Is that the light at the end of the tunnel we see or an approaching train? As 2025 draws to a close it seems a good time to step back from the torrent of press releases and news stories to consider what exactly AI is doing for/to us and what might be in store next year. The Economist reported this week that evidence AI is changing the labor market remains weak but that the experience of software development and call center teams points to a future where work that consists of repetitive tasks that are easily verified and do not require significant context are ripe for disruption. However, this is constrained by companies ability to make their data consumable by the models they use for training.

The impact on our day to day experience is still not clear. On one hand we are being promised, less meetings and three day weeks but on the other we are also being warned of increasing stress as the speed of execution and decision making speeds up to keep pace with our hyperactive models and agents.

Whether you see the light at the end of the tunnel or an imminent collision with some implacable force probably says more about attitudes to change than our ability to predict the future, but I would not get too excited about the prospect of a 3 day week just yet.

More AI less meetings

As AI systems take over routine drafting, summarization, and data retrieval, it has the potential to impact how teams organizes and collaborate. The promise is less meetings and more time for meaningful work. We can only hope!

Fortune Magazine: AI is changing the shape of the work week

Less meetings more work

Is it done yet? While AI should boost throughput and responsiveness, it can also raise expectations for constant availability and faster turnaround, eroding work-life boundaries.
Knowledge at Wharton: The AI efficiency trap

As 2025 winds down these trends are worth watching
ETC Journal: Emerging trends to track

So What?

  • There is an emerging tension between efficiency and sustainability
  • After 3 years we still don’t know where AI is taking us.

Tools & Resources

Until Next Week

Try using AI to challenge your thinking rather than confirming your assumptions

Retrospective is a weekly exploration of AI engineering, knowledge work, and the practices that connect them.