Get AI summaries of any video or article — Sign up free
The Future of Obsidian with CEO Stephan Ango and Andy Polaine thumbnail

The Future of Obsidian with CEO Stephan Ango and Andy Polaine

Nicole van der Hoeven·
6 min read

Based on Nicole van der Hoeven's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Stephan Ango (Kepano) anchors Obsidian’s long-term strategy in privacy, note longevity, and extensibility, with plain-text Markdown treated as the durable layer.

Briefing

Obsidian’s CEO Stephan Ango (Kepano) frames the product’s future around three non-negotiables—privacy, note longevity, and extensibility—arguing that the app’s long-term value depends more on plain-text files than on any single interface. The core pitch is simple: Obsidian should stay fast, local-first, and customizable, even if that means accepting rough edges like plugin “jank” and a steeper learning curve for newcomers.

Ango’s own entry point explains the philosophy. After years of using wiki-style note systems—starting with TiddlyWiki and earlier personal experiments—he says Obsidian arrived in April 2020 with the right combination of minimal design and practical speed: quick switching, prompts, and a file navigation approach that felt closer to a local-first “code editor” workflow. He emphasizes why local storage matters: cloud tools introduce latency and friction, while offline access enables uninterrupted writing, including on flights. That speed and control also connect to a broader belief that digital information must persist for centuries; Markdown, as a plain-text format, is treated as the durable layer that can outlive today’s apps.

From a leadership standpoint, Ango treats the growing “Obsidian clone” ecosystem as a healthy sign rather than a threat—so long as clients don’t drift into app-specific syntaxes that break interoperability. He calls the current moment a “Cambrian explosion,” where multiple tools can explore different ways of thinking. The risk, he says, is when innovation turns into fragmentation: too many proprietary extensions can make notes harder to move and harder to preserve.

A major trade-off shows up in how Obsidian handles plugins. Ango and Andy Polaine compare the plugin system to how modern mobile platforms make the whole device feel like the active app: Obsidian is fundamentally a Markdown-based editor and viewer, but powerful plugins can transform the experience. That freedom comes with downsides—new users may need to install many plugins, and community add-ons can feel less native. Still, Ango argues that extensibility is a core design principle, and “choosing to be bad” at cohesion is the price of letting users tailor workflows.

Looking ahead, the roadmap prioritizes “primitives” rather than bloated feature sets. Metadata is a key focus: improving how users add and query structured information, building toward easier use of DataView-style workflows without forcing everyone into SQL-like syntax. Publish and collaboration are also active areas, with improvements aimed at external sharing of entire vaults, shared vaults for internal teams, and single-note viewing through links. Task management is treated more conservatively—mainly making tasks more queryable and improving due-date handling—while Canvas and metadata work take precedence.

AI enters the conversation with a privacy-first constraint. Ango says community plugins already integrate GPT-style tools, but he prefers local-first, private AI where inference happens on the user’s machine. He points to the technical and hardware barriers of local training/inference and suggests Obsidian will likely watch plugin experiments before deciding what belongs in core.

Finally, Ango acknowledges Obsidian’s biggest usability gaps: sharing and collaboration remain harder than the solo note-taking experience, and accessibility feedback is ongoing. He ends by crediting the community for making the tool feel welcoming and collaborative—an asset he sees as central to Obsidian’s long-term adoption.

Cornell Notes

Stephan Ango (Kepano), CEO of Obsidian, ties the product’s future to three pillars: privacy, longevity of notes, and extensibility. He argues that plain-text Markdown is the durable layer that should survive app churn, while local-first performance keeps writing fast and reliable. Obsidian’s plugin ecosystem is treated as both its strength and its weakness: it enables users to transform the editor into different workflows, but can also create “janky” experiences and a higher setup burden. Roadmap priorities emphasize primitives like metadata (to make structured querying easier), plus publish and sharing improvements, while AI and task management are approached cautiously to preserve privacy and avoid feature bloat.

What “hook” pulled Stephan Ango into Obsidian, and why does it matter to his product philosophy?

Ango describes a long personal history with wiki-style note systems, especially TiddlyWiki, which used bracket-style linking. His key requirement was a minimal, simple interface that still delivered the speed and workflow patterns he expected from code editors. When Obsidian launched in April 2020, it offered quick switching, prompts, and file navigation that felt immediately better than his heavily customized TiddlyWiki setup. That early experience shaped his emphasis on local-first control, offline use, and fast interaction—principles he later connects to privacy and longevity.

