Multi-stakeholder Discussion on issues about Generative AI

Knowledge Graph of Debate

Session report

Moderator – Yoichi Iida

In the field of AI governance, the importance of interoperability between different policy frameworks was underscored. This recognition stemmed from the understanding that transparency and predictability are crucial for ensuring effective governance and regulation of AI technologies. Japan, in particular, has pursued a non-binding software approach in AI governance.

The Hiroshima AI Process was launched in May and aimed at establishing high-level guiding principles and a code of conduct for AI actors. The process consisted of working group meetings held online, where priority risks, challenges, and opportunities presented by generative AI were discussed. The initiative received positive sentiment and was perceived as a step towards shaping responsible AI development and deployment.

One of the key arguments put forward was the need for open and inclusive discussions on the risks and challenges associated with AI. Yoichi Iida emphasised that by openly addressing these concerns, it would be possible to harness the full potential of AI while ensuring its safe and responsible use. This argument received positive sentiment as it was seen as a way to promote innovation and economic development by making the best use of the technology.

The positive potential of new AI applications and their potential to improve society and drive economic development were emphasised by Yoichi Iida. The sentiment towards the benefits of AI applications and systems was positive, with a focus on the positive impacts they can bring to various sectors of the economy. Panelists were invited to share their company’s services and the benefits they can provide to society, further highlighting the potential of AI applications.

Adapting AI solutions to local conditions was deemed vital for their successful implementation in communities. Speakers provided examples illustrating how different regions and communities have diverse needs and priorities when it comes to technology and AI usage. Failure to consider the realities and contexts of different communities may result in AI solutions that fail to meet their intended purpose. This neutral argument stressed the importance of understanding and incorporating local conditions into AI development.

The concept of interoperability between different AI frameworks was identified as another important aspect of AI governance. The ability of different systems to work together effectively was seen as necessary to deliver comprehensive and efficient solutions. This neutral argument emphasised the need for compatibility and collaboration between different AI frameworks to address complex challenges and achieve desired outcomes.

The potential of AI among digital technologies was highlighted and received positive sentiment. It was acknowledged that AI has far-reaching opportunities and potentials in various fields, presenting possibilities for transformation and innovation. This positive sentiment towards AI’s potential emphasised its significance as a driving force in the digital age.

Support for collaborations among companies, governments, and international organisations to facilitate the benefits brought about by AI was expressed. The World Bank’s active involvement in development support activities in the digital field was cited as an example, reinforcing the importance of collaboration to expand impacts and create effective AI solutions. This positive argument emphasised the need for collective efforts in harnessing the benefits of AI.

Promotion of collaboration among different stakeholders for AI evolution was seen as crucial. Yoichi Iida highlighted the potential for collaboration among different AI players in the ecosystem, and the Hiroshima process was identified as a means to foster such collaboration. International organisations such as the World Bank were urged to promote collaboration and share their knowledge and experiences. This positive sentiment demonstrated the recognition of stakeholder involvement as a catalyst for AI advancement.

Efficient AI development was seen as dependent on multi-stakeholder involvement. It was argued that finding the best solutions for enhancing AI projects involved the collaboration of public and private sector entities, as well as the World Bank. The involvement of different types of players in the AI ecosystem was deemed necessary to pave the way for efficient AI development. This positive argument reflected the need for diverse perspectives and expertise in shaping AI initiatives.

The notion of capacity building programmes was discussed, with the suggestion of extending these programmes to players from other countries in collaboration with the World Bank. This positive sentiment indicated an openness towards multinational collaboration in capacity building, recognising the importance of knowledge sharing and skill development in AI.

In conclusion, this summary highlights key discussions and arguments regarding AI governance and the potential of AI applications. It underscores the importance of interoperability, open discussions on risks and challenges, adaptation to local conditions, stakeholder collaboration, recognising AI’s potential, and capacity building programmes. The involvement of various stakeholders and international organisations is advocated to harness the benefits of AI in a responsible and inclusive manner.

Luciano Mazza de Andrade

Luciano Mazza de Andrade, the Director for Science, Technology and Innovation and Intellectual Property at the Brazilian Ministry of Foreign Affairs, holds a significant position in driving technology and innovation in Brazil. He recognises the potential of Artificial Intelligence (AI) in addressing the challenges faced by different countries.

Andrade highlights the transformative power of AI in various sectors. He emphasises how AI solutions can improve the provision of public services and e-government, enhance agriculture and food security, and contribute to the advancement of healthcare and education. In Brazil, there is a specific focus on AI innovation, supported by a legal framework designed to boost start-up growth. However, Andrade also stresses the importance of adapting AI to the local needs and communities of developing countries, as AI models are primarily trained on English language data, which may not reflect the realities and nuances of these nations. Additionally, AI models can also contain biases that may perpetuate inequalities.

