AI Won't Make Everyone A Tool Builder
Benedict Evans challenges the narrative that AI democratizes tool creation, arguing that problem identification, workflow design, and enterprise adoption remain critical barriers. Value shifts to opinion and taste as execution costs drop.
The prevailing narrative that AI democratizes tool creation and renders traditional software obsolete is fundamentally flawed. Benedict Evans argues that while AI lowers technical barriers, it does not alter the core human and organizational constraints that dictate software success. The shift is evolutionary, not revolutionary, mirroring historical transitions like no-code and spreadsheet databases.
The Illusion of Universal Tool Building
Evans deconstructs the 'everyone becomes a builder' myth through three critical barriers. First, most professionals are immersed in execution and lack the cognitive distance to identify automation opportunities; they do not see the problems that software could solve. Second, domain expertise does not equate to software architecture. A master lawyer or salesperson possesses distinct skills from those required to design effective legal tech or sales enablement platforms. The ability to perform a task is unrelated to the ability to systematize it for others. Third, enterprise environments demand rigorous security, compliance, and cross-functional alignment. Individual tools cannot bypass the need for permissioning, audit trails, and integration with multiple systems of record.
Adoption Dynamics and Market Realities
Historical patterns persist in the AI era. Bottom-up adoption captures only approximately 5% of the enterprise market. Successful software deployment still requires evangelism, budget justification, and integration with existing infrastructure. Evans highlights that shadow IT, such as departmental Excel files, often evolves into institutional software like SAP or Workday when accountability, bug fixes, and scalability become necessary. AI mirrors previous shifts like no-code, expanding the perimeter of tool creation without displacing the need for supported, secure enterprise solutions. The bottleneck remains organizational adoption, not technical capability.
Value Migration to Opinion and Workflow Design
As execution costs approach zero, competitive advantage migrates to 'opinion' and 'taste.' LLMs generate average outputs based on existing patterns; value accrues to those who can define unique workflows and challenge standard practices. The differentiator is no longer writing code but articulating precise requirements and designing novel solutions that transcend generic AI capabilities. Prompt engineering remains vital, not as a technical trick, but as the ability to flowchart and articulate complex processes—a skill most professionals lack. Success depends on knowing what to build and how to sell the solution internally, rather than merely having the tools to build it.
Key insights
-
High performance in a domain does not translate to the ability to design effective software solutions for that domain.
Impact: Companies should recruit dedicated product architects rather than assuming subject matter experts can build superior tools.
-
Bottom-up software adoption captures only a marginal share of the enterprise market, requiring top-down evangelism and budget justification.
Impact: SaaS vendors must maintain robust enterprise sales motions and compliance frameworks despite AI-driven democratization trends.
-
As AI lowers execution costs, competitive advantage shifts to unique product opinions, workflow design, and strategic taste.
Impact: Businesses must prioritize differentiated value propositions and curated user experiences over feature parity or technical implementation.
-
Effective AI utilization depends on the ability to articulate complex workflows, a skill distinct from coding or prompt syntax.
Impact: Organizations should invest in process mapping and workflow documentation to maximize AI agent effectiveness.
Action items
-
Conduct workshops to help teams identify hidden automation opportunities by mapping current processes and pain points.
Impact: Uncovers latent efficiency gains that professionals overlook due to execution focus.
-
Structure product teams to pair domain experts with dedicated software architects to bridge the skill gap.
Impact: Ensures tools are built with rigorous design principles rather than just domain knowledge.
-
Create comprehensive adoption strategies addressing security, compliance, and cross-team integration for new AI tools.
Impact: Mitigates shadow IT risks and accelerates scaling of internal solutions.
Quotes
“Most people don't see the problem that you're going to automate.”
“The mark of great software is you see something and you think, oh, wow, that's a great idea. And you can't.”
“Your value as a lawyer is doing something that isn't exactly how anybody would do it.”