Artificial Intelligence (AI) & Emerging Technologies
Chat GPT, Generative AI, and Machine Learning
Speaker 1: GABRIELA RAMOS, Intergovernmental Organization, Intergovernmental Organization
Speaker 2: Anriette Esterhuysen, Civil Society, African Group
Speaker 3: Stefaan G Verhulst Verhulst, Technical Community, Western European and Others Group (WEOG)
Speaker 4: Alexandre Barbosa, Civil Society, Latin American and Caribbean Group (GRULAC)
Speaker 5: Dawit Bekele, Technical Community, African Group
Speaker 6: Dorothy Gordon, Civil Society, African Group
Siva Prasad Rambhatla, Civil Society, Asia-Pacific Group
Changfeng Chen, Civil Society, Asia-Pacific Group
Marielza Oliveira, Intergovernmental Organization
Fabio Senne, Civil Society, Latin American and Caribbean Group (GRULAC)
Yves POULLET, Intergovernmental Organization, Western European and Others Group (WEOG)
Xianhong Hu, Intergovernmental Organization, Western European and Others Group (WEOG)
Noémi Bontridder, Civil Society, Western European and Others Group (WEOG)
Round Table - 90 Min
- How to articulate the private and the public governance of the foundation models?
- How to organize a global debate on that technology in order to avoid counterproductive competition between regional approaches?
- Do we need a moratorium?
What will participants gain from attending this session? The contribution of speakers from different regions of the world and from different disciplines will allow the participants to have a first approach of the different ways in which the States and the private actors of the world of generative AI consider the risks linked to this technology and envisage their governance, including in their concrete developments. Furthermore, the speakers will give a first analysis of the different legal and ethical aspects (notably, privacy, multiculturality, intellectual property, harmful content, …). The meaning and the application of the ethical principles of the UNESCO Recommendation will be highlighted.
Generative AI like LLaMA, LaMDA, GPT-4, PaLM 2 developed by different major key players in our digital world refer to a category of AI algorithms supervised or not that generate new outputs based on the data they have been trained with. Unlike traditional AI systems that are designed to recognize patterns and make predictions, generative AI creates new content in the form of image, text, audio, and more. Some of them (BERT, DALL-E, FLORENCE or NOOR) are qualified as “foundation model”, since they are trained on a vast quantity of data resulting in a model that can be adapted including by individuals to a wide range of downstream tasks, preprogrammed or not.
The irruption of that new category of AI system in all sectors – like computer programming, artistic creation, education, personal, scientific innovations, healthcare, security systems, personal interaction, … – creates not only new risks as regards our individual liberties (privacy, freedom of expression, exploitation of our vulnerabilities, …) and potential collective discrimination but also societal risks as uniformization of opinions, environmental damages, competition infringements, impacts on the democratic functioning of our societies and violations of the rule of law. These systemic risks call for new methods of public and private governance, or even for a moratorium in order to allow a public discussion on the stakes of this technology and its limits.
Accordingly, the IFAP Information Ethics Working Group (IEWG) is proposing a workshop devoted to how some countries or regions are approaching the issue in light of the ethical principles established by the UNESCO Recommendation on the Ethics of Artificial Intelligence and, based on the analysis, to trigger policy reflections on how these ethical principles could be applied to the development of AI in countries and regions as well as in national jurisdictions.
The recording of the session will be published on the webpages of the Information Ethics Working Group (IEWG) of UNESCO and of its members’ institutions (in particular academic partners from Morocco, India, China, Germany, Belgium and America). The publication of an academic text by the IEWG can also be expected.
This workshop is part of the work of the IEWG, which is developing a series of workshops in several regions of the world.
Hybrid Format: We encourage hybrid participation to include in the discussion people from different countries, in particular from developing countries. Our online moderator, Xianhong Hu, has already organized several online events and therefore has experience in encouraging interactive debates among participants, which we aim for.