To fully harness the benefits of AI, developing countries must establish adequate infrastructure and robust governance frameworks. Andrade points out that without these foundational elements, it will be challenging for these countries to take full advantage of AI’s capabilities and potential.

Furthermore, Andrade highlights the significance of dialogue and cooperation in the global AI landscape. He particularly praises Japan’s leadership role in this field and suggests engaging with different development banks to leverage investments. He believes that dialogue and cooperation are vital for sharing experiences, best practices, and building the necessary national capabilities. Andrade suggests that Japan should strengthen dialogue with other international organizations to avoid fragmentation of initiatives. Incoherence in narratives and policies can hinder progress, so building momentum at the United Nations for an inclusive dialogue is essential.

In conclusion, Luciano Mazza de Andrade underlines the importance of AI in addressing global challenges. However, he emphasizes the need to adapt AI to local needs and communities, establish adequate infrastructure and governance frameworks, and foster dialogue and cooperation. By considering these aspects, countries can fully harness the potential of AI for sustainable development and inclusive growth.

Hiroshi Maruyama

Hiroshi Maruyama’s company is heavily invested in hardware development and aims to broaden the applications of generative AI. One notable accomplishment is the development of an energy-efficient supercomputer, which received recognition by ranking highly in the Green 500 supercomputer ranking. This demonstrates the company’s commitment to sustainability and reducing energy consumption in the tech industry.

Additionally, Maruyama’s company utilized a deep learning method to improve the speed of new material discovery. By harnessing the power of AI, they have accelerated the process of discovering new materials, which has implications for industries such as medicine, manufacturing, and energy.

An interesting collaboration highlighted in the research is with Cowell Corporation, resulting in the creation of a virtual human generative model. This joint effort showcases Maruyama’s company’s expertise in combining AI technologies with virtual modelling, with potential applications in entertainment, virtual assistance, and virtual reality.

Maruyama himself is a strong advocate for pushing the boundaries of AI beyond human perception. He believes in the immense potential of this technology and its ability to transform various sectors. To support this, he cites ChatGPT’s scalability law, which highlights the limitless scalability of AI language models. Additionally, his company’s development of the Matalantis software for materials informatics demonstrates their commitment to expanding AI into new technological domains.

In a noteworthy collaboration, Maruyama’s company played a crucial role in the creation of the W3C virtual human generative model. This signifies the significance of their contributions towards advancing virtual human technology, ensuring ethical and responsible development in this field.

Despite these achievements, Maruyama expressed concerns about the energy consumption and cost of current hardware technology. He believes that existing hardware technology is too energy-consuming and expensive. However, his company has developed their own accelerator technology, enabling the creation of one of the most energy-efficient supercomputers. This demonstrates their commitment to addressing the challenges associated with energy consumption and cost in hardware development.

In conclusion, Hiroshi Maruyama’s company is at the forefront of hardware development and expanding the applications of generative AI. They have achieved significant milestones, including the creation of an energy-efficient supercomputer, collaborations in virtual human generative modelling, and breakthroughs in materials informatics. Maruyama’s passion for pushing the boundaries of AI beyond human perception is evident in his company’s accomplishments. However, he also recognizes the need to address energy consumption and cost issues in current hardware technology, taking steps to mitigate these challenges. Through their innovations and commitment to responsible development, Maruyama’s company is contributing to the advancement of AI and shaping its future.

Amrita Choudhury

The use of Artificial Intelligence (AI) in emerging economies has the potential to bridge the divide between these economies and developed countries, as well as improve public distribution systems. AI can be a powerful tool for development, helping emerging economies leapfrog traditional barriers. Its application in agriculture, for instance, can facilitate smart farming, maximize crop benefits, and assist in climate control. Furthermore, AI has the potential to enhance public distribution systems if accurate and relevant datasets are used.

However, it is important to acknowledge that AI algorithms, often developed in the Global North, can display inherent biases when applied in diverse global south contexts. This is attributed to the varying socio-cultural variables, environments, and genealogical distinctions across regions. Therefore, it becomes crucial to respect local cultures and customize AI solutions to specific localities to ensure inclusive and equitable outcomes.

Greater transparency, accountability, and local customization should be guiding principles in AI development. Respecting local cultures and conditions is essential, as is using AI responsibly to avoid or minimize biases. Promoting dialogue between industry innovators, regulators, and civil societies will allow for a collaborative shaping of the direction AI applications should take. This will contribute to sustainable development and progress, aligning with SDG 16: Peace, Justice, and Strong Institutions.

Regarding regulations, the development of AI should be guided by strong frameworks and best practices rather than restrictive regulations. This approach allows for innovation and growth, aligning with SDG 9: Industry, Innovation, and Infrastructure. Restrictive regulations could hamper the growth of small and medium enterprises and hinder innovation. Thus, collaboration, dialogue, and capacity-building around AI are encouraged.

