From the article:

Executives are personally active on AI at rates that match or beat their teams. CEOs at companies of 201 or more people went from 9% to 36% in six months, the largest jump of any cut in this report, suggesting the most senior leaders are learning the technology by using it rather than reading about it. Company size comes from third-party enrichment, so this cut covers fewer workspaces than the rest of the report.

[...]

Two years ago, fewer than one issue in a thousand was created by AI. Teams now use AI to write just under half of everything created in Linear, and at the current pace it will soon author more than people and integrations combined.

[...]

Teams that connected a coding agent roughly tripled their weekly pull requests over two years, from 21 to 65, while teams without one went from 8 to 10. These teams were already higher-output before coding agents existed, so the levels aren’t directly comparable, but each cohort against its own baseline tells a clean story, and nearly all the growth sits on the agent side.

[...]

The clearest indication of AI’s influence on product development is the dramatic output gains experienced by teams using coding agents over the last two years. We have no way of knowing whether this increased output led to positive business outcomes, but it shows a very clear correlation between AI adoption and acceleration.

Perhaps more intriguing is the makeup of that adoption, and how it appears to be blurring roles. Senior leaders are doing more of the hands-on IC work, adopting AI aggressively to help them do it, and non-engineers are committing code. The suggestion that everyone in an organization is becoming a “builder” seems to be directionally true.

Those gains haven’t shown up as time saved, though. Time spent on existing tasks in Linear held while AI usage appeared as a new layer of work, meaning the overall time spent on product development is going up rather than down. As far as we can observe, teams are working more, not less, suggesting AI has a Jevons paradox quality beyond token consumption.

Many will rightfully argue that looking at pull requests indicates motion rather than value, which is certainly true, but it’s still a step forward from measuring tokens. A mechanical refactor might burn lots of tokens while a meaningful bug fix or code review doesn’t, so token spend and value don’t line up at all, and using one as a proxy for the other will be remembered as a relic of AI’s early days.