DFiR debate series: Artificial Intelligence in Research and Innovation
Artificial intelligence has the potential to revolutionise our society. This is true not least for the Danish research and innovation actors who, with the use of artificial intelligence, are expected to be able to increase the quality, scope and relevance of investments in research and innovation. However, the technology may also bring significant changes to the framework for research and innovation, including familiar processes such as peer review and research applications. Such significant changes to the research and innovation landscape must be accompanied by considerations about ethics and safety, and it ought to be discussed whether the relevant Danish communities are seizing the new opportunities to a sufficient extent.
In 2023/24, the Danish Council for Research and Innovation Policy will be hosting a series of debates on the potentials and risks of artificial intelligence for the Danish research and innovation ecosystem. The debate series is held in collaboration with the Ministry of Digital Government, the Pioneer Centre for Artificial Intelligence and the Knowledge Center for Foundations in Denmark.
The purpose of the debate series is to qualify public dialogue on the use of artificial intelligence in research and innovation. This applies both to conversations taking place around the lunch tables in innovative companies and at Danish universities, to the public dialogue and to political discussions.
Working group and process
DFiR's working group for this project consists of Frede Blaabjerg, Steffen Bohni and Tine Jess. The project is considered a dialogue-initiating platform rather than an analysis forming the basis for recommendations.
DFiR debate meeting I: The Danish ecosystem for Artificial Intelligence - balances and barriers
Time: Tuesday 12 December 2023, 15:30-17:00
Location: Ministry of Higher Education and Science, Børsgade 4, 1215 København K.
The event will be recorded: If you are unable to attend the event, you may register in order to receive a link to access the recording after the event.
The debate will focus on potentials rather than risks, and the conversation will be centred around balances and barriers in the Danish ecosystem. In DFiR's view, a healthy ecosystem surrounding artificial intelligence for research and innovation consists of competences and capacity within four pillars:
- Data (e.g. public data in datavejviser.dk)
- Algorithms (e.g. computer scientists who can develop algorithms to process data)
- Computing power (e.g. DeiC and private providers who can run the algorithms)
- Application (e.g. researchers or entrepreneurs with insights into problems in health or climate)
A discussion of the balances between these pillars will form the basis of our debate. It is important for there to be a reasonable level of competences and capacity within each pillar, and for the processes and networks that connect them to be well functioning.
A key question is this: Are we focusing investments correctly, and is there particular potential to be gained from increasing investments in specific pillars? Should we invest in additional computing power or are funds better spent investing in research projects that fund more PhD students in computer science? Are there specific barriers within each pillar and can they be addressed through policy initiatives or regulation? A separate question relates to the balances between application areas. Are there greater potentials in some research areas than others and are certain areas overlooked in the current funding landscape? Furthermore, questions will be asked about private actors' access to the ecosystem. How do we support a private market that applies research within AI? We will not be able to realise the full potential of this area if only publicly funded research communities have proper access to it.
Debate II: Artificial Intelligence and Risks - from plagiarism to geopolitics
Time: Monday, 22 January 2024, 15:00-17:00
Location: Statens Naturhistoriske Museum, Øster Voldgade 5-7, 1350 København K.
The event will be recorded: If you are unable to attend the event, you may register in order to receive a link to access the recording after the event.
In the second debate, we focus on the risks and pitfalls of increased use of artificial intelligence in Danish research and innovation. It is about the concrete challenges in everyday life in laboratories and research libraries, where large language models, such as ChatGPT, have made it easier to fabricate false references and plagiarise the work of others. In the debate, we will ask how these new challenges ought to be handled. Should it be left to each institution to develop guidelines or is there a need for national or European initiatives?
The debate will also invite more fundamental ethical considerations about the consequences of the use and dissemination of broad artificial intelligence in research and innovation. Where are the ethical pitfalls? A central discussion concerns the biases that we risk reproducing when using artificial intelligence. We ask which forums are taking on the task of developing the conversation about ethical issues. Finally, the use of artificial intelligence will be discussed in the context of security policy. Where is the balance between an ethically sound approach and the risk of losing the race with China and the US? And where should Denmark place itself?
Debate III: Artificial Intelligence and new conditions for research and innovation
Time: Wednesday 28 February 2024, 15:00-17:00
Location: Danish National Research Centre for Working Environment, Lersø Parkallé 105, 2100 København
The event will be recorded: If you are unable to attend the event, you may register in order to receive a link to access the recording after the event.
The third debate is about the conditions for research and innovation, and we will ask whether the emergence of artificial intelligence gives us reason to rethink the organisation of the research and innovation system.
In the existing system, researchers spend a lot of resources on assessing and evaluating each other. This applies to application and grant processes as well as to the procedures connected to publishing, e.g. in journals or books. This system was developed at a time when there was far less research and when individual researchers could be expected to have an overview of a much broader research front. This is not the reality today, where few people have the opportunity to have deep insight across fields. The current system is costly and the many rejections of grant applications are emphasised as a major challenge for the development and retention of employees at public research institutions. There are also ongoing discussions about whether the current system is fair and leads to the promotion of the best ideas. For innovation efforts, companies spend a lot of time applying for funding through systems that operate with relatively long processing times. This is a hindrance to a well-functioning innovation promotion system.
The debate will be a discussion of whether artificial intelligence can be used in the future development of the conditions for research activities. This includes peer review and publication patterns, as well as application and grant structures. Can artificial intelligence change our crediting and resource allocation traditions and for example develop alternatives to inefficient application processes? We will focus on the challenges and opportunities faced by foundations, institutions and publishers.
Contact
Document Actions