AI’s computational power and its ability to recognize patterns across vast amounts of data make it well-suited for understanding and predicting human behavior. This can provide commercial advantages, but the field is highly controversial and characterized by high entry barriers.
This blog post is the fifth in a series on AI’s disruption of innovation
This blog post is the fifth in a series exploring the disruptive impact of a technology based on a new form of intelligence that is self-learning, universally enabling, and allows for deep customization:
- Where does AI impact today?
- What is new and disruptive about AI?
- AI’s effect on robotics and experimental automation
- AI’s effect on climate and meteorological research
- AIs impact on social sciences
- AI’s effect on teaching, learning, and psychotherapy
- AI’s impact on biotechnology and pharmaceuticals
- AI’s effect on materials research and quantum chemistry
- AI’s effect on theoretical physics and mathematics
- What should investment strategies in AI take into account?
The best predictor has the greatest competitive advantage
Investment is built on prediction. Whoever can best anticipate the market can achieve the highest returns. Similarly, the movements of financial markets reflect the events and trends that the majority of market participants believe in. This applies to both financial behavior and broader patterns of consumer behavior. For this reason, simulations of behavior and sociological patterns hold potentially great financial and commercial value..
But decoding and then predicting the behavior of millions of actors requires enormous amounts of foundational data. It also requires immense computational power to integrate these data and extract structural insights. The real value-creating element in this process is therefore the development of new methods for analyzing, interpreting, and simulating social and cultural phenomena and interactions.
AI’s pattern recognition across massive datasets creates strong opportunities
A core characteristic of AI is that overlaps and digital processing can occur in a vector-based way rather than relationally. This is akin to how the human brain seeks patterns rather than functioning deterministically according to strict 1:1 relationships. For this reason, AI is a particularly well-suited catalyst for social interactions. Examples include:
Analyzing large psychometric datasets or social media for early signals of commercial trends and tendencies, including through agent-based simulations of social interactions. More controversially, identifying early political or security-related signs of widespread mis- or disinformation.
Moreover, AI’s characteristics open up possibilities for interdisciplinary research. This includes, for example, a broader sociological understanding of how norms and ethics evolve over time and under what circumstances. Such understandings are of great importance for the development of teaching and communication methods.
… which becomes even stronger when AI can also hyper-personalize communication
AI has proven to be particularly well-suited for hyper-personalizing dissemination and communication. This enables, among other things, its use in psychotherapy and emotional support (see the next blog post). But in addition, AI has also shown itself capable of predicting people.
In July of this year, for example, a group of researchers from the Institute for Human-Centered AI published in Nature that they had created a cognitive “foundation model” called Centaur.
- Centaur was trained on Psych-101, a comprehensive dataset containing information from more than 60,000 people who made millions of decisions across 160 experiments.
The researchers found that the model was able to outperform all previous models for predicting human behavior — even in scenarios not seen before. The project thus suggests how AI may be able to anticipate our everyday choices, the deeper the access we grant it to our thoughts and decisions.
For this provides a basis for persuasion, something history has many examples of
Back in 2018, researchers from Oxford, Cambridge, the Future of Humanity Institute, and OpenAI outlined a number of scenarios. They focused on how AI could be misused for malicious purposes such as hacking or broader cyber-offense, including politically motivated disinformation. These remain just as relevant today.
At that time, the scandal surrounding Cambridge Analytica (2016–2018) was still fresh in memory. Here, advanced models built on psychometric data had made it possible to hyper-personalize political messaging (e.g., using OCEAN personality profiles). It is naïve to believe that such use of AI could only occur in the U.S. or only by one political party; it was simply there that the first structured application was uncovered.
But it is such a complex field that only few have the critical mass to pursue it
The area is complex and requires the use of as many complex datasets as possible. This creates high entry barriers, and therefore the field generally suits Big Tech best. Google DeepMind and the LLM providers OpenAI and Meta are particularly focused on analyses of discourse and interaction — that is, information diffusion. In addition, Microsoft and IBM are both developing models to simulate social scenarios.