My research examines social cognition across minds and machines.
Human social capacitiesāfrom detecting persuasion to learning from opinionāremain out of reach for even the best artificial intelligence (AI) systems.
Clarifying the computations underlying these social capacities can therefore shed light on human cognition while helping us engineer better AI systems.
The following projects exemplify these dovetailing aims:
How does opinion influence belief?
My research has shown that people can draw rational inferences from othersā opinions (Psych. Sci., 2024), and that societal dissent rarely changes our beliefs on important issues, from abortion to vaccination (PsyArxiv, 2024). Our Paths to Persistence Model explains such robustness through three factors that distill decades of cross-disciplinary research (Nat. Rev. Psy., 2025) and that we can use to explain when interventionsāsuch as persuasive conversations with AI (Arxiv, 2025)āwill succeed or fail.
How can AI systems vigilantly avoid manipulation?
For AI agents to be effective in the real world, they must critically evaluate opinions diverging from their own (Decision, 2024): for example, LLMs should discount claims made in sales pitches, while trusting reliable experts. My research has provided formal models of how such vigilance should operate rationally, and shown that human inferences are aligned with this model (Proc. CogSci, 2024). In ongoing research, we have shown that making salient the key components of this model through prompt-engineering improves LLMsā capacity for vigilance.
In past research, I studied when deeply important beliefsāsuch as whether immigration restrictions are moralācan be changed through courses in moral philosophy (Cognition, 2023); and have studied how people decide how to decide (Cognition, 2022).
I was born and raised in Istanbul (merhaba!); studied economics and cognitive science at Pomona College (CA); and completed my PhD in Psychology at Princeton (NJ). I am presently advancing research alongside a fantastic community of collaborators.
I want science to be more inclusive and rigorous. Here are some resources that can help with that:
Feel free to contact me at oktar[dot]research[at]gmail.com with regards to research / collaboration / mentorship /⦠- I love talking about science. If you would like to send me anonymous feedback, click here.