Bootstrapping a writing language

Jeremy Yuille
→→ ←←
Published in
3 min readJul 30, 2024

Last weekend, I sat down with an idea that had been brewing in my mind for months. By Sunday evening, LMScript was born – a tool that I believe could change the way we interact with AI language models. But LMScript is more than just a set of shorthand commands for LLMs; it’s a statement about the future of human-AI collaboration.

Why LMScript?

For a while now, I’ve been frustrated by the narrative surrounding AI, particularly large language models. There’s this persistent idea that AI is here to replace us, to make human skills obsolete. But that’s not how I see it. I’m more aligned with Marshall McLuhan’s view of technology as an extension of ourselves.

LMScript is my response to this narrative. It’s a tool designed to augment our capabilities, not replace them. With LMScript, I wanted to create something that would help people interact with LLMs more effectively and efficiently, turning these powerful AI models into collaborative partners rather than potential replacements.

Democratizing AI Interaction

One of my core goals with LMScript was accessibility. I wanted to create something that anyone could use, anywhere, without needing coding experience or paid accounts. In a world where AI capabilities are becoming increasingly crucial, I believe it’s essential to level the playing field. LMScript is my small contribution to democratizing AI interaction.

A Performance of Ideas

In many ways, creating LMScript felt like a performance. It was a way of expressing and exploring ideas I’ve been pondering about our relationship with technology. It’s not just a tool; it’s a conceptual piece that questions how we interact with AI and what that interaction could look like in the future.

Collaborating with AI to Create an AI Tool

Perhaps the most meta aspect of developing LMScript was collaborating with AI assistants in the process. I worked with ChatGPT-4 and Claude, using their capabilities to help organize my ideas and extend the concepts I was working with. This experience was a vivid demonstration of the kind of human-AI collaboration I’m aiming to facilitate with LMScript.

Navigating Power Dynamics

During the development process, I had an interesting interaction with Claude that made me reflect on the power dynamics at play. Claude pointed out that there’s a complex relationship between the system I inhabit (as a human developer) and the system the LLM represents. This observation has stuck with me, and I’m still contemplating how to address this in future versions of LMScript. It’s a reminder that as we develop these tools, we need to be mindful of the implicit power structures we might be creating or reinforcing.

Looking Ahead

As I continue to develop and refine LMScript, I see it as more than just a practical tool. It’s a proof of concept for rapid AI tool development, a platform for exploring human-AI collaboration, and hopefully, a catalyst for discussions about our relationship with AI systems.

In this AI-driven world we’re moving into, I believe it’s crucial that we develop tools thoughtfully, always keeping the human element at the center. With LMScript, I’m not just asking what AI can do, but how we want to interact with it and what role we want it to play in our lives and work.

LMScript is my vision of a future where humans and AI are true collaborators, each enhancing the capabilities of the other. It’s an exciting vision, and one that I hope will contribute to shaping the future of human-AI interaction.

As I continue to work on this project, I’m excited to see how others will use and adapt LMScript. After all, it’s released under a Creative Commons license because I believe in the power of community-driven development. I’m looking forward to seeing how this tool evolves and what new ideas it might inspire.

In the end, LMScript is more than just a weekend project. It’s my contribution to an ongoing dialogue about AI, human capability, and the future of work and creativity. I hope it sparks thought, inspires innovation, and maybe, just maybe, helps reshape our relationship with AI for the better.

--

--

Jeremy Yuille
→→ ←←

Principal @WeAreMeld Melbourne. Designer, coach, learner, seeker, finder, explorer.