the EU AI Act by the end of 2023. The result is a risk based approach pioneer regulation that focuses on high risk AI.
Policy recommendations
AI regulation-risk based approach. Adopt a proportionate and risk-based approach in regulation recognizing that AI is a multi-purpose technology-and regulatory requirements should be calibrated to the particular risk and use case. Over the coming weeks, legislators should continue to work on a version of the EU AI Act that is aligned with international efforts, stays close to a risk based approach and that is consistent with existing rules and regulations in the EU. Regulations should account for the complexity of the AI ecosystem – providers, integrators, data providers, deployers, users, and others – to ensure AI gets the proper oversight it needs. AI regulations should be underpinned by international standards, which appropriately allocate obligations to the entity that is best suited to control the risks that regulators seek to address.
Responsible AI. Promotion of best practices. AI responsibility practices are a top priority. Transparency, auditability, fairness evaluations are all means to ensure products are safely and responsibly deployed. There are many tools, techniques, and methods to assist deployers as they seek to safely operate the product. Establish a clear AI governance and conduct ethical reviews of the systems, avoiding bias and incorporating privacy and security safeguards. Public-private collaboration is needed. Governments, tech companies, academics and research institutions must work together to address AI challenges. No single organization will be able to build the AI future on its own. On the contrary, there is a need to exchange knowledge, concerns and best practices among industry, civil society and governments to maxmize AI benefits while mnimizing downsides.
Cloud first policies promotion by Governments. Cloud is an enabler of emerging technologies like AI/GenAI. Public cloud services ought to be considered first for the public sector’s’ IT infrastructure and take precedence over solely on-prem and private cloud solutions – provided that all the necessary security requirements and international standards are met by the Cloud service provider.
