Future signals
In the final section of the State of Open Infrastructure report, we highlight a few emerging topics and trends that we’ve seen increasingly discussed in the open infrastructure and adjacent spaces over the past year. Inspired by Nesta’s Future Signals (2023), our team pondered recent developments and key tensions and their impact on the investment in and adoption of open infrastructure in research and scholarship. By sharing our thoughts, we hope to hold space for further discussions with the broader community.
The impact of AI on infrastructure in research and scholarship
With the rapid advancement in artificial intelligence (AI) models, algorithms, and tools and their increasing prevalence in our day-to-day lives, it is unsurprising that AI has come up in almost every conference and community conversation in the open science, open source, and open infrastructure spaces last year. AI is redefining digital research and scholarly communication infrastructure and funders’, institutional decision makers’, and researchers’ technological investment and adoption practices.
Redefining infrastructure for research and scholarship
Today, AI tools can readily execute or assist in many tasks that humans currently perform in the research production and communication lifecycle. They can create and enhance metadata, classify images, translate between languages and schemas, conduct data analysis, improve peer review, and more (Watkins, 2023). Similarly, AI algorithms are increasingly used by researchers and scholars in their research and academic workflows, e.g. automated transcription services to transcribe research interviews, or an image segmentation tool that is dependent on a machine learning algorithm. Inevitably, AI — the models, the training datasets, and even the hardware it depends on — is becoming part of research and scholarship infrastructure.
AI as an enabler for better open research infrastructure
Automation creates an opportunity for people, including those who are part of the human infrastructure in scholarly communication and research, to reduce effort on tedious tasks and to focus their labour on the aspects of their roles best suited to humans, exercising their creativity, social-emotional intelligence, and complex problem-solving skills.
There is a significant risk of funders and decision-makers promoting AI as a cost-reduction strategy, rather than as a way to achieve more with the same (or increased) resources and to empower human workers. While we acknowledge this risk, thinking of AI as an enhancement to rather than a replacement for human labour can expand the potential of what open research infrastructures can achieve. Take open data-sharing platforms as an example. If a curation team leverages AI to perform a series of quality checks and metadata enhancements during or immediately after the data submission process, they not only reduce human effort on tedious or repetitive tasks, but free up time to further improve data reusability: performing deeper quality checks than time might otherwise allow, providing support for authors, or developing creative approaches to promoting data reuse.
Centering principles and values in infrastructure decision-making
The potential applications of AI in research production and communication and research stakeholders’ interest in applying emerging AI tools also present novel challenges for research-performing organizations. A May 2023 UNESCO survey showed that less than 10% of schools and universities had developed institutional policies or guidance on the use of generative AI applications in education (UNESCO, 2023). While many institutions have unveiled policies since then (and EDUCAUSE has begun collecting information on its members’ AI policies — see EDUCAUSE, n.d.), it remains to be seen whether policy development to encourage the responsible use of AI in research will keep pace with the proliferation and evolution of AI tools. In particular, the black-box nature of many of these tools makes it difficult, if not impossible, to validate generative AI outputs. We will be watching with interest to see how institutions, research funders, publishers, and research infrastructure providers continue to adapt to this rapidly evolving landscape.
With AI becoming part of open research infrastructure, for everyone who cares about making informed decisions about the technologies we use, including us at IOI, the important question is how can we understand the implications of our technology choices in order to make better ones? As we look across the open infrastructure and ethical AI conversations, we are excited to see the parallels that exist: the considerations around transparency, accountability, and governance, the discussions around principles and values (Decklemann, 2023), and the recognition that openness is a spectrum (Ramlochan, 2023). As AI becomes part of research infrastructure, we see opportunities to use our experience working with infrastructure service providers and adopters to surface information to help adopters make more informed decisions. Of course, the “supply chain” here makes this complicated: how does the governance model of the training data that algorithm A depends on, that is used by infrastructure B, that is depended on by infrastructure C, affect the community using infrastructure C? We have no answers at the moment, but we are encouraged by fellow travellers on this journey.
