[Read more session reports and live updates from the 13th Internet Governance Forum]
Many emerging technologies deserve discussion, as do the socio-economic and ethical dimensions of artificial intelligence (AI). Understanding how to adapt emerging technologies will help to increase the inclusive digital economy for different parties.
The main session was co-moderated by Mr Vladimir Radunovic (Civil Society, Europe), DiploFoundation, with Ms Olga Cavalli (Government, Latin American and Caribbean Group and Caribbean Group (GRULAC)), President of the ISOC Argentina Chapter.
Cavalli introduced the session by saying that emerging technologies are changing our lives: AI, machine learning, and jobs performed by robots. In his turn, Radunovic started by asking the panellists to define the emerging technologies. He emphasised their roles in job development, economics, security issues, and transparency.
Mr David Redl, Government, US, Assistant Secretary for Communications and Information at the United States Department of Commerce, started by providing background on the NTIA, an agency of the US Department of Commerce. He explained that the agency is working on a free and open Internet, a long-standing commitment to the multistakeholder process, and on minimal barriers to global exchange of information.
Redl also highlighted the promise of emerging technologies; for example, the future of AI. He pointed out the role of connectivity. Part of NTIA's mission of is to expand the use of spectrum and broadband around the US because most emerging technology will rely on its connection to the Internet. He mentioned that the NTIA has made progress on many of these issues.
Redl said that blockchain has broken the cyber-physical barrier and he gave an example of programming e-commerce sites. He explained that when these sites are programmed, it is just a matter of transfer. Furthermore, he noted the existing North-South divide vis-à-vis cryptocurrencies.
Ms Lorena Jaume-Palasi, Civil Society, Germany, WEOG Algorithmic Decision Making Manifesto-Algorithm Watch, started with comments on transparency related to AI and to the emerging technologies in general. She emphasised that transparency is not a value in itself, it is a tool. Moreover, she said that it is the very first step towards a more complex set of rules and norms that will make artificial intelligence efficient.
However, according to Jaume-Palasi, a transparency code only is not going to help civil society on a very basic level. She pointed to the role of social discussion concerning AI and addressed the understanding of the human object of technology and potential human conflicts.
Mr Maarten Botterman, Civil Society, WEOG, ICANN Board, Dynamic Coalition on the Internet of Things (IoT), said that we need to take into account ethical issues. He noted that developers and employers need guidance, particularly in consideration that it is all about people in the end; we need to be more pro-people in development itself. Bottermann highlighted that ethical considerations come to us in two ways. The first one is, does a product give people the transparency needed? Do we secure transparency well enough? And the second concerns privacy issues.
Ms Layla El Asri, Private Sector, Research Manager, Western European and Others Group (WEOG), commented on the ethical perspective from a private-sector view. First, she said that ethical concerns are really complex. It changes over time and it is not something on which everyone agrees. Second, she discussed how to take into account ethical considerations when you build an artificial intelligence with a mathematical formulation.
According to El Asri, fairness should play a big role in algorthms. Moreover, she highlighted interpretability: Can you understand what your model does? Can you understand when it fails? According to her, these questions can help us to understand the way AI algorithms perceive the world.
She pointed to building models so that they can work with humans and ethical issues in this area. Finally, she noted that building ethics into AI through fairness and ensurance that it will treat different groups equally.
Mr Satish Babu, Chair, APRALO, ICANN At-Large, Asia-Pacific Group, highlighted two points. His first point was about accountability, and the dichotomy between developers and users. He said that civil society has no power regarding the development of algorithms; algorithms are driven not only by code, but also by data.
His second point was that the instruments used in building software or code can involve ethical concerns. He mentioned a project at the Massachusetts Institute of Technology that takes into human concerns into the design process itself, not just in the coding.
The session ended with further discussion from the Dynamic Coaltion on Emerging Technologies, in which the panellists stressed the main ideas of ethics and transparency.
By Gilles D. Bana