Collaboration is necessary due to the cross-border nature of technologies. The interconnectedness of AI and its global reach requires cooperation among countries and stakeholders. By working together, they can address common challenges, share knowledge, and foster partnerships that contribute to achieving SDG 17: Partnerships for the Goals.

In developing countries, training and capacity-building play a crucial role in ensuring the effective and responsible use of AI. By investing in quality education, governments and organizations can equip individuals with the skills needed to leverage AI technologies for their benefit. This aligns with SDG 4: Quality Education.

It is also important for AI systems to respect rights and promote gender equality. Embedding these principles in AI systems ensures that they do not perpetuate discrimination or biases. This aligns with SDG 5: Gender Equality and SDG 16: Peace, Justice, and Strong Institutions.

Lastly, the security of AI systems must be prioritized to prevent misuse. State actors attacking different countries and bad actors infiltrating the system could have significant consequences. Protecting AI systems from these risks contributes to maintaining peace, justice, and strong institutions as outlined in SDG 16.

Entities like the World Bank can play a significant role in providing training and sharing best practices. By supporting capacity-building efforts and offering guidance to governments and stakeholders, they can help maximize the positive impact of AI development. This aligns with SDG 17: Partnerships for the Goals.

In conclusion, the use of AI in emerging economies has the potential to drive development and bridge the gap with developed countries. However, it is crucial to address biases, respect local cultures, and customize AI solutions accordingly. Transparency, accountability, collaboration, and capacity-building are important factors to ensure responsible and inclusive AI development. By addressing these aspects, AI can contribute significantly to achieving various Sustainable Development Goals.

Daisuke Hayashi

Daisuke Hayashi, a Senior Digital Development Specialist at the World Bank, focuses on digital infrastructure, international cooperation, and scaling up digital skills. The World Bank has been actively involved in digital development, supporting the growth of AI technology, and expanding digitalisation in developing countries. Hayashi’s work involves addressing the challenges associated with achieving consensus on AI within the international community.

Daisuke Hayashi recognises that getting consensus within the G7 countries on AI is a difficult task due to the potential and associated risks of AI technology. This acknowledgement highlights the intricate and complex nature of reaching an agreement on AI policies and regulations between different countries.

The World Bank has been actively working on building digital infrastructure and skills to bridge the gaps between connected and unconnected areas. They have supported infrastructure construction projects aimed at improving connectivity in developing countries. Additionally, the World Bank’s Digital Development Partnership focuses on capacity-building initiatives to expand and develop digital skills in these countries. This support is crucial for promoting development and reducing poverty in these regions.

The World Bank also emphasises the importance of establishing effective regulatory frameworks for AI. They collaborate with private companies and public sectors to find the best solutions for creating an efficient regulatory framework in the field of AI. This demonstrates their commitment to ensuring that AI technology is developed and implemented responsibly, taking into consideration ethical and legal considerations.

In conclusion, Daisuke Hayashi’s work at the World Bank focuses on digital infrastructure, international cooperation, and scaling up digital skills. The World Bank has been actively involved in digital development, supporting the growth of AI technology, and expanding digitalisation in developing countries. They acknowledge the challenges in achieving consensus on AI within the international community and are in favour of establishing effective regulatory frameworks for AI. Overall, the World Bank’s efforts in building digital infrastructure and supporting skill development play a crucial role in promoting economic growth and prosperity in developing regions.

Bonifasius Wahyu Pudjianto

Bonifasius Wahyu Pudjianto, a key figure at the Ministry of Communication and Informatics in Indonesia, is extensively involved in promoting IT literacy and nurturing the start-up ecosystem. His role primarily focuses on digital iteration and the start-up ecosystem. He is responsible for handling the IT-sector capabilities of all people and promoting IT literacy among the population. Pudjianto’s efforts align with SDG 9, which emphasizes industry, innovation, and infrastructure.

AI technology has rapidly gained prominence in Indonesia, contributing significantly to various sectors, including healthcare, education, skill development, poverty alleviation, and environmental humanitarian aid. AI solutions have improved remote healthcare access, and start-up companies have developed online courses for learners in rural areas. Moreover, AI solutions are being developed for disaster response and early warning systems. The widespread use of AI in Indonesian society has had a positive impact.

However, concerns regarding the ethical use of AI have been raised, and guidelines are being formulated to regulate its usage. These guidelines aim to ensure inclusivity, humanity, security, democracy, openness, credibility, and accountability in AI utilization, reflecting the need to consider individual rights and ethics.

