Ria Viswanathan

About

I have a background in psychology and social science research, and I am currently interested in forecasting, AI’s societal impacts, and the welfare of all beings affected by increasingly capable AI systems—including those using these systems today, and potentially, the systems themselves in the future.

I’m actively thinking about where I can have the most impact, and testing my fit across several areas to figure that out. My current hypothesis is that digital minds/AI welfare research (particularly the societal aspect<sup>1What the public will think about digital minds matters enormously, and it’s far from obvious that their views will track what experts and philosophers conclude. For example, you can imagine scenarios where AI systems intuitively feel sentient to most people, even though experts believe they lack the relevant underlying mechanisms (or the reverse, which would be a huge moral catastrophe!).

Regardless of where the empirical evidence lands, this seems likely to become a major societal issue, and preparing early makes sense.</sup>) could be where I’m best placed to contribute,^{2I hold this with appropriate uncertainty — both about my personal fit and more generally about which cause areas matter most. I am very open to other ideas (or pushback on my current ones) about where I might be best placed to contribute, if you have any!} and I’m actively testing this by working on related projects.

At the same time, I think it’s important to understand what the future could look like more broadly, which is why I’m also working in forecasting.

Research interests: AI societal impacts (and resilience), moral circle expansion, AI welfare, artificial sentience beliefs, animal welfare, human-AI interaction, psychology of effective altruism, institutional decision-making, movement formation/strategy, loss of control risk.

I also created a personal research agenda to track ideas I would like to explore further (I will keep adding to it over time). If you are working on something similar and/or would like to collaborate on any of these ideas, please reach out! Also feel free to leave comments/feedback on the document.