A group of GPAI experts propose a raodmap to encourage innovation and the public funding of research where market fails to address pressing issues.
To what extent can AI resources facilitate the discovery of new drugs? Answering that question was the mission of the team of experts on the AI and pandemic response working group co-chaired by Alice Oh, Associate Professor in Computer Science at the Korea Advanced Institute of Science and Technology (KAIST). Their findings are rather reassuring: it seems that AI offers immense potential in terms of pharmaceutical development. Not only can this technology help us to discover new classes of efficacious drugs, but it can also facilitate the design of smart and targeted therapies. With its vast resources, AI also has the potential to dramatically improve the speed and cost of conducting clinical trials, and give us a more detailed understanding of the basic science behind the mechanics of drugs and diseases.
An inadequate ecosystem
However, according to the experts, the current ecosystem restricts artificial intelligence from delivering its full potential. This situation is the result of a lack of financial investment in areas where the level of R&D is simply not sufficient, such as data and open science, which can provide input for more powerful and data-intensive AI algorithms.
In addition to high levels of funding, the experts are also calling for a cultural transformation in both the academic world and the pharmaceutical industry to embrace open data as integral to their drug discovery and development research. “This change will catalyse research in areas of high social impact, such as the treatment of rare diseases and the development of new solutions to address the threat of antibiotic resistance”, says the report.
The need for incentivization an public support
However, this transition can only be made successfully where there is a proactive policy of support from national governments. So the group og GPAI experts is recommending public authorities to invest in multidisciplinary academic research into AI-driven drug discovery. “Some government subsidies should be specifically earmarked for funding the construction of high-quality open data sets, as well as interdisciplinary collaborations to develop and test AI algorithms for tasks of particular relevance to public health, such as the development of new antimicrobials”, say the experts.
Governments’help is also called upon to improve AI literacy at every level of the ecosystem, and to focus on providing access to high-quality training programmes and the development of tools and resources that facilitate AI adoption. The working group’s recommendations also include the creation of innovation procurement programmes to stimulate and encourage biotechs, pharmaceutical companies and public-sector research centres to move beyond academic prototypes to launch industrial-scale drug development. These are all recommendations will be discussed at the 2021 GPAI Summit.