Digital literacy is recognized as a vital component of societal capacity building, contributing to SDG 4 (quality education) and SDG 8 (decent work and economic growth). It is crucial for individuals to understand how to utilize AI and other technological advancements effectively and responsibly. Collaboration between industries and emerging start-ups is also seen as a key driver of innovation and economic growth.

To support the growth of start-ups, venture capital and financial engagement are essential. During the tech winter, many start-ups faced difficulties, leading to a decline in the start-up ecosystem. Venture capital and institutions like the World Bank play a crucial role in providing the necessary support and funding for start-ups to thrive, contributing to economic growth.

In conclusion, Bonifasius Wahyu Pudjianto’s work at the Ministry of Communication and Informatics in Indonesia focuses on promoting IT literacy and nurturing the start-up ecosystem. AI technology is widely utilized across various sectors, but ethical considerations and regulatory frameworks must be established. Strengthening digital literacy, fostering collaboration between industries and start-ups, and increasing financial engagement are identified as crucial factors for sustainable growth. Through these efforts, Indonesia can harness the potential of technology and innovation to drive progress and development in line with the relevant Sustainable Development Goals.

Melinda Claybaugh

Meda is actively developing AI products with a core focus on connecting people. These products utilize generative AI technology, allowing users to create and share images within the apps. This enhances the user experience and promotes interaction and engagement among users. Meda’s objective is to foster social networking and connect individuals.

In addition to connecting people, Meda is also investing in open source tools and products to democratise access to AI. They have recently launched the Lama2 open source large language model, allowing developers and researchers to utilise it in the field of AI. Meda aims to empower individuals and communities by democratizing access to AI tools, leading to a more inclusive and accessible AI landscape.

Furthermore, Meda’s commitment to making a positive impact extends to their Data for Good program. This program tackles societal challenges through the use of AI. It supports translations between 200 languages, including low resource languages, facilitating better communication and understanding among different cultural and linguistic backgrounds. Additionally, their Data for Good program includes the Relative Wealth Index, which helps governments increase social protection program coverage using artificial neural networks. Meda leverages AI technology to assist governments in making informed decisions and addressing social inequality.

In conclusion, Meda’s development of AI products that connect people, investment in open source tools, and the implementation of their Data for Good program demonstrate their commitment to creating a positive impact. These efforts contribute to achieving Sustainable Development Goal 9, focusing on industry, innovation, and infrastructure. Meda’s initiatives highlight the transformative power of AI for the benefit of individuals and society as a whole.

Natasha Crampton

Natasha Crampton, Microsoft’s Chief Responsible AI Officer, is responsible for implementing the company’s responsible AI principles. She works closely with engineering teams to ensure that Microsoft’s AI technologies adhere to ethical and responsible standards. Crampton also defines the policies and governance approach for AI implementation within the company.

In her external-facing role, Crampton takes the knowledge gained from Microsoft’s responsible AI practices and actively participates in public policy discussions. She advocates for the development of new laws, norms, and standards that promote responsible AI systems.

Microsoft is at the forefront of developing a suite of co-pilots, AI-powered products that enhance productivity and creativity in coding. These co-pilots assist both experienced and novice coders in accomplishing tasks more efficiently. For instance, GitHub Copilot allows users to code in plain language, making coding more accessible to non-coders. Microsoft’s suite of co-pilots has been widely accepted for their ability to enhance productivity and foster creativity in coding.

Microsoft also believes that AI technology has the potential to bridge gaps in access and communication, particularly in linguistically diverse communities. Initiatives like the Be My Eyes platform, which uses OpenAI’s GPT-V model, aid blind or visually impaired individuals by providing textual descriptions of visual information. In India, Microsoft has deployed an AI-enabled chatbot that allows users to access government services in their local languages, overcoming the linguistic barrier.

Effective multi-stakeholder collaboration is crucial for addressing the challenges posed by AI technologies. Microsoft advocates for identifying specific problems and directing resources towards finding solutions. The Christchurch Call serves as an example of a successful multi-stakeholder initiative, bringing together governments, civil society, and industry to work collectively towards addressing the issue at hand. Building on existing frameworks has also proven effective in multi-stakeholder collaborations, rather than reinventing the wheel.

Tackling the digital divide and focusing on upskilling are vital steps in fully harnessing the potential of AI technology. Many parts of the world still lack access to AI due to the digital divide. By addressing this divide and providing the necessary resources and training, individuals and communities will be better equipped to leverage AI’s benefits.

Overall, Natasha Crampton and Microsoft are committed to implementing responsible AI principles while driving innovation and inclusivity. Through their suite of co-pilots and efforts to bridge gaps in access and communication, they are demonstrating the positive impact AI technology can have. Effective multi-stakeholder collaboration, building on existing frameworks, and addressing the digital divide are essential steps in fully realizing the potential of AI.

Speakers