Why does Ango treat Markdown (and plain text) as more important than the app itself?

He argues that apps change or disappear, but plain text can persist for centuries. Digital files have existed for only about 70–80 years, while books have lasted thousands. Obsidian is therefore positioned as a temporary interface over a durable storage format: even if Obsidian becomes obsolete, Markdown notes should remain usable. This also explains his caution about app-specific syntaxes that could reduce interoperability across tools.

How does Obsidian’s extensibility create both value and friction?

Ango frames extensibility as a core design principle, but he acknowledges the trade-off: deciding what belongs in core versus community plugins is a constant challenge. New users may need many plugins, and community add-ons can feel “janky” or less cohesive. Andy Polaine compares the plugin mechanism to Apple’s iPhone model—turning the whole experience into whatever “app” the user activates—so Obsidian can become something more than a basic Markdown viewer. The friction is real, but the flexibility is the point.

What does “primitives” mean in Obsidian’s roadmap, and what’s the current example?

“Primitives” refers to foundational capabilities that others can build on, rather than a single monolithic feature set. Ango highlights metadata as a major focus: improving how users add structured data (currently via YAML frontmatter and DataView-style custom fields) and making metadata easier to use as a first-class concept. He notes DataView’s syntax is powerful but nerdy and painful to write in YAML, so Obsidian wants to reduce that friction and enable richer developer ecosystems.

How does Obsidian approach AI without breaking its privacy and local-first principles?

Ango says Obsidian already has GPT-style integrations through community plugins like Text Generator, and the plugin ecosystem can move quickly. But he prefers AI that is local-first and private, with inference happening on the user’s machine and minimal data leaving the device. He points to the difficulty of local training/inference (GPU intensity and setup complexity) and suggests waiting for more mature open-source model options before attempting core-level solutions. In the meantime, community plugins fill the gap.

What sharing and collaboration problems remain, and how does the team think about them?

Ango and Nicole discuss that Obsidian excels for isolated personal use but struggles when multiple people need shared repositories or coordinated access. Ango distinguishes several sharing use cases—publishing entire vaults to the external world, sync-based shared vaults for internal teams, link-based single-note viewing for people who don’t want to learn Obsidian, and true collaboration workflows akin to Google Docs/Notion. The team’s challenge is sequencing these features while staying consistent with privacy principles, especially because encryption limits what a browser-based editor could do.

Review Questions

  1. Which three pillars does Ango name as Obsidian’s core design priorities, and how does each one shape product trade-offs?
  2. What arguments does Ango give for why plain-text Markdown should outlast Obsidian as an app?
  3. How do Ango and Polaine justify the plugin system despite user complaints about setup complexity and “jank”?

Key Points

  1. 1

    Stephan Ango (Kepano) anchors Obsidian’s long-term strategy in privacy, note longevity, and extensibility, with plain-text Markdown treated as the durable layer.

  2. 2

    Local-first performance is presented as a productivity advantage over cloud tools, especially for offline use and low-latency editing.

  3. 3

    The rise of Obsidian clones is viewed as healthy “ecosystem diversity,” provided tools avoid drifting into app-specific syntaxes that harm interoperability.

  4. 4

    Obsidian’s plugin model is a deliberate trade-off: it enables transformative customization but can create a less cohesive experience for new users.

  5. 5

    Roadmap priorities emphasize foundational “primitives” like metadata and improvements to publish/sharing, rather than expanding into a bloated all-in-one suite.

  6. 6

    AI integration is approached cautiously, with a preference for local-first, private inference aligned to Obsidian’s privacy principles.

  7. 7

    Sharing is treated as a multi-problem area (publishing, shared vaults, single-note links, collaboration), each requiring different solutions under the same privacy constraints.

Highlights

Ango argues that the app can become obsolete, but Markdown/plain text should remain usable for centuries—making file format durability the real product.
The plugin ecosystem is defended as a feature, not a bug: powerful add-ons can turn Obsidian into something “completely else,” similar to how mobile platforms make the whole device feel like the active app.
Metadata is singled out as a key primitive: improving how users add and query structured information could unlock richer workflows without forcing everyone into painful YAML or SQL-like syntax.
AI is framed as a privacy and performance problem first; local-first inference is the goal, but hardware and model maturity determine how quickly core support can arrive.
Sharing is broken into distinct use cases (vault publishing, shared vaults, link-based note viewing, and collaborative editing), and the team prioritizes them based on feasibility and principles.

Topics

Mentioned