A new artificial intelligence (AI) system, developed by international researchers, is turning heads worldwide for its uncanny ability to predict human decisions with unprecedented accuracy—raising both hopes of revolutionary applications and urgent questions about privacy and ethics. This breakthrough, recently published in the journal Nature, introduces the AI model “Centaur”, which has outperformed decades-old cognitive models in forecasting how people think, learn, and act across diverse scenarios (studyfinds.org).
Centaur’s creators set out with an ambitious goal: develop a single AI system capable of predicting human behaviour in any psychological experiment, regardless of context or complexity. To achieve this, they compiled a massive “Psych-101” dataset spanning 160 types of psychological tests—ranging from memory exercises and risk-taking games to moral and logical dilemmas—amassing data from over 60,000 people and more than 10 million separate decisions. Unlike traditional models tuned for specific tasks, Centaur was trained to generalise, understanding the plain-language descriptions of each experiment.
Built on technology similar to the language models behind tools like ChatGPT, but with specialised enhancements, Centaur’s capabilities were astonishing: when tasked with predicting responses in familiar and even entirely new experimental setups, it displayed an ability to mirror human strategies, capture nuanced patterns of uncertainty, and adapt to unforeseen variations in choice structures. According to the research team, “[t]he human mind is remarkably general… Not only do we routinely make mundane decisions…but we also tackle complex challenges, such as figuring out how to cure cancer or explore outer space.” (Nature, subscription may be required).
What sets Centaur apart is not only its accuracy but also its depth of psychological realism. Upon analysing the AI’s inner workings, neuroscientists found that its information-processing patterns had begun to mirror those observed in the human brain—even though it was not expressly trained to do so. When its internal AI states were compared to neural scans collected while people performed the same psychological tasks, Centaur’s “neural alignment” with the brain was far stronger than previous AI models. This suggests that, in learning to predict our behaviour, the AI even reverse-engineered aspects of how the human mind itself operates.
The implications of such a tool could be transformative for industries like marketing, education, mental health, and product design—any field where anticipating human judgement is valuable. For instance, understanding how people make financial or moral decisions could inform the development of more user-centred online platforms, mental health diagnostic tools that adapt to individual thought patterns, or digital tutors that personalise learning experiences for Thai students. On the other hand, such detailed “mind-reading” powers heighten concerns about surveillance, manipulation, and consent, especially as “digital footprints” become ever more revealing. In Thailand, as in other digital societies, there is growing awareness of data privacy issues—especially with rising AI use in social networks and digital marketing (The Bangkok Post).
Centaur’s training largely drew on information from educated, Western populations, a limitation the researchers acknowledge. As noted in their statement, future expansions will target more globally diverse data—an area highly relevant for Thailand’s multicultural and rapidly digitising society. “We have created a tool that allows us to predict human behaviour in any situation described in natural language—like a virtual laboratory,” lead author and research spokesperson said, promising to release datasets and models publicly so that scholars worldwide—not just industry giants—can refine and test the technology.
This openness could allow Thai academic institutions or government bodies to test Centaur’s abilities using localised datasets—for example, simulating choices within the context of Thai education, local financial behaviours, or traditional community values. For sectors like mental health, where the gap in personalised, early-detection tools is profound, a model that can anticipate risk behaviours or social withdrawal may prove invaluable. Thai experts in cognitive science and ethics urge a balanced approach, emphasising the need for strong, adaptive digital rights frameworks to accompany such innovations. “AI that understands human cognition can be a force for good—but without transparent safeguards, the risks to autonomy and privacy could be significant,” commented a director at a leading Thai research institute.
Historically, Thais have shown adaptability when integrating new technologies, from early adoption of social media and mobile payments to the rapid evolution of online learning during the pandemic (World Bank). However, initiatives that touch on “prediction” or “profiling” in the digital realm are often met with public scrutiny—rooted in Buddhist traditions that emphasise self-awareness and ethical introspection, yet wary of tools that erode individual agency.
Looking forward, Centaur and its derivatives may accelerate scientific discovery by allowing researchers—and perhaps, in future, anyone—to run “virtual experiments” on millions of simulated agents reflecting true-to-life decision-making. The research team themselves used Centaur to spotlight a previously unknown human decision strategy, showing that AI’s value may not just be in copying us, but in expanding the frontiers of psychological knowledge. For Thailand, whose universities and startups are increasingly engaged with global AI research, such openness invites collaboration: adapting, testing, and critiquing new theories with local context in mind.
What should Thai readers do as this AI revolution unfolds? For students and educators, it is a reminder of the value in developing not just information-processing skills but also critical thinking and digital literacy—so Thais can interpret, not just accept, what AI tells us about ourselves. For health professionals and policymakers, caution and curiosity are warranted: careful study is needed to tailor the benefits of predictive AI, while strong rules are needed to protect personal information and prevent misuse. Finally, for anyone navigating Thailand’s digital transformation: stay engaged with science news, participate in public debates on technology and privacy, and support efforts to ensure that AI “mind-reading” always serves—and never subverts—the rights and wellbeing of individuals and communities.
References:
- “AI thinks like humans – with unprecedented accuracy”, StudyFinds, 2 July 2025.
- “AI that predicts human behaviour could help reshape psychology”, Nature, 3 July 2025.
- “Thailand takes steps to shield personal information as AI grows”, The Bangkok Post, 2025.
- “Country Profile: Thailand’s digital transformation”, World Bank.