Project summary

We are seeking funding for a 6-month pilot to test whether probabilistic forecasting can improve public reasoning and preparedness around AI governance, misinformation, institutional trust, and social risk in Greece.

Greece is a useful pilot case for this work. It is a small European democracy that has experienced repeated economic, social, institutional, environmental, and political shocks over the last fifteen years, but it currently lacks public infrastructure for systematic, accountable forecasting. There are polls, political commentary, academic research, and policy analysis, but no visible mechanism for turning claims about the future into clear forecasting questions, assigning probabilities, tracking updates, and evaluating accuracy over time.

This project will build a first version of that infrastructure at a modest scale.

We will design 40–50 forecasting questions about Greece, recruit an initial community of forecasters, involve a small group of domain experts, publish a simple public forecast tracker, and produce a final report on whether this model could be expanded into a permanent Hellenic forecasting lab.

The broader question behind the project is not only “what might happen in Greece?” but whether small and medium-sized democracies can build low-cost forecasting capacity to reason better about emerging risks, especially around AI, misinformation, public trust, and institutional resilience.

What are this project's goals? How will you achieve them?

The main goal is to run a focused, 6-month proof of concept for public-interest probabilistic forecasting in Greece.

We want to test whether a small research team can:

1. Convert vague public concerns into clear, time-bounded forecasting questions.
2. Recruit a mixed group of forecasters, including researchers, analysts, students, civil society actors, journalists, and interested citizens.
3. Involve domain experts in AI, public opinion, media, politics, technology, and civil society.
4. Track forecasts over time rather than collecting one-off opinions.
5. Publish the results in a transparent and accessible way.
6. Evaluate the quality of the forecasting process and identify whether the model should be expanded.

The pilot will focus on questions in five broad domains:

- AI adoption and AI governance in Greece
- AI-generated misinformation and information integrity
- Institutional trust and democratic resilience
- Social mobilisation and public opinion shifts
- Climate, energy, and social risk

Examples of forecasting questions we may include:

- What is the probability that at least one major AI-generated misinformation incident will receive national media attention in Greece before the end of 2027?
- What is the probability that a Greek public institution will issue formal guidance on generative AI use before the end of 2027?
- What is the probability that reported use of generative AI tools by Greek businesses will exceed a specified threshold by the end of 2027?
- What is the probability that trust in a major Greek institution will fall below a specified threshold in at least two national surveys by the end of 2027?
- What is the probability that Greece will experience at least one large-scale mobilisation around housing, cost of living, climate, or institutional accountability before the end of 2027?

The project will proceed in five stages.

First, we will design the forecasting question set. Each question will have a clear resolution criterion, a time horizon, and an identified source or method for determining the outcome.

Second, we will recruit participants and experts. We aim to involve at least 100 forecasters and 10–15 domain experts during the pilot.

Third, we will collect and update forecasts. Participants will be asked to assign probabilities, provide short reasoning, and update their forecasts when relevant new information appears.

Fourth, we will publish a simple public forecast tracker. This does not need to be a complex platform at the pilot stage. A lightweight dashboard or public tracker is sufficient for testing the model.

Fifth, we will produce a final report in English. The report will summarize the forecasting questions, aggregate forecasts, participation data, methodological lessons, and recommendations for whether and how to continue the project.

How will this funding be used?

We are requesting funding to cover the core research, coordination, technical, and reporting work required for a 6-month pilot.

Indicative use of funds:

- Research design and forecasting question development
- Recruitment and coordination of forecasters
- Expert outreach and interviews
- Data collection and background research
- Survey and forecasting workflow setup
- Development of a simple public forecast tracker or dashboard
- Analysis of forecasts and forecast updates
- Methodology note on question design, resolution criteria, and scoring
- Two short interim updates
- One final public report
- Basic administration, communication, and project management

The funding would primarily support researcher time. The project is labor-intensive: the main work is designing good questions, recruiting the right participants, maintaining the forecasting process, checking resolution criteria, communicating with forecasters, and writing up the results.

We are intentionally proposing a modest pilot rather than a full institution. The goal is to produce a clear proof of concept before seeking larger funding or building a permanent organization.

Who is on your team? What's your track record on similar projects?

The project is led by two Greece-based researchers with complementary backgrounds in social research, political analysis, survey design, and data analytics.

The first team member is a PhD candidate in Political Science at Aristotle University of Thessaloniki, with research interests in political sociology, social movements, social and solidarity economy, and sustainable development. His experience includes qualitative and quantitative research, questionnaire design, data collection, interviews, social movement research, adult education, and academic publishing.

The second team member is a research professional with experience in political research, market research, sentiment analysis, big data analytics, survey design, statistical analysis, and client-facing research reports. His technical skills include Python, SPSS, and Power BI, and he has worked on political, social, and market research projects, including public opinion and campaign-related research.

Together, the team combines:

- Political sociology and social research
- Survey design and questionnaire development
- Qualitative and quantitative methods
- Public opinion and political research
- Sentiment analysis and big data analytics
- Statistical analysis and data visualization
- Experience producing research reports for academic, political, social, and institutional audiences

We have not previously run a public forecasting tournament or forecasting lab. This is why we are proposing a bounded pilot rather than claiming to have an established institution. However, our existing experience is directly relevant to the first version of this project: designing research questions, collecting and analyzing data, interpreting social and political developments, and communicating findings clearly.

We will also seek feedback from people with experience in forecasting methodology, AI governance, public opinion research, and democratic resilience during the pilot.


What are the most likely causes and outcomes if this project fails?

The most likely failure mode is that we do not recruit enough active forecasters or expert contributors to make the pilot useful. To reduce this risk, we will start with a modest target of around 100 forecasters and 10–15 experts, rather than trying to build a large community immediately.

A second risk is that the forecasting questions are not well designed. Poorly specified questions would make the results less useful. To mitigate this, we will focus heavily on question design, resolution criteria, and external feedback before launching the full question set.

A third risk is that some questions may take too long to resolve. Many important social and political questions have long time horizons. We will therefore include a mix of shorter-term and longer-term questions, and the final report will focus not only on resolved outcomes but also on process quality, participation, forecast updating, and methodological lessons.

A fourth risk is that the project is perceived as local Greek political analysis rather than as a forecasting and epistemic infrastructure pilot with broader relevance. We will address this by publishing the final report in English and framing Greece as a case study for small and medium-sized democracies facing similar challenges around AI, misinformation, institutional trust, and social risk.

If the project fails to become a permanent lab, we still expect useful outputs: a public set of forecasting questions, a methodology note, a small network of interested forecasters and experts, and a final report on what worked and what did not. This would still be valuable for future forecasting efforts in Greece or similar contexts.

How much money have you raised in the last 12 months, and from where?

We have not raised dedicated funding for this project in the last 12 months.

This is a pre-institutional pilot led by an early-stage research team. We are applying for initial seed funding to test whether the model is worth expanding.

If the pilot is successful, we may later seek additional funding through forecasting-focused grants, AI governance funders, European research programs, university partnerships, or the creation of a dedicated nonprofit entity in Greece.