Nathan Wrigley interviews Luke Carbis about the future of WordPress plugins, covering plugin discovery, the surge of AI-generated submissions, ethical design, and possible new standards for the plugin directory.
Who is Luke Carbis
Luke has been active in WordPress for about 20 years: developer, product lead, agency and hosting roles, plugin entrepreneur, Core contributor, WordCamp organiser, and a member of the Plugin Review Team. He also co-hosts the Crossword podcast.
The problem: plugin overload and discoverability
Luke explains that the plugin directory has been flooded with many more submissions in the last year, driven heavily by AI-assisted plugin generation. Submissions have multiplied, making it harder for quality plugins to stand out and for users to find the right tool. Although the Plugin Review Team has grown and improved tooling (keeping review wait times to roughly a week), sheer volume and duplication remain the primary UX problem: how do good plugins surface among thousands offering similar features?
How wordpress.org search and rankings work
The directory’s search is open source and influenced by a range of signals: keyword matching, recent reviews, author responsiveness in support forums, and notably active installs, which are hard to fake and carry strong weighting. Still, creators and users feel the system could be improved: open algorithms invite gaming, and a heavy reliance on active installs disadvantages new, quality plugins.
Practical experiments and proposals
Luke suggests practical changes that could be implemented without requiring wholesale governance shifts:
– Connect wordpress.org accounts to WordPress installs: Let users link their wordpress.org identity across sites (similar to the Connectors API), making personal favourites and recommendations available directly in the Add Plugins UI.
– Allow trusted external repositories for installs: Let users register Git repositories or premium plugin sources in their wordpress.org profile. Those entries could appear under an “untrusted sources” (or similar) view in Add Plugins so developers and agencies can quickly install their own custom or commercial plugins without repackaging zips manually. The central directory remains controlled, but individuals get more discoverability and convenience for their own plugins.
– Experiment with an official commercial marketplace: While controversial, Luke imagines a model where premium plugins can be sold through the directory, with a modest platform fee. He proposes roughly 8% (3% payment processing + 5% to the Foundation or similar fund) to support contributors, events, and reviewer compensation. That could streamline licensing, updates, and discoverability for paid offerings — but would require transferring directory control to the Foundation and is politically sensitive.
Pros and cons of a commercial directory
Introducing payments can attract criticism from those who see WordPress as strictly free/open-source at the point of use. Luke argues that a properly governed marketplace could fund contributor work, WordCamps, and plugin reviewers, and could revive product momentum in the ecosystem. Opponents fear commercial pressures, preferential ranking for paid plugins, and a split community. Luke acknowledges these risks but sees potential benefits if governance and transparency are strong.
AI as an agent of change, not just a tool
AI is reshaping plugin submissions and broader community dynamics. Luke and Nathan discuss two distinct AI-related issues:
– Product integration: Building optional AI features into WordPress (for content generation, editing assist, etc.) is OK when optional and user-controlled.
– Project/tooling reliance: Using AI to generate documentation, PRs, or contributions raises ethical and cultural questions. There is concern for younger contributors: many Gen Z members express strong resistance to AI, viewing it as environmentally harmful or de-skilling. Over-emphasising AI in project contributions could alienate potential contributors.
Transparency and AI disclosure
To address both practical and ethical concerns, Luke proposes an AI disclosure header for plugins. This would be a voluntary metadata field indicating the level of AI involvement across a spectrum (from minor assistance to full AI-generated code). Visible disclosure on plugin pages would let the market and reviewers collect data about AI usage, help users make informed choices, and encourage accountability. Luke acknowledges this could be gamed, but sees it as a starting point for research and policy.
Leadership, strategy, and the project’s future
The conversation touches on the need for clear direction in WordPress. Luke notes a renewed activity from project leadership and feels the project needs decisive stewardship to adapt to rapid external change. He argues a strong leader could provide the focus required to respond to AI and other shifts — though that could invite controversy. Nathan and Luke both weigh the trade-offs between community-led, consensus-based decision-making and more directive leadership styles in a period of potential disruption.
Practical realities for plugin authors
Many plugin authors use wordpress.org simply as a convenient host for their own client or agency installations. The ability to register and quickly deploy custom or premium plugins across sites would reduce friction and lower the bar for product launches. Luke argues that reducing distribution and licensing burdens makes launching a product easier and could reinvigorate innovation in the WordPress product space.
Final thoughts and next steps
Luke believes the plugin ecosystem can be improved through pragmatic experiments: better account-to-site connections, optional external repo support, AI disclosure metadata, and thoughtful debate about a commercial marketplace. Each idea has trade-offs; governance, transparency, and community input would be critical.
If you want to dig deeper, Luke co-hosts the Crossword podcast at crossword.fm, where he explores WordPress topics in greater detail.
This conversation highlights three recurring themes: the technical challenge of scaling review and discovery systems, the ethical and cultural questions around AI, and the political debate over whether and how to introduce commercial mechanisms into an open-source directory. Luke frames these not as solved problems but as areas for community experimentation — with careful attention to governance, inclusivity, and long-term sustainability.