Theo - t3․gg

Your prompts kinda suck

Oct 18, 2026 32 min
prompt engineeringsoftware engineeringllm agentsdevelopment productivity
Watch on YouTube Follow Theo - t3․gg on Rundown — free

Summary

AI summaries can be incomplete or wrong. Verify anything important against the original video.

This video argues that engineers often over-engineer their AI prompts by focusing on strict formatting rather than utilizing AI as a dynamic tool. By treating prompts as conversations and leveraging modern AI features to handle complex tasks, developers can significantly improve their productivity and build more robust applications.

The video challenges the prevalent trend of 'prompt engineering' as a complex, specialized skill. The creator argues that, rather than focusing on precise prompt structures, engineers should treat AI agents as flexible tools that can perform dynamic tasks within a codebase. By stepping back from the urge to micromanage input and output, developers can unlock the true potential of AI agents to handle intricate processes autonomously.

The creator demonstrates this philosophy through practical, hands-on examples, specifically within the context of his own project, T3 Code. He showcases how to use LLMs to automate tasks like code reviews, analyze system performance, and even triage incoming error reports. These examples illustrate a shift in mindset: moving from asking the AI to perform a single, rigid task to giving it the autonomy and data context necessary to execute complex workflows.

Ultimately, the video encourages viewers to shift their focus away from crafting perfect prompts and toward building the necessary infrastructure—such as tools, data pipes, and context windows—that empower AI agents to be more effective. This shift, the creator argues, transforms AI from a basic productivity tool into a powerful extension of an engineer's own capabilities, allowing them to solve problems that were previously impractical to address.

Concepts & takeaways

Locked

Key Points

Locked

Worth watching if: You are a software engineer interested in leveraging AI to optimize your development workflow and build more sophisticated, autonomous agent systems. It is particularly relevant for developers who want to move beyond basic ChatGPT usage to integrate AI directly into their technical projects.

Sign in to unlock the full extract

Every claim, key point, and timestamp for this Theo - t3․gg video — plus a daily email of every channel you follow.

Sign in with Google

No credit card. Free tier forever.

Watch on YouTube