Conformity to popular, not average, opinions: Models, data, and evolution

A free platform for explaining your research in plain language, and managing how you communicate around it – so you can understand how best to increase its impact.

link.growkudos.com

Model redefining conformity excels against real-world data

In a new paper in PNAS, SFI Complexity Postdoctoral Fellow Kaleda Denton, SFI External Faculty Fellow, and Science Board Fellow Marcus Feldman, and independent researcher Jonathan F. Johannemann offer...

santafe.edu

Kaleda K. Denton - Cats

Pictures of my foster cats and kittens! For more information about adopting and fostering, feel free to reach out to me at kdenton@ucla.edu

kaledadenton.com

SFI welcomes Complexity Postdoctoral Fellow Maike Morrison

Complexity Postdoctoral Fellow Maike Morrison builds mathematical tools to quantify and compare biological variation. Her work draws on ecology, population genetics, and information theory to study th...

santafe.edu

docs.google.com

456: Kaleda Denton | Following the Majority - How Conformity Shapes Culture and AI

In this episode of the Armen Show, Armen Shirvanian interviews Dr. Kaleda Denton, a post-doctoral fellow at the Santa Fe Institute. They discuss the interdisciplinary nature of the Institute, the concept of niche construction in biology, and the dynamics of decision-making, particularly the differences between following the mean versus the majority. The conversation also explores the limitations of game theory, the role of cultural evolution in cooperation, and the impact of AI on accessibility and learning. They delve into the importance of reducing polarization while maintaining diversity in viewpoints, emphasizing the need for mixing and collaboration across different fields and perspectives. In this engaging conversation, Dr. Kaleda Denton and Armen Shirvanian explore a variety of topics including the distinction between majority and mean opinions, the challenges of reading in the digital age, and the implications of large language models (LLMs) on culture. They discuss the importance of communication in society, the influence of close circles versus wider communities in learning, and the role of influential figures in personal development. The conversation also delves into the concept of memes and cultural transmission, the impact of role models on cumulative culture, and the question of human uniqueness in the context of cultural evolution. Check out Kaleda's website at https://www.kaledadenton.com/ 00:00 Introduction to the Podcast and Guest 03:01 The Santa Fe Institute: A Unique Interdisciplinary Environment 05:56 Niche Construction and Extended Evolutionary Synthesis 08:56 Mean vs. Majority: Insights from Research 12:08 Game Theory and Its Limitations 15:03 Cultural Evolution and Cooperation 17:59 The Role of AI in Learning and Accessibility 20:54 Conformity Across Species and LLMs 23:59 The Importance of Current Information 27:03 Final Thoughts on AI and Human Interaction 36:57 The Quirks of AI Hallucinations 38:53 The Evolution of Language Models 39:35 Agency and Individuality in AI Interactions 41:41 The Flattening of Ideas and Language 43:42 Cognitive Offloading and Its Consequences 45:04 Reducing Polarization Through Interaction 48:39 The Importance of Mixing Perspectives 50:02 The Dangers of Majority Influence 53:51 Understanding Mean vs. Majority Opinions 01:04:32 Risks and Opportunities in Cultural Evolution 01:10:02 The Impact of Headlines on News Consumption 01:11:46 The Need for Positive News 01:15:04 Learning from Close Circles vs. Wider Communities 01:17:23 Influential Figures in Personal Development 01:18:44 The Evolution of Human Uniqueness 01:22:38 Understanding Memes and Cultural Transmission 01:23:10 The Multiplicative Costs of Altruism 01:36:09 The Role of Role Models in Cumulative Culture

tinyurl.com

nationalacademies.org

Do AI Models Perform Human-like Abstract Reasoning Across Modalities?

OpenAI's o3-preview reasoning model exceeded human accuracy on the ARC-AGI benchmark, but does that mean state-of-the-art models recognize and reason with the abstractions that the task creators inten...

arxiv.org