liveBook · Manning

Code generation

As explained in Machines that Think

The automated production of source code by a computational system from specifications, examples, or natural-language instructions.

Code generation lets people request software changes in ordinary language instead of starting with architecture, boilerplate, and implementation details. It can make product creation faster and more accessible, but shifts effort toward understanding, testing, securing, and maintaining the result. section 8.2.5

Typical requests

A builder might ask for a landing page, database connection, authentication, dashboard, or interface change. Generated code can appear to work while hiding errors or accumulating fragile shortcuts, so review and maintenance remain important. section 8.2.5

How Machines that Think covers it

How deep it goes
Describes the workflow and organizational consequences, but not code-generation implementation mechanics.
Examples
Names requests for landing pages, database connections, authentication, dashboards, and interface changes.
What you can do after
Build or modify software through ordinary-language requests while recognizing the need to review, test, secure, and maintain the result.
How it connects
Links code generation to software development, user interface design, cybersecurity, testing, and technical debt.

Also explained in

Compare how 21 books explain code generation →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Machines that Think. The section links point to the parts of the book it draws on.