"Predictable Randomness" or How to Make Forecasts in Uncertainty Age

XV International Conference "Foresight and Science, Technology, and Innovation Policy" took place in Moscow on 12-14 November 2025. Experts from various countries discussed whether classical forecasting methods are truly powerless against "wild cards" predictions and which trends will transform labor market, economy, and our understanding of humanity tomorrow.

©istock

perplexity.ai

At the Session 3 “Global Foresight and Regional Dimension” international experts exchanged global practices for managing futures study in the face of growing uncertainty.

Professor Kuniko Urashima from Nagoya University (Japan) discussed long-term forecasting methods evolution in Japan under pressure of such events as "wild cards".

Japan was a pioneer in science and technology Foresight, and over its half-century history, it has evolved from a narrowly focused technological forecast to a comprehensive analysis that takes into account long-term impact of technology on society and the economy.

Initially, the primary method was a Delphi survey of experts. However, over time, the Foresight focus has shifted from predicting specific technologies to shaping a vision of the future, and now studies involves not only scientists but also representatives of industry, politics, culture, and ordinary citizens.

Since 2018, the Foresight process in Japan has been structured in several stages: from horizon scanning and creating future visions to scenario planning and developing specific recommendations.

Professor Urashima paid special attention to factor of unpredictability. According to her, modern Foresight must consider possibility of "wildcards events"— hardly predictable but highly influential phenomena. The speaker identified these as breakthroughs in AI and quantum computing, environmental disasters, geopolitical conflicts, and even shifts in philosophical paradigms.

A striking example of "wild cards" impact is COVID-19 pandemic. As Delphi surveys showed, after the pandemic, importance of topics related to security, digitalization, and societal resilience increased sharply.

Designing for the future in an uncertain environment requires embracing unpredictability, creating adaptive systems, fostering interdisciplinary collaboration, and investing not only in efficiency but also in resilience Professor Urashima concluded.

Professor Ozcan Saritas from the Rochester Institute of Technology (UAE) and HSE University (Russia) focused on technological turbulence (“The Future’s Coaster: Peaks of Progress & Drops of Disruption”). While the world is becoming more complex and society struggles with one set of problems, technology is creating new risks, triggering a cascade of unforeseen consequences. Recently a key challenge is control over AI. The success of human control over artificial intelligence will determine what the world will be like by 2050: doubling economy and beginning of Mars colonization, or a society in which reasoning machines take away jobs from people, and AI algorithms, using synthetic narratives, actively intervene in everyday life, changing and imposing artificial values ​​on people. But the most likely scenario is one in which peaks of progress, supported by international cooperation and ethical governance of AI, are followed by devastating declines, conflicts over resources, an erosion of trust in science and institutions, and a split in society between “enhanced” and ordinary people. Expert is convinced that human control over AI reduces the risk of negative scenarios.

Foresight projects in Malaysia are considered as a key tool for identifying future threats and promising opportunities.  Malaysian Industry-Government Group for High Technology (MIGHT) is operating as an agency for strategic Foresight under auspices of the Ministry of Science, Technology, and Innovation.

MIGHT director Mr. Mohd Nurul Azammi Mohd Nudri explained how his agency identifies upcoming challenges in healthcare and public safety, reduces the vulnerability of society and economy to consequences of increasing political turbulence and climate disasters threat using Foresight. The expert emphasized six factors, identified through Foresight, that shape the country's strategic priorities: achieving efficiency and fiscal sustainability, demand for justice, ensuring sustainable development, changing citizen expectations, technological race, and ensuring sovereignty.

Alexey Mitkin, Director of Strategic Programmes and Innovative Development at PJSC ‘Rostelecom’ shared the company's approach to strategic planning through monitoring of global trends which combines qualitative methods and big data analysis. To achieve this, the company uses specialized software with an integrated AI module. ‘Rostelecom’ has been monitoring and analyzing global trends since 2017, using a database consisting 54 million patents and 80 million scientific publications. To demonstrate how trends and priorities are changing, speaker compared trend ratings for 2017 and 2025. Thus, mobile networks, which occupied the rating first line in 2017, ended up in third place in 2025, AI moved from second to third place, e-health technologies, on the contrary, rose from third to second place, and trend of information security jumped from seventh to fourth place. The expert believes that such ratings make it possible to track the dynamics of trends and the growth or decline of interest in certain technologies. Such ratings make it possible to track trends dynamics and growth or decline of interest in certain technologies, the expert believes. Among the fastest-growing trends in 2025, the speaker noted head-mounted displays (+77% in news), drones (+58% in scientific articles), augmented reality technologies (+56% in scientific articles), data center networks (+38% in news), wearable computers (+35% in news), brain-computer interfaces (+23% in patents) and human behavior analysis technologies (+14% in patents). ‘Rostelecom’ also identifies weak signals. This year, these include vulnerabilities in quantum cryptography (quantum hacking), distributed quantum computing, nanotechnology for targeted drug delivery, parasocial (one-sided) relationships, AI hallucinations, and explainable AI.

Summing up the session, Professor Alexander Chulok noted that Foresight studies should not rely on a single method; a combination of methods is needed, determined by the research objectives. Furthermore, it is important to consider socioeconomic factors, not just technological ones.