> Markdown version of [/videos/100136-designing-high-quality-software-with-ai?t=1716](https://www.wearedevelopers.com/videos/100136-designing-high-quality-software-with-ai?t=1716). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Designing high quality software with AI

Does AI make developers lazy and software mediocre? Discover how Notion leverages Next.js playgrounds and rigorous verification loops to ship ambitious, high-quality code.

- **Speakers:** [Brian Lovin](https://www.wearedevelopers.com/@brian-lovin)
- **Event:** World Congress 2026 Europe
- **Published:** July 9, 2026
- **Duration:** 30:46
- **URL:** https://www.wearedevelopers.com/videos/100136-designing-high-quality-software-with-ai

## Summary

AI brings massive productivity gains, making software significantly faster and cheaper to build. However, this often results in "plausible software"—predictable, functional, but lacking aspiration and craft, much like the tradeoffs seen in modern architecture. The optimistic view, however, reveals that AI drastically raises the ceiling for ambitious builders. To achieve this, product designers and engineers must evolve into quality stewards, treating prompt engineering as the industry's most critical meta-skill to reliably translate human intuition into machine-readable intent.

At Notion, this cultural shift relies on a philosophy of "demos over memos," where product requirement documents (PRDs) are increasingly written for autonomous agents rather than humans. To bridge the famously painful gap between Figma mocks and production, teams leverage a shared Next.js "Prototype Playground." This isolated, multiplayer repository forces designers to work intimately with actual code, exposing real-world UX constraints like network latency, missing loading states, and browser edge cases early on. It also serves as a crucial sandbox for mastering AI tools, equipped with specialized sub-agents like a "Figma verifier" and effortless one-click zero-risk deployment hooks.

Scaling high-quality AI generation at an organizational level demands a deeply shared design vocabulary; when agents understand internal systems perfectly, even a low-fidelity "napkin sketch" can predictably output robust code. Notion structures this through a "Software Lab" database acting as an asynchronous dark factory, where simple state changes trigger complex background agent processing. To prevent AI from making developers lazy, teams must enforce rigorous verification loops. This means scoping tasks over days rather than minutes, running adversarial AI code reviews, and building tools like visual diff explainers that actively compound our understanding of what the machines are writing under the hood.

**Keywords:** ai software quality, plausible software, prompt engineering skills, demos over memos, prototype playground, next.js prototyping, figma-to-code handoff, design engineering, shared ai vocabulary, autonomous coding agents, adversarial ai review, software lab workflows, asynchronous ai collaboration, automated pull requests, ux edge cases, ai-generated prds, visual code diffs

## Chapters

1. **Defining software quality as behavior encountering reality** (00:09) — The definition of software quality rests on how effectively a product behaves when encountering real-world conditions.
1. **Comparing productivity gains against perceived quality improvements** (01:26) — An initial surge in software creation speed often yields more output without raising overall quality standards.
1. **Architecture analogies for the commoditization of software design** (02:34) — Parallels between modern affordable housing and software highlight a shift toward reliable but uninspiring deliverables.
1. **Plausible software and the limits of current generation models** (04:54) — Generative models currently excel at assembling median-tier applications that reveal cut corners upon deeper user interaction.
1. **Embracing ambition to build uniquely beautiful software with AI** (06:36) — Democratized contribution and self-healing repositories empower engineering teams to elevate the ceiling for software craftsmanship.
1. **Transitioning from visual canvas tools to terminal based environments** (10:02) — Shifting from visual editors to terminal-based environments reveals the deep difficulty of resolving the last mile of design polish.
1. **Articulating design intuition as a core prompting meta skill** (13:10) — Experienced developers must learn to verbalize their deeply ingrained muscle memory to guide coding models effectively.
1. **Progression levels for mastering artificial intelligence coding agents** (15:24) — Prompting proficiency ranges from accepting basic autocomplete suggestions to orchestrating multiple autonomous workers using custom organizational harnesses.
1. **Prioritizing pull requests over memos in autonomous agent workflows** (16:40) — Emphasizing functional demos shifts product requirement documents from targeting human stakeholders to explicitly instructing autonomous agents.
1. **Prototyping playgrounds for direct designer engagement with material code** (17:29) — Dedicated production-like sandbox repositories allow creative teams to prototype directly in code without risking live service stability.
1. **Integrating autonomous sub-agents for design verification and user research** (22:43) — Deploying continuous environment monitors and design verifiers accelerates component iteration and generates short-lived links for user research.
1. **Establishing a shared vocabulary for low fidelity system design** (23:51) — Formulating a common language enables developers to bypass polished mockups and build robust features directly from rough concepts.
1. **Software labs as environments for human and agent collaboration** (25:24) — Database-driven internal platforms seamlessly coordinate task tracking and asynchronous handoffs between human implementation and automated review.
1. **Running extensive agent implementation sessions with persistent verification loops** (26:52) — Structuring broader conceptual tasks and enabling adversarial code evaluations lets developers focus on system planning while models execute backlogs.
1. **Automating user bug triage and resolutions using Slack agents** (28:36) — Automated internal processes capture user feedback to instantly investigate repositories and compile initial remediation pull requests.
1. **Creating visual code artifacts to enhance pull request comprehension** (29:51) — Generating explicit visual state breakdowns prevents cognitive laziness and drastically simplifies comprehensive pull request evaluations.

## Related Moments

- [Moving beyond demos to build production-ready software](https://www.wearedevelopers.com/videos/100042-it-s-a-great-time-to-be-a-builder-leveraging-ai-for-good) (from "It's a Great Time to be a Builder: Leveraging AI for Good")
- [Motivations for adopting AI to enhance developer productivity](https://www.wearedevelopers.com/videos/1266-navigating-the-ai-revolution-in-software-development) (from "Navigating the AI Revolution in Software Development")
- [Using AI to rapidly prototype intuitive application features](https://www.wearedevelopers.com/videos/1377-rethinking-intelligence-ai-accessibility-and-the-future-of-inclusive-work-artur-ortega) (from "Rethinking Intelligence: AI, Accessibility, and the Future of Inclusive Work - Artur Ortega")
- [Designing complex software architecture in the era of AI](https://www.wearedevelopers.com/videos/1365-wearedevelopers-live-the-weekly-developer-show-with-chris-heilmann-and-daniel-cranney) (from " WeAreDevelopers LIVE - the weekly developer show with Chris Heilmann and Daniel Cranney")
- [Navigating developer bottlenecks and human accountability](https://www.wearedevelopers.com/videos/100265-fireside-chat-in-conversation-with-werner-vogels-cto-of-amazon-com) (from "Fireside Chat - In conversation with Werner Vogels, CTO of Amazon.com")
- [Shifting developer workloads and realistic AI productivity gains](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions")

## Related Articles

- [Never delegate the understanding](https://www.wearedevelopers.com/magazine/749-never-delegate-the-understanding)
- [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care)
- [Transforming Software Development: The Role of AI and Developer Tools](https://www.wearedevelopers.com/magazine/527-transforming-software-development-the-role-of-ai-and-developer-tools)
- [What is Software Engineering in the Age of AI?](https://www.wearedevelopers.com/magazine/640-what-is-software-engineering-in-the-age-of-ai)

## Related Jobs

- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [Principal Product Manager, Agent Platform](https://www.wearedevelopers.com/jobs/ext/277541-principal-product-manager-agent-platform) at **GitHub**
- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**
- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**
- [AI Full Stack Engineer](https://www.wearedevelopers.com/jobs/ext/1354435-ai-full-stack-engineer) at **Almedia**