Three Factors that are Critical to Building Human-Centered AI
By Michael Schönstein and Kristin Lena Keveloh
Artificial Intelligence (AI) is everywhere and it’s both an opportunity and a challenge. As our society stands on the brink of these disruptions, we can design systems that enable and promote the responsible use of AI, putting humans at the centre.
What are the implications for our society when machines become more and more capable? How can we leverage new technologies to make tomorrow's work better for us humans?
A panel discussion that took place on 28 January 2019 in Berlin and was co-hosted by LinkedIn and the German Ministry of Labour and Social Affairs’ Policy Lab Digital, Work and Society addressed implications of AI and other emerging technologies on our society.
Three factors emerged as critical for building a human-centered use of AI and other emerging technologies:
Social skills and ongoing learning are key for the future of work. Panellists agreed that there always will be jobs. Nonetheless, the nature of work will certainly change. Therefore, it is obligatory to guarantee high levels of employability among the workforce. Focusing on foundational social skills – such as teamwork and communication skills – combined with a mindset of ongoing learning is the key to ensure high levels of employability that goes beyond specific tasks.
Addressing diversity is key to building better AI. Currently there is a significant lack of diversity within the tech industry. This deficit is reflected in the products developed and, more significant, in the data sets used to develop AI. When developing and deploying AI, an important question to ask is what subsets of the truth we are seeing in the data sets we work with. Addressing this two-fold absence of diversity is essential for building a human-centered AI. Employers need to go the extra mile in order to find diverse AI talent by reaching out to candidates beyond their traditional networks.
Regulation is necessary, and requires transparency and trust. There was general agreement that regulation of AI and other emerging technologies is important and practical. Panellists cited the insurance industry as an example of a regulated industry that is still capable of innovation and growth. Doing so requires both regulators and innovators to establish a trustworthy and open relationship where they are learning from each other to craft laws and build technologies to minimize risks and benefit society.
The panel discussion marked the kick-off of a transatlantic dialogue series by the newly established Policy Lab Digital, Work & Society, hosted in the Federal Ministry of Labour and Social Affairs. The Lab is an interdisciplinary and agile entity that combines the functions and working practices of a traditional think tank and those of a modern future lab, aimed to develop strategies to tackle the changes ahead.
The panel was moderated by Geraldine de Bastion, co-founder of Konnektiv and included Bjorn Böhing, State Secretary at the Federal Ministry of Labour and Social Affairs; Allen Blue, co-founder of LinkedIn; Laura D’Andrea Tyson, former Chair of the US President's Council of Economic Advisers and professor at the Haas School of Business of the University of California, Berkeley; Chris Nadine Kranzinger, data analytics strategist at QuantCo; and Michael Schönstein, Head of Strategic Foresight and Analysis, Policy Lab Digital, Work & Society of the Ministry of Labour and Social Affairs.