23.04.2024
Speech at the reception hosted by the New Zealand Ambassador to France for the launch of A guide for policy-makers: evaluating rapidly developing technologies including AI, large language models and beyond.
Your Excellency Ambassador Caroline Bilkey, thank you for hosting the science community on the occasion of the visit of the Honorable Judith Collins KC, Minister of Science, Innovation & Technology to Paris.
I am Sir Peter Gluckman, former chief science advisor to the prime ministers of New Zealand, and now president of the International Science Council (ISC).
The ISC is hosted in Paris by the government of France and brings the world’s active scientific organisations in both natural and social sciences together including national scientific academies, international scientific unions and social science associations together with many other organisations such as the global and national young academies.
Its role is complex – to be the primary interface between the active scientific community and the multilateral system especially regarding the myriad challenges of environmental, social and technological rapid change. Indeed tonight we release our contribution to addressing one of these challenges – more about that momentarily.
The ISC promotes international science cooperation through a myriad of partnerships and affiliates. Many diverse organisations you may be familiar with operate under the ISC banner such as the International Network on Governmental Science Advice, the World Climate Research Program, the World Data System, and the Scientific Committee on Antarctic Research. We take a leadership role in science diplomacy, work with UNESCO and other organisations on freedom and responsibility in science and trust in science and to take a key role in exploring and promoting how science can best contribute to policy making at every level form global to local. The ISC is also a key leader in exploring how science itself must evolve both in developed and developing countries to meet the challenges we all face.
New Zealand may be a small country, geographically isolated but emotionally and culturally unique in its connections to the Pacific, Europe and Asia. We are only 5 million people but are an active collaborative member of the science community. New Zealand is well known for recognising that its characteristics allow it to play roles that are valued on the world stage.
I am particularly grateful that for the last decade the New Zealand government has supported the critically important ISC committee on freedom and responsibility of science which is one of the most important and longstanding components of the International scientific apparatus.
But New Zealand has another unique contribution in bringing diverse perspectives into complex issues. This was demonstrated 15 years ago when Minister Collins was in an earlier government when it was arguably the first country to really appreciate the importance of big data to social policy decisions. But the social and ethical issues were recognised at the beginning and enormous strides were made. This led to the OECD asking the New Zealand team, led by myself and working globally through INGSA, to develop a framework for considering the interface between technology and society, providing concepts that became important in their early digital policy work.
And that early work leads us to today. Eighteen months ago, as debates over the impact of AI started to ramp up, I was approached by leading scientists to consider would the ISC take on an independent role as the assessor of rapidly moving AI technologies. The idea was superficially appealing but after many discussions in New York, Geneva, Paris, Brussels and with policy makers and academics, we recognised not only was that impractical but that there was a glaring gap in everyone’s conversation that needed addressing and hence the report we are releasing today.
With a technology such as AI, it is relatively easy to write high-level principles that in a utilitarian sense have little impact. But then to jump to discussion of generic governance and regulation becomes almost meaningless for such a fast-moving technology. What would you regulate and why? What risks need management. The huge ontological gap was obvious and so the ISC working largely through Hema Sridhar from New Zealand who led the writing team set out to fill the gap. This involved many discussions. An initial draft document was circulated last October for feedback and now a definitive but short document entitled A guide for policy-makers: evaluating rapidly developing technologies including AI, large language models and beyond is being released today.
The report is not intended to be either dogmatic or definitive. And while framed around AI, it could be applied with minor amendments to any rapidly moving technology such as synthetic biology. What frames it, is a structure that thinks through the many domains of self, social, commercial, and national life that need to be considered whenever a technology, or more likely a specific application is considered, for development and release, and the question becomes whether there is a need for regulation or other forms of governance.
The guidelines provide a provisional checklist of all the dimensions, both positive and negative, that merit active consideration. We suggest that all these need to be considered even if some may not be relevant in a particular context. This ensures that any claim or application is tested for both its value and risk and appropriate choices made. In this way it is rather like a pilot’s checklist. Even if the pilot thinks there is nothing to ask, she is still required to check.
In this sense the proposed approach is both highly novel and highly utilitarian. It paves a way ahead for governments to be more logical and balance restriction with supporting innovation, societies to get beyond extreme utopian or dystopian perspectives, and for developers to be more cognizant of the issues that these pervasive technologies might create and allow a more practical and pragmatic approach to the complex issues when to regulate and when not to.
The report tries to be agnostic to ideology, culture, or application but rather to be a guide to policy makers on how to approach one of the most critical issues of our times.
This report is a reminder of the critical value of international collaboration in science beyond empirical research, and what and how the International Science Council can uniquely contribute to essential global and national discussions where an integrated scientific perspective is critical.
It is also a reminder that small countries have unique value in science – it is the multidimensional thinking they can bring to the table and their ability to work globally and across cultures that I think makes New Zealand such a valuable partner.
