If you thought the speed of writing code was your problem, you have bigger problems - Andrew Murphy

Value of the Information & Strength of the Argument

The value lies in applying a well-established industrial theory to modern AI-driven software development, offering a systematic way to identify where AI investments actually pay off. The argumentation is coherent and compelling: he builds from personal anecdotes (weekend projects) to empirical statistics (IDC, GitHub) to theoretical lenses (Theory of Constraints, Lean). The reasoning is logical and grounded in real-world experience, though it occasionally leans on assertion rather than rigorous data. The speaker’s credibility is strengthened by his dual role as a hands-on maker and a fractional CTO, and he explicitly positions himself as an ‘AI realist’ rather than pessimist or evangelist. However, some claims (like the 7.2% stability drop) lack precise source details, and the talk does not explore potential counterarguments or limitations of the theory’s application to software.

Scientific Rigor, Source Quality, Title Accuracy

The speaker references IDC and GitHub research, as well as foundational books like ‘The Goal’ and ‘The Phoenix Project’, but does not provide direct URLs or detailed methodologies, reducing traceability. The title perfectly captures the central message and is neither sensationalized nor misleading. The structure is clear: he introduces a problem, explains the theory, provides evidence, and offers actionable advice. While the talk is an opinion piece, its alignment with known industry discussions (e.g., DORA metrics) adds credibility. The lack of a bibliography or visual data tables is a minor weakness, but the narrative flow and practical examples compensate.

The title accurately reflects the core thesis: coding speed is not the primary bottleneck in software development, and larger systemic issues are often overlooked.

7/10

The talk cites specific research (IDC, GitHub) and references established models like Theory of Constraints, but relies heavily on personal anecdotes and executive experience rather than a structured academic study. The arguments are plausible and well-reasoned, but some claims lack detailed evidence.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 01:20Personal story about building automated chicken coop with AI, illustrating the gap between weekend projects and production software.
• 11:00Introduces Theory of Constraints with a simple factory analogy and the one-sentence summary: every system has one bottleneck.
• 19:30Reveals IDC research that engineers spend only 16% of their time coding, raising the question of where the real bottleneck lies.
• 27:00After explaining how AI accelerates coding, he shows the bottleneck moves to discovery and review, using a pipeline diagram.
• 38:00Advises engineers to stop writing code when PRs back up, to avoid compounding work-in-progress waste.

• NDC Conferences — Official conference website.
• NDC Oslo — Specific conference website.
• The Goal — Book by Eliyahu Goldratt, referenced as the origin of Theory of Constraints.
• The Phoenix Project — Book by Gene Kim et al., described as an homage to The Goal.
• IDC research on software engineer time — Mentioned in the talk: engineers spend 16% of time in coding.
• GitHub research on AI coding — Cited for statistics on code speed, churn, and delivery throughput.

• Theory of constraints - Wikipedia — Provides foundational explanation of the bottleneck concept used in the talk.
• Lean software development - Wikipedia — Directly relates to waste and optimization in software processes, echoing the talk's themes.
• DORA metrics — Industry-recognized research on software delivery performance, aligning with the talk's focus on stability and throughput.

• ndcoslo.com