We’re launching Lyria 3.5 in Google Flow Music, with advances across musicality, lyrics, vocals, and creative control
deepmind.google·6d ago
AI Engineering
Lobsters·17d ago·5 min read
Software and AI - Plotting vs. Pantsing
TL;DR
A divide exists in AI usage among software engineers, influenced by their approach to problem-solving. Some prefer 'plotting,' where solutions are planned before coding, while others favor 'pantsing,' where coding drives understanding.
✦ Why It Matters
Engineers should assess their coding style to choose AI tools that align with their problem-solving approach.
Key Takeaways
- Identify if your current work is plotting or pantsing.
- Use AI assistants to enhance understanding in existing codebases.
- Leverage AI delegates for efficiency in greenfield projects.
- Align AI tools with your development approach for optimal results.
Full Summary
In software development, 'plotting' refers to a top-down approach where solutions are planned before coding, while 'pantsing' is a bottom-up method where understanding evolves through coding. The article discusses how these styles affect the use of AI, distinguishing between 'delegates' that autonomously edit code and 'assistants' that support human developers.
Many developers face challenges when working with existing codebases, often requiring a pantsing approach to build understanding. In contrast, greenfield projects allow for more plotting as there are fewer constraints.
The author argues that using the appropriate AI style can enhance understanding and efficiency, suggesting that AI should complement the developer's current approach rather than dictate it.