This feels awkwardly premature and dubious. I think "we're adding features to support better context modelling and execution with LLMs" would have sufficed.

I've thought of Linear as a careful, measured, thoughtful company in the past so this seems out of the blue, like there's some kind of existential crisis occurring over there.

At ZAR once we had pervasive ingestion into an organization-wide knowledge graph in place and working well, the next step was to ditch Linear and replace it with a homegrown experiment tracking system that focuses all product engineers on empirical data and scientific method applied to how we prioritize work.

It's the only way to actually encourage high-agency, high-ownership behavior. Working from a backlog is actively counterproductive!

This industry is becoming so boring.

Yikes. I guess writing your own eulogy isn't just a Stanford MBA exercise anymore.

Show HN: justanissuetracker.com

  Code Intelligence. Linear can understand, answer questions about, and debug your codebase.

  Code Diffs. Review code within a fast, modern interface built for both humans and agents to iterate together.

  Linear Coding Agent. Linear writes code and automatically fixes bugs. Powered by frontier models, enhanced with native Linear context and tools

  Agents are not mind readers. They become useful through context. Customer feedback, internal ideas, strategic direction, decisions, and code all need to be captured in a system that humans and agents can work from together.

Internal ideas and strategic direction are usually discussed on Slack/Teams/meetings. Not on Linear.

Decisions can indeed be tracked on Linear, as an issue.

I think a true AI agent would simply sidestep tools like Linear. Tools like Linear won't be needed.

I think a true AI agent will simply be another employee. It gets added to Slack channels. It joins Zoom meetings. It gets access to company files. It gets access to feedback forms. It scours the internet for feedback on the company.