Don't view AI as just a feature set. Instead, treat "intelligence" as a fundamental new building block for software, on par with established primitives like databases or APIs. When conceptualizing any new product, assume this intelligence layer is a non-negotiable part of the technology stack to solve user problems effectively.

Instead of chasing the latest hyped AI model, focus on building modular, system-based workflows. This allows you to easily plug in new, better models as they are released, instantly upgrading your capabilities without having to start over.

Don't just sprinkle AI features onto your existing product ('AI at the edge'). Transformative companies rethink workflows and shrink their old codebase, making the LLM a core part of the solution. This is about re-architecting the solution from the ground up, not just enhancing it.

Instead of codebases becoming harder to manage over time, use an AI agent to create a "compounding engineering" system. Codify learnings from each feature build—successful plans, bug fixes, tests—back into the agent's prompts and tools, making future development faster and easier.

Railway encourages its team to use AI not just for coding but to build massive test benches and prototypes of future product concepts. This allows them to validate complex ideas for free, accelerate learning, and in some cases, skip incremental roadmap items to build the final vision sooner.

Instead of asking an AI for a one-off task, identify recurring workflows and have the AI turn them into a "skill." This creates a reusable asset that dramatically improves efficiency and output quality over time, turning the user into a system builder.

The initial rush to adopt AI resulted in superficial features like text rephrasing tools. That era is over. The next, more valuable phase of AI product development requires creatively embedding AI's reasoning capabilities into core product workflows, moving beyond simple generative tasks to create genuine, contextual automation.

The 'compound startup' model, building a broad suite of integrated products, is now supercharged by AI. Because AI makes building software 10x faster, companies can and should pursue extreme product breadth to create a single, unified platform that customers prefer over siloed point solutions.

Instead of building a single-purpose application (first-order thinking), successful AI product strategy involves creating platforms that enable users to build their own solutions (second-order thinking). This approach targets a much larger opportunity by empowering users to create custom workflows.