We usually define the new frontier of knowledge as everything related to the development of Artificial Intelligence and the multiplying possibilities unlocked by supercomputing systems. However, there is a significant imbalance between the efforts devoted to applying these tools to technical problems and commercial developments, and the possibilities they offer for addressing complex societal challenges (socio-ecological transitions, the transformation of the care system, tackling inequality, etc.).
A good example of this disconnect is the growing investments in “digital twins”. The Barcelona Supercomputing Centre, for instance, is working on a macro-project to enable the planet Earth to be simulated computationally. This digital twin, which combines climate models and supercomputing, will make it possible to simulate terrestrial scenarios for 50 or 100 years from now with greater accuracy than current models. By simulating potential scenarios, all relevant stakeholders can take more innovative decisions informed by multiple layers of data.
These same technologies are being used in medicine and in mobility or urban planning supported by the EU and other governments at large scale. The missing gap is that the tech teams leading these simulations consider it much more difficult, or practically impossible, to simulate social systems in which one of the fundamental variables is human behavior.
This is, therefore, the true frontier of knowledge and the greatest opportunity for social innovation in the near future. Our community has the necessary knowledge and skills to partner with AI and language experts and supercomputing infrastructures to learn how to simulate socio-technical scenarios. These tools will allow us to combine real experimentation with digital simulation and feed the algorithms with data from the social innovation process. We just need to overcome the traditional skepticism towards digital solutions and incorporate a robust ethical framework to manage data.
According to Google and Microsoft researchers (Rida Qadri, Michael Madaio, and Mary L. Gray), we must focus on “unpacking the social nature of data and the social scientific work it takes to develop culturally inclusive training datasets”. We need new methods for data annotation work that feature interpretation and deliberation of data’s social meanings. This is the new scientific adventure for “modeling social dynamism”. They wonder what might design paradigms look like that could support the polyvocality and multiple perspectives inherent in our complex, heterogeneous social world. In their opinion, these are novel approaches to dataset schema that might encode multiple narratives into each data object. Instead of giving a single answer about what a good care solution is, it invites all relevant stakeholders and citizenry to clarify what kinds of care solution are possible and relevant for their communities. “Instead of relying on singular datasets that supposedly capture global cultural values, such interfaces might become a supporting partner in a journey of cultural discovery” and social innovation.
Existing computational models allow us to work with multiple scenarios and visualise different variables and combinations of a given issue. Collaboratively interpreting (scientific agents, public institutions, and social organisations) the impact of one public policy investment or another, or the potential responses of different perceptual patterns to a given measure, can become essential tools when making innovative decisions. The scientific knowledge is ready, democratic institutions desperately need it to reduce the risk of innovation policies, but the social innovation community must provide the necessary interface to connect the data with real experimentation to better inform the decision-making process. International consultancies will charge ridiculous fares offering how to jump directly from data analysis to decision making but these models will fail if there is no experimentation correcting the model in real time.
Let’s take a real example: collectively simulating and interpreting the positive and negative impacts of deploying a new innovative policy to tackle homelessness, such as Housing First, in a city like Malmö. The objective would be (1) to compile all existing data to better understand all dimensions of this issue, including the mapping of the variety of actors involved and their interconnections; (2) to understand all the social perceptions that exist on this issue in Malmo; and (3) to simulate what would happen if we developed different implementation modalities of Housing First in the city that could even be contradictory to each other. The added value of working in this way lies in the possibility of analysing and simulating different options while introducing a portfolio of experimentation running in parallel. The city of Malmö could collectively and openly interpret the impact of offering smaller or larger homes with different types of services, as well as compare their advantages and disadvantages with the interventions currently being developed. This process would allow all the relevant stakeholders and the public to participate in the sensemaking process and agree on a new portfolio of interconnected interventions co-created in a much more consensual manner and with greater scientific rigor. Working in this way would also strengthen the social innovation capabilities to prove the evidence of the intervention and to compare it with potential randomized control scenarios.
This example is just one possibility. These tools could be applied to all social innovation interventions aiming at producing systems change. Those who learn how to connect their supercomputing and artificial intelligence capabilities with social innovation will soon lead the way to tackle complex societal challenges. All require comprehensive approaches and solutions that have the sustainability of the planet and the well-being of people in all its dimensions as their fundamental objective. Unfortunately, we still lack the necessary infrastructure to collectively address them. There are strong social, political, business and institutional networks operating globally, but they were not designed to address complex problems collaboratively in a digital age.To achieve this, we need to build new capacities by investing in dedicated AI and supercomputing infrastructure for social innovation and to create social data catalogues that collect, annotate and integrate diverse social datasets with cultural context and multiple interpretations.
We normally think that it is a question of adaptation, of incremental innovation. Changing little by little until we reach a better scenario. However, the most pressing societal challenges call for a more disruptive evolution of the social innovation community. Whether we are capable of developing our own infrastructure in the field of supercomputing and artificial intelligence might determine the value that social innovation can bring to society in the near future.