The infrastructure powering diamond open access
Diamond Open Access (Diamond OA) is, in many ways, a new name for an old phenomenon – completely free online journals have been in existence in some form since at least the late 1980s. In recent years, there has been a heightened focus and urgency (as well as the creation of the label “Diamond”) to “no-pay” forms of publishing in reaction to the proliferation of the Article Processing Charge (APC) business model and the associated discussions of equity and disparities in publishing (see, for example, the 2023 reports and ongoing work from the Open Access Scholarly Publishing Association).[1]
This resurgence of interest has prompted recommendations of business models and sustainability for the publishing operations (OASPA, 2021), as well global events such as the Diamond Open Access Conference which held its second gathering in October 2023, co-organized by Redalyc, UAEMéx, AmeliCA, UNESCO, CLACSO, UÓR, ANR, cOAlition S, OPERAS, and Science Europe (Saenen et al., 2024).
Moving away from APCs
Funders have been reconsidering and evolving their approaches to open access publishing and APCs. Diamond OA is at the heart of large-scale projects such as DIAMAS in Europe, a three-year, European Commission funded effort to bring together 23 organisations across 12 countries to investigate models to advance non-profit publishing models that do not charge authors or readers as a counterpoint to APC-based OA.[2] Efforts such as cOAlition S and Plan S,[3] a consortium of research funding and performing organizations dedicated to making full and immediate open access a priority, have also converged on Diamond OA and “publish, review, curate” models of publishings as means to address current challenges in scientific publishing. The Bill and Melinda Gates Foundation recently announced that they will cease supporting APCs in 2025 (Bill & Melinda Gates Foundation, 2024). With APC inflation and proliferation being driven by funders’ willingness to pay for them as part of their mission to make the outputs of their funded research available to the public, we see these movements shifting back to the original declarations of the OA movement in a much more mature ecosystem for open access publishing. The backlash against APCs may swing the funding pendulum back towards Diamond OA and the infrastructures that power it; alternatively, it may result in the further development of other ecosystems, such as preprint and peer review.
Diamond OA is described as the equitable means to making knowledge a public good (Manifesto on Science as Global Public Good: Noncommercial Open Access, 2023), calling back to the original declarations on Open Access from the early 2000s (e.g. the Budapest Open Access Initiative[4]) and recognizing that some profit-driven motives have shifted open access publishing away from the original intent of the Open Access movement. The future signal of Diamond OA is that we are revisiting the roots of the OA movement and reckoning with the reality that the current open access publishing landscape, rife with rapidly inflating and proliferating APCs, is very far from the free access ideal envisioned in the early years of the OA movement. We are keenly interested and monitoring efforts for more equitable models with investments in infrastructures as well as incentives to drive more concerted shifts towards these models.
The risks of reinventing “plumbing” infrastructure
The existing ecosystem of publishing infrastructures, human and digital, are often dependent on other infrastructures (e.g. open standards, metadata infrastructure, and collective care infrastructures). This set of behind the scenes infrastructures has been named the “plumbing” infrastructure that powers open science (Pfeiffer et al., 2024). We see an important opportunity to fortify the plumbing of the interdependent open infrastructures that already power Diamond OA in order to build a robust open toolchain that can realistically compete with proprietary publishing workflows. Commercial publishers benefit from mature (usually closed) workflow systems that enable the full spectrum of publishing activity, reliably delivering content from author to reader. Shiny new and visible tools tend to attract funding (Skinner, 2019), leaving existing open infrastructures with the challenge of growing, evolving, and investing in research and development without large infusions of funding. As a buzzy topic, Diamond OA risks replicating this investment in new infrastructures at the expense of those that have powered these activities for decades. Investing instead in the plumbing — existing interdependent infrastructures that power the full lifecycle of Diamond OA (from standards and identifiers, to repository infrastructures, access protocols, discovery, and beyond) — bolsters the infrastructure for the global research community to fully participate in the Diamond OA movement.
