From Telemetry to Empathy: Measuring AI in Your Teams

On This Podcast

In this conversation, we examine Multitudes' finding that developers using agentic coding tools are committing 19.6 percent more outside their typical working hours, adjusted for time zone and personal preference. Lauren explains that while many engineering leaders assume this reflects renewed joy in coding, the survey and interview data pointed to a different, less comfortable driver: delivery pressure combined with a steep AI learning curve that leaves developers squeezing in extra hours to keep pace. We also discuss how relying on telemetry alone would have missed this nuance entirely, underscoring the value of pairing quantitative data with qualitative interviews. The pattern is complicated further by a difficult economic climate and layoffs, which heighten anxiety about job security even as leaders proclaim AI adoption as purely positive.

In this conversation, we examine what separates successful AI rollouts from failed ones, starting with Lauren's insistence that leaders must articulate a clear why rather than leaving teams to guess at the goal. We also discuss the concept of commander's intent, borrowed from military mission briefings, as a model for giving engineers the outcome and reasoning they need to adapt when things go wrong. Lauren shares that peer-to-peer AI demos, run by trusted colleagues who share the same codebase and challenges, consistently outperform top-down mandates or vendor-led training because they carry more credibility. Daniel adds his own experience running workshops that surface fears and ideas from engineers directly, reinforcing that listening to the people closest to the work is itself a powerful lever for adoption.

In this conversation, we examine the growing anxiety among engineers, particularly seniors, about what AI-driven productivity means for the future of junior hiring and mentorship. Lauren's research found that individual contributors and team-level managers were far more likely to raise concerns about the talent pipeline organically than senior leadership, suggesting a disconnect between those closest to the work and those setting strategy. We also discuss why she believes this is an area where individual contributors hold real influence, since it is engineers themselves who must volunteer to mentor and support new hires for junior programmes to succeed. The conversation closes on a more hopeful note, with Lauren citing organisations that are still hiring juniors precisely because they are AI-native, onboard faster, and represent a talent pool with less competition than before.