AI is blurring the lines between product, design, and engineering, with non-engineers at companies like Notion now shipping code, accelerating development but risking the loss of deep specialization.
As AI commoditizes technical execution, the most valuable human skills in product development are becoming 'taste' (the ability to predict what will resonate with users) and 'agency' (the initiative to build and change things).
The most successful products are built around a single, exceptionally well-executed 'tiny core' feature (e.g., GitHub's pull request, Figma's real-time collaboration) rather than an accumulation of many features.
The software industry is poised for a shift away from rigid, single-purpose SaaS tools towards more malleable, general-purpose software, as AI makes it easier for users to build their own solutions.
8 quotes
Concerns Raised
The trend of merging roles could lead to a loss of deep, specialized expertise in design and engineering.
While the quantity of software has increased due to AI, the overall quality and reliability have not.
Companies will soon face a difficult reckoning with the return on investment (ROI) of their AI expenditures.
The centralization of power among a few large AI labs is a significant risk if the gap with open-weight models widens.
Opportunities Identified
Empowering non-engineers to ship code can dramatically accelerate product development and innovation cycles.
AI will cause the principles of software engineering to spread to all business domains, automating and optimizing functions like HR and marketing.
Focusing on user empowerment by making existing teams 'superheroes' (like Kubernetes) is a more successful strategy than trying to replace them (like Heroku).
As AI reaches a 'good enough' state for many tasks, companies can differentiate on user experience, speed, and cost-efficiency.