The importance of sustainable investment in human infrastructure and labour
Renewed investment and commitment to developing these support structures for open infrastructures is a key pathway that we see for the Diamond OA ecosystem to ensure a robust technical ecosystem into the future. Investing in Diamond OA is not only about investing in digital infrastructure, but equally (and if not more) the human infrastructure powering it. If the needs of editors and staff who operate a journal are not taken into consideration, the effects can be dire, as seen in the continued rise in journal editorial boards resigning from subscription model journals. Recent examples of this mass resignation have circled around pressures to produce higher numbers of articles at a lower quality and with lowered standards in publication services (Abels & Flynn, 2024; Oransky, 2024; Weinberg, 2023). In that regard, It is important to call out the realities that have shifted since the first online open access journals were published in the 1980s: editing and reviewing used to be a regular part of an academic’s work, but increasingly, especially in the US, there has been increased pressure to publish a higher quantity of articles to achieve “success” in academia and less focus on services like editing or reviewing. This is coupled with the ongoing increase in faculty workloads at higher education institutions in response to institutional budget cuts, which provides further constraints on scholars’ time and energy (Marcus, 2021). In order to sustain Diamond OA models and operations, it is critical to invest in the human infrastructure and labour for them in a sustainable manner by making space for them in the regular course of scholarly labour. An increase in investment in the technical infrastructure cannot be successful without a complementary increase in investment in the people who use the tools.
With a re-energized focus on Diamond OA, we see a path towards shoring up the infrastructures that have long powered this method of publishing and creating a viable open access publishing ecosystem that heeds the call to return to the roots of open access.
Diverse visions for digital sovereignty and the impact on open knowledge infrastructure
Digital sovereignty can be defined as the right of a nation, region, or other political entity to assert control over its digital infrastructure and data, on its own behalf and on behalf of its citizens (Pohle and Thiel, 2020). From the development of the European strategy for data to the CHIPS and Science Act in the US, recent years saw a growing, evolving conversation and rapid policy development around digital sovereignty, prompted by nations’ and regions’ desires to (re)gain control over and independence in their digital infrastructure and data.
Distinct motivations and visions for digital sovereignty
We provided a working definition for digital sovereignty for this section, but we would be remiss if we did not acknowledge the diversity of approaches to defining and motivations for embracing digital sovereignty. The concept has been embraced by authoritarian governments in order to exert social control over their citizens, and by more democratic regimes to protect citizens’ personal data and businesses’ economic interests in an information economy. In Latin America, the loss of trust in how big data corporations and developed countries utilize data is one of the root causes of the drive towards digital sovereignty (Bosoer, 2022). The enclosure of the digital commons and its integration into infrastructure developed, maintained, and licensed or sold by for-profit entities which are subject to regulations applied to their home jurisdictions and those of their users makes geography more relevant than early internet visionaries might have anticipated. Taken to its extreme, this phenomenon has the potential to fall victim to the “Galápagos Syndrome” (2024), by which infrastructures become separated and segregated into smaller, detached and non-interoperable components.
Impact on global digital research infrastructure
We wonder what impact the geographic “anchoring” effect of the digital sovereignty movement will have on the free flow of knowledge and research around the world, and the infrastructures that support it. As we think about developing and maintaining sustainable, interoperable, and global digital research infrastructure, digital sovereignty can impact how these technologies are deployed locally. For example, cloud-based technologies need to work on local servers and cloud infrastructures in order for them to be used by researchers in a certain country. Digital sovereignty regulations may also have an increasing influence on institutional procurement policies (IT-Planungsrat, 2021), affecting what technologies researchers and scholars can use. Thinking about solutions that are going to be most successful because of their abilities to aggregate from data sources worldwide, increased regulations and restrictions on how the data can be shared, motivated by digital sovereignty, may lead to additional challenges and barriers in these solutions’ development.
