AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Writing serves as a means to communicate ideas, influence actions, and enact change. It allows us to attract attention, widen the space of communication, and share our perspectives with others. Effective writing can persuade, inspire, and engage readers, shaping their understanding of the world. It is a way to convey complex concepts, observations, and insights, whether it is to inform, entertain, or provoke thought. By writing, we participate in a larger conversation, contributing to the collective knowledge and understanding of society. However, it is important to recognize that writing is not merely about expressing ourselves, but also about connecting with readers and adapting our message to their perspective and context.
Writing involves capturing and conveying the essence of reality and personal experience. It is a way to make ideas and experiences more tangible and communicable. Through writing, we strive to make the world feel more real, whether by describing vivid sensory details, capturing the essence of a particular event or scene, or engaging readers with personal insights and observations. Writing is not just about abstract concepts, but about grounding them in lived experiences and the physical world. However, this embodied knowledge and connection to reality may be difficult to replicate for language models, as they lack the embodied understanding and personal perspective that humans possess.
Writing is deeply intertwined with cultural context, social understanding, and consensus. It is through writing that we both gain new perspectives and seek to influence others. Shared cultural references, jargon, and values shape our writing and how we perceive and interpret it. At the same time, writing helps us enter the perspectives of others, expanding our understanding and fostering social consensus. However, language models may struggle to capture this contextual richness and the diverse cultural understanding that humans bring to writing. Their dependence on training data, which is often biased and limited, raises concerns about representation, accuracy, and the preservation of cultural perspectives and diversity in the writing process.
Writing is deeply connected to cognitive processes such as perception, action, and embodied knowledge. It is a reflection of our complex interactions with the world, where perception and action are intertwined. Writing represents an extension of our thoughts, experiences, and understanding that allows us to communicate and reflect upon them. Language models, although capable of generating text, lack the lived experiences and embodied understanding that shape human cognition and the writing process. The inherent limitations of language models in perceiving and interacting with the physical world may limit their ability to fully capture the nuances and richness of human cognitive processes in writing.
Language models, like AI-driven algorithms, are becoming an integral part of our lives, filtering and organizing information to help us function in the future. While there are concerns about privacy and security, personalized AI trained on individual users seems to be the direction we're heading. Understanding the limits of language models in their cultural context is important, as humans are inherently embedded in their cultural surroundings, unable to function outside of them. The concept of the embodied mind challenges the idea of rationality as a separate entity from the body, emphasizing the interdependence of the entire system. Language itself is influenced by embodiment and culture, making physical references to convey even abstract concepts. Activism and autopoiesis theories emphasize the importance of engagement and activity in learning and understanding. While language models can augment certain tasks, it is crucial to recognize that they lack the embodied experience and sensory perception humans possess.
Action and perception are deeply linked, and language models cannot fully understand this connection. Humans go beyond writing to take action based on their experiences and beliefs. Language models, even with their advanced capabilities, cannot truly experience the world through perception and interaction. They lack the 3D sensual depth and embodied understanding that humans possess. While some argue that language models can replace certain tasks, such as preserving and presenting condensed information, their inability to grasp the complexities of the real world limits their effectiveness. Writing is a difficult translation task, bridging the gap between experience, thought, and language. Language models should be used as tools to supplement human activity, not as substitutes for it, and products should be designed with a human-in-the-loop approach for a balanced future.
This is a continuation of the conversation I had with Maggie previously: "The Dark Forest and Generative AI." We discuss why we write, versus why to use AI and LLMs to produce writing, and what effects that loss is likely to have.
Discussed in this episode:
Clerestory by Bryan Kam • Infrequent updates at Substack • All my work plus exclusive content at Patreon
Show notes https://pod.fo/e/1720c4
Listen to all your favourite podcasts with AI-powered features
Listen to the best highlights from the podcasts you love and dive into the full episode
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
Listen to all your favourite podcasts with AI-powered features
Listen to the best highlights from the podcasts you love and dive into the full episode