Current chat interfaces are compared to the command-line: they require users to learn a specific, procedural way of communicating ('prompt engineering'). New interaction models, which allow for natural, multimodal communication, could be AI's 'GUI moment,' democratizing access by letting users focus on the task, not the tool.

Simply offering the latest model is no longer a competitive advantage. True value is created in the system built around the model—the system prompts, tools, and overall scaffolding. This 'harness' is what optimizes a model's performance for specific tasks and delivers a superior user experience.

The challenge in using AI effectively is often prompt engineering, not model capability. A potential solution is a social platform where users can follow experts, discover their prompts, and be 'catalyzed' by others' creativity. This democratizes access to AI's full potential beyond one's own ingenuity.

Complex prompting is a transitional phase for AI interaction, not the end state. Truly useful AI tools will abstract this complexity away, using agents to translate user intent into optimal prompts. The focus should be on creating intuitive, directorial controls rather than teaching users to be prompt engineers.

Current AI hits like ChatGPT are "single-player" tools. The next breakthrough will be a social, multiplayer experience built on network effects and user-generated content, an area where technically-focused AI labs are culturally weak.

Formal, top-down AI integration protocols like MCP failed due to inefficiency and high context usage. The more successful approach was the bottom-up, community-driven emergence of 'Skills'—shareable, specific prompts that reflect how people were already organically using the technology.

Just as you use different social media apps for different purposes, you should use various specialized AI tools for specific tasks. Relying on a single tool like ChatGPT for everything results in watered-down solutions. A better approach is to build a toolkit, matching the right AI to the right problem.

The current back-and-forth prompting model is a "product overhang" that limits AI's potential. The future lies in giving agents a high-level goal, access to tools and data, and letting them run for extended periods to figure out the execution details, functioning more like an autonomous employee than a simple tool.

With top models reaching comparable performance, differentiation is moving to the "harness"—the user interface, tool integrations, and agentic workflows. OpenAI's ChatGPT Work, an extension of its Codecs interface to general knowledge work, shows that the system surrounding the model is now as crucial as the model itself for user adoption and value.