The world of research is rapidly changing. The past 20 years have seen huge developments in the way research is conducted, with many research tasks dramatically changed. There is a growing consensus that we are at the start of a 4th industrial revolution, with the rise of the Internet of Things, 3-D printing, nanotechnology, biotechnology, 5G, new forms of energy storage and quantum computing.
It is evident that research has been changed by these new technologies and in particular the growing use of machine learning and other artificial intelligence techniques to augment the work of research staff and alter their experience of research. In some cases, it may even replace them altogether. However, the rapid pace at which AI is developing means that there are many unanswered questions about exactly how technology is changing research. What will the long-term effects be? How should policy makers approach the increasing use of AI in research?
This paper describes the progress of AI through history and how it has led researchers to develop a range of techniques including natural language programming and computer vision. We then examine these techniques and their practical application across different scientific fields, and the ways in which this has augmented research processes. This leads us to question the exact role of AI and the extent to which it can replace human researchers. We look at both the challenges posed by the advancement of AI and the challenges that may hinder this progress. Finally, we raise questions to be considered for the future of AI in research and for policy making in the second phase of this research project.
- Artificial intelligence and robotics have made significant progress since the Second World War. The exponentially increasing computational capacity of computers has, over the last decade, enabled the use of machine learning and artificial neural networks which have dramatically increased the potential of artificially intelligent systems.
- Across many fields and industries, research is adopting a range of tools and techniques enabled by artificial intelligence, including: natural language processing, computer vision, automated monitoring, automated experiment selection and the prediction of physical systems.
- There is an open question as to whether this automation of research will free up and empower researchers to be more creative and productive, or whether it will instead replace them entirely. This may vary from field to field.
- It is also uncertain how long the current rate of progress in artificial intelligence can keep up due to hardware limitations. Further, the costs involved may be prohibitive for all but the richest countries and companies to deploy these technologies in research.
From this, we have a number of questions about how best to forecast the impact of the convergence of 4th industrial revolution technologies, including artificial intelligence, on research and what policy interventions may be needed to ensure the most beneficial outcomes for researchers, their output, and society at large. We will consider these questions in detail in the second phase of this research project.