Beyond Clarity Workshop 2025 Proposal

Capital B with water inside

Natural language is fundamentally discrete. Language draws hard boundaries. Language structures how we consciously conceptualize and perceive the world. However, language is inherently limited, with discrete words, concepts, and compositions that are unable to capture the full richness of the world, sensation, perception, thoughts, and beyond.

Neuroscience offers insights into how the brain forms discrete or continuous percepts and memory from sensory signals. Discrete representation of continuous quantities such as space, color, duration, and intensity, often incurs quantization error and loss of information. Could we and should we rid ourselves from the binding of discrete low-level abstractions and linguistic concepts?

Moreover, recent developments in large language models (LLMs) introduce continuous vector space representations that transcend the discreteness in languages. We explore the interplay between these domains, asking:

This workshop aims to bring together deep thinkers to lead discussions on these questions, providing space for open exchange across neuroscience, artificial intelligence, philosophy, art, mathematics, and more. We aim to foster new insights that transcend the boundaries.


Discussion Leaders

Participants