Luca Garzino Demo

The way we're thinking about generative AI right now is fundamentally individual. We see this not just in how users interact with models but also in how models are built, how they're benchmarked, and how commercial and research strategies using AI are defined. We argue that we should abandon this approach if we're hoping for AI to support groundbreaking innovation and scientific discovery. Drawing on research and formal results in complex systems, organizational behavior, and philosophy of science, we show why we should expect deep intellectual breakthroughs to come from epistemically diverse groups of AI agents working together rather than singular superintelligent agents. Having a diverse team broadens the search for solutions, delays premature consensus, and allows for the pursuit of unconventional approaches. Developing diverse AI teams also addresses AI critics' concerns that current models are constrained by past data and lack the creative insight required for innovation. The upshot, we argue, is that the future of transformative transformer-based AI is fundamentally many, not one.

Preregistration is designed as a solution to questionable research practices, safeguarding the epistemic integrity of scientific work. Despite its benefits, concerns remain that it might inadvertently suppress creativity and exploration, diminishing epistemic diversity. To engage with this debate, I employ an agent-based model to analyze how different preregistration regimes influence scientific communities. I argue that the coexistence of different research types is heavily dependent on the existing reward structures in academia, rather than on preregistration itself. Interestingly, I find that universally mandatory preregistration, compared to mere incentivization, promotes a more balanced coexistence of research methodologies, contingent upon how rewards are distributed between confirmatory and exploratory research. In light of arguments and results, I propose practical improvements for implementing preregistration, including a two-tier publication system for exploratory research and the adoption of registered reports. This approach aims to foster an environment where both diversity and integrity can thrive, addressing the underlying institutional biases.