Narrative trajectories
Five ways through
the same territory.
Each trajectory has a center, an emotional shape, and a cost.
The question is not which contains the most ideas. It is which makes the audience feel one coherent argument unfolding.
Agency Through Feedback
One loop for building; the same loop widened into a way of meeting an unknowable future.
Disorientation → recognition → practical control → sober agency
The Finite Human and the Tireless Producer
The limiting resource is not generated output, but the human capacity to preserve judgment and direction.
Awe → overload → recognition → a humane practice of limits
From Plausible Output to Dependable Systems
A prompt can produce an outcome; software makes it observable, repeatable, governable, and scalable.
Impressive result → suspicion → decomposition → earned confidence
Selective Ownership
What does it mean to own work you did not write and cannot fully hold in your head?
Confession → discomfort → distinction → responsible stewardship
Rebuilding Professional Identity
If writing code becomes less scarce, what should a software engineer understand their value to be?
Loss → inventory → possibility → deliberate reinvention
Fast comparison
What each path optimizes.
Use this view to choose the audience experience before arguing about individual slides.
| Path | Strongest at | Main risk | Center of gravity |
|---|---|---|---|
| Stay in the loop | It satisfies every promise in the abstract while giving the practical and philosophical material one underlying problem. | The repeated loop must widen in meaning; otherwise the ending can feel like a recycled engineering metaphor. | Personal rupture · The finite human · Milestones |
| The finite human | It names an experience many practitioners feel but lack language for, while connecting attention directly to engineering quality. | It can feel introspective or process-heavy and needs an explicit answer to why software engineers are especially relevant. | Cognitive saturation · Expansion and compression · Activity without progress |
| From output to system | It gives a precise, practical answer to why software engineering still matters. | It can become a technical AI architecture lecture with the career, community, and identity arguments attached afterward. | System around the model · Failure taxonomy · Outcome vs system correctness |
| Ownership after authorship | It speaks directly to engineers who feel simultaneously empowered and fraudulent, and makes ownership genuinely substantive. | It can sound like a rationalization for weak code review unless stakes, evidence, and the limits of delegation are explicit. | Meeting confession · Production–judgment gap · Expertise as anchor |
| Identity after authorship | It makes the title do real work and gives the audience permission to grieve the old craft while choosing a new identity. | Without a tightly integrated demonstration and specific examples, it can drift into generic future-of-work speculation. | Career story · Lost mastery · Four dispositions |