Jul 24, 2026 · 18m · another-podcast
Most people and companies aren't tool-builders
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Benedict Evans and Toni Cowan-Brown discuss why emerging AI and no-code tools will not turn most workers into software builders. They argue that organizational complexity, enterprise compliance, and the critical distinction between domain knowledge and product design taste keep traditional software development dynamics firmly in place.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 99.8% of the talking time here. How this is scored →
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Toni directly interrupts Benedict to question whether his claim that a billion people use AI for newsletters is an actual statistic.
Hardest push from the hosts ▶ 16:54 Refuting the death of prompt engineeringBenedict bluntly dismisses prevailing industry claims that prompt engineering is obsolete, arguing that articulating requirements remains deeply difficult for most humans.
Biggest teaching moment ▶ 5:31 Improvised vs institutional software distinctionBenedict clarifies why running a department on a sprawling Excel file does not make users tool builders, contrasting ad-hoc workflows with institutional, maintained software.
The host holds their own ▶ 8:51 Historical analysis of enterprise software adoptionBenedict demonstrates deep domain knowledge by detailing how grassroots enterprise sales only worked for rare outliers like Slack while traditional sales cycles remain essential.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| The Three Core Barriers to Becoming Tool Builders | 7 | 0 | 1 | 1 | Benedict establishes a clear analytical framework outlining the three core reasons most professionals do not become tool builders. The exchange is structured and collaborative as Toni prompts Benedict to unpack his thesis. | |
| Organizational Collaboration and the Frame.io Case Study | 8 | 1 | 2 | 3 | Benedict cites real-world VC analysis from his time at Andreessen Horowitz looking at Frame.io to explain multi-stakeholder enterprise workflows. When Toni asks whether Excel users and personal agents bypass central software, Benedict distinguishes between shadow improvised tools and institutional systems. | |
| Bottom-Up Adoption Limits and Historical Technology Cycles | 8 | 0 | 1 | 2 | Benedict breaks down historical technology cycles, debunking the Silicon Valley myth of pure bottom-up enterprise sales adoption by noting it only captures a tiny fraction of the market. Toni engages constructively, reinforcing the point that ease of creation does not guarantee distribution. | |
| Product Taste, LLM Outputs, and Domain Expertise | 7 | 1 | 3 | 4 | Benedict explains why LLMs produce average output and why domain taste differs from software product taste. Toni introduces mild friction by calling out Benedict when he uses an exaggerated statistic about newsletter creators. | |
| The Enduring Human Bottleneck in Tool Creation | 7 | 0 | 1 | 3 | Benedict forcefully rejects the common tech claim that prompt engineering is dead, comparing prompting to articulating complex tasks to junior staff. Both hosts conclude in full agreement that human behavior remains the fundamental bottleneck in software adoption. |