Open-source technologies as enablers of digital sovereignty
Open-source technologies are increasingly seen as enablers of digital sovereignty as they can “cultivate trust through openness, direct involvement, and preserving entities’ autonomy” (Nordhaug & Harris, 2021). In some regions and countries, there is increasing governmental support for and investment in the development of open-source technologies to advance digital sovereignty of the region/country. The Sovereign Tech Fund, supported by the German Federal Ministry for Economic Affairs and Climate Action, for example, “supports the development, improvement, and maintenance of open digital infrastructure” to ensure that it is “available, accessible and secure [...] for digitalization in the public interest.”[5] We also observe an increasing number of regional and national Open Source Programme Offices, notably in Europe (Osborne et al., 2023). On the other hand, in the Majority World, where governments and the private sector simply do not have the capital or technical expertise to invest at scale for needed digital public infrastructures, we see a potential for open-source technologies to offer a viable, cost-effective alternative to the technological infrastructure provided by big tech and/or powerful, well-resourced nations.[6] All this can increase global investment in building local capacity to develop and maintain open infrastructure.
As developments surrounding digital sovereignty continue, we see a need to revisit the (also highly contextual) motivations for the pursuit of open digital research infrastructure. If enabling open and equitable access to and participation in research and scholarship is indeed the goal and motivation for the creation and deployment of this infrastructure, then we recognize that significant investments are needed in building trust in global open infrastructures. If independence and sovereignty regarding data and digital infrastructure are the goals and motivation, then significant efforts are needed to foster collaboration and cooperation in developing connections and networks using open infrastructure that can enable nations to work together while retaining their own domains of control.
References
Abels, K. & Flynn, S. (2024, March 9). Resignation of co-editors of Syntax. LINGUIST List. https://linguistlist.org/issues/35/835/
Bosoer, L. (2022). Digital Sovereignty: Voices from Latin America. https://blogs.eui.eu/latin-american-working-group/digital-sovereignty-voices-from-latin-america/
Bill & Melinda Gates Foundation. (2024). Policy Refresh 2025 Overview. https://gatesfoundationoa.zendesk.com/hc/en-us/articles/24810787662100-Policy-Refresh-2025-Overview
Decklemann, S. (2023). Wikipedia’s value in the age of generative AI. https://medium.com/freely-sharing-the-sum-of-all-knowledge/wikipedias-value-in-the-age-of-generative-ai-b19fec06bbee
EDUCAUSE. (n.d.). Artificial Intelligence (AI). https://library.educause.edu/topics/infrastructure-and-research-technologies/artificial-intelligence-ai
European Commission. (n.d.). A European strategy for data. https://digital-strategy.ec.europa.eu/en/policies/strategy-data
Galápagos Syndrome. (2024, March 31). In Wikipedia. https://en.wikipedia.org/wiki/Gal%C3%A1pagos_syndrome
IT-Planungsrat. (2021). Strategy for strengthening the digital sovereignty of public administration IT: Strategic objectives, approaches and implementation measures. FITKO (Federale IT Cooperation). https://www.it-planungsrat.de/fileadmin/it-planungsrat/foederale-zusammenarbeit/Gremien/AG_Cloud/20210104_Strategy_for_strengthening_the_digital_sovereignty_of_public_administration_IT_1.0.pdf
Komminoth, L. (2023, July 21). “Can Africa Achieve ‘Digital Sovereignty’ in an Era of Big Tech?” African Business. https://african.business/2023/07/technology-information/can-africa-achieve-digital-sovereignty-in-an-era-of-big-tech.
Manifesto on Science as Global Public Good: Noncommercial Open Access. (2023, October 27). Cumbre Global Sobre Accesso Abierto Diamante. Retrieved April 11, 2024, from https://globaldiamantoa.org/manifiesto/#/
Marcus, J. (2021, April 30). Some universities’ response to budget woes: Make faculty teach more courses. The Washington Post. https://www.washingtonpost.com/local/education/college-faculty-course-workload-budgets/2021/04/30/d5d2ee1e-a904-11eb-8c1a-56f0cb4ff3b5_story.html
Nesta. (2023). Future Signals — what we’re watching for in 2023. https://www.nesta.org.uk/feature/future-signals-2023/
Nordhaug, L. M. & Harris, L. (2021). 26. Digital public goods: Enablers of digital sovereignty. In OECD, Development Co-operation Report 2021: Shaping a Just Digital Transformation, OECD Publishing, Paris, https://doi.org/10.1787/ce08832f-en
OASPA. (2021). Major OA Diamond Journals Study completed: Report emphasizes diversity and sustainable pathways for diamond Open Access. https://www.oaspa.org/news/major-oa-diamond-journals-study-completed-report-emphasizes-diversity-and-sustainable-pathways-for-diamond-open-access/
Oransky, I. (2024, February 12). Econ journal board quits en masse because Wiley ‘appeared to emphasize quantity over quality’. Retraction Watch. https://retractionwatch.com/2024/02/12/econ-journal-board-quits-en-masse-because-wiley-appeared-to-emphasize-quantity-over-quality/
Osborne, C., Boehm, M., & Jiménez Santamaría, A. (2023). The European Public Sector Open Source Opportunity: Challenges and Recommendations for Europe’s Open Source Future. https://www.linuxfoundation.org/hubfs/LF%20Research/European%20Public%20Sector%20Open%20Source%20Opportunity%20-%20Report.pdf?hsLang=en
Pfeiffer, N., French, A. L., Vines, T., Gibson, J., & Chandramouli, I. (2024, March 21). Open Science Plumbing: Infrastructure Enabling and Catalyzing Policy Implementation. Retrieved from osf.io/v9hy2
Pinto, R. A. (2018, July 16). “Digital Sovereignty or Digital Colonialism?” Sur - International Journal on Human Rights 27. https://sur.conectas.org/en/digital-sovereignty-or-digital-colonialism/.
Pohle, J., & Thiel, T. (2020). Digital sovereignty. Internet Policy Review, 9(4). https://doi.org/10.14763/2020.4.1532
Ramlochan, Sl. (2023). Openness in Language Models: Open Source vs Open Weights vs Restricted Weights. https://promptengineering.org/llm-open-source-vs-open-weights-vs-restricted-weights/
Saenen, B., Ancion, Z., Borrell-Damián, L., Mounier, P., Oliva Uribe, D., Papp-Le Roy, N., & Rooryck, J. (2024, February 20). 2nd Diamond Access Conference Report. 2nd Diamond Open Access Conference, Toluca, México. https://doi.org/10.5281/zenodo.10684544
Skinner, K. (2019, July 23). Why Are So Many Scholarly Communication Infrastructure Providers Running a Red Queen’s Race? https://educopia.org/red-queens-race/
Weinberg, J. (2023, May 1). More Details on the Dispute Between JPP & Wiley. https://dailynous.com/2023/05/01/the-dispute-between-jpp-wiley/
UNESCO. (2023). UNESCO survey: Less than 10% of schools and universities have formal guidance on AI. https://www.unesco.org/en/articles/unesco-survey-less-10-schools-and-universities-have-formal-guidance-ai
Watkins, E. (2023). Guest Post: AI and Scholarly Publishing - A (Slightly) Hopeful View. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2023/05/10/guest-post-ai-and-scholarly-publishing-a-slightly-hopeful-view/
Feedback
- https://www.oaspa.org/news/equity-in-oa/
- https://diamasproject.eu/
- https://www.coalition-s.org/
- https://www.budapestopenaccessinitiative.org/
- https://www.sovereigntechfund.de/
- We have already seen accelerated investment by big tech giants in Latin America, Asia, and Africa as they strive to expand their user base (Pinto, 2018; Komminoth, 2023).
Find us on