The New Stack Podcast

chevron_right

Stop Writing Code, Start Writing Docs

whatshot 28 snips

Oct 31, 2025

Guest

Keith Ballinger

Keith Ballinger, Vice President at Google Cloud Platform Developer Experience, shares his insights on the evolving landscape of AI coding tools. He discusses the Gemini CLI and its role in enhancing developer workflows, advocating for writing clear documentation to collaborate with AI effectively. Ballinger believes programming will shift from writing code to articulating intent. He explores the resurgence of the command line, the importance of great writing for programmers, and how AI can transform software development into a cooperative effort.

01:03:25

forum

Ask episode

web_stories

AI Snips

view_agenda

Chapters

menu_book

Books

auto_awesome

Transcript

info_circle

Episode notes

volunteer_activism

ADVICE

Document Intent Before Using AI

Slow down to speed up: write a user guide, spec, and technical design before using AI on complex projects.
Treat AI like a coworker and feed it clear plans to get reliable, large-scale results.

question_answer

ANECDOTE

Conductor Repo Walkthrough

Keith demonstrated Conductor: a repo-driven workflow that creates docs, plans, and prompts to guide agents.
He used it to build Aether, a programming language prototype, by iterating tests and examples.

volunteer_activism

ADVICE

Verify With Tests And Examples

Use tests and examples as verification: have the AI write tests and implement features until tests pass.
Inspect examples and test outputs rather than every line of generated code.

Get the Snipd Podcast app to discover more snips from this episode

Scope of Google Developer Experiences

00:16 • 1min

chevron_right

Balancing AI with Other Developer Projects

01:42 • 1min

chevron_right

Keith's Career Path and Background

03:01 • 5min

chevron_right

How Keith Started Using AI Tools

07:53 • 4min

chevron_right

Documenting Intent to Work With AI

12:13 • 5min

chevron_right

Building a Language for LLMs

16:51 • 2min

chevron_right

Testing and Verification With AI

18:44 • 2min

chevron_right

Inspiration from Personal Pain Points

21:07 • 2min

chevron_right

Main Limitations of Early Agent Workflows

23:35 • 51sec

chevron_right

Why the CLI Is Resurgent for AI

24:26 • 1min

chevron_right

Evolving Terminal and Shell Interfaces

25:55 • 2min

chevron_right

Integrating CLI with IDE Workflows

27:43 • 1min

chevron_right

Programming Languages and Abstraction Shifts

29:06 • 3min

chevron_right

Natural Language vs. Formal Interfaces

32:03 • 2min

chevron_right

The Evolving Skillset for Programmers

34:19 • 1min

chevron_right

Addressing AI's Black-Box and Reliability Issues

35:20 • 3min

chevron_right

State of Gemini CLI and Next Steps

38:30 • 3min

chevron_right

Workflows: From Code to Deployment with Agents

41:57 • 17sec

chevron_right

Background Subagents and Monitoring

42:14 • 2min

chevron_right

Tooling Bootstrapped by the CLI Itself

44:29 • 47sec

chevron_right

Context Management Challenges

45:15 • 2min

chevron_right

State Management Across Agents and Repos

46:47 • 2min

chevron_right

Cloud's Role in Managing Agent Infrastructure

48:40 • 2min

chevron_right

Using Agents for Scientific Research and Papers

50:43 • 4min

chevron_right

Publishing and AI Authorship Norms

54:31 • 2min

chevron_right

Future AI-Driven Conferences and Entertainment

56:15 • 1min

chevron_right

Impact of Full No-Code/Vibe Coding

57:27 • 3min

chevron_right

Speculation on Dynamic, On-the-Fly UI Generation

01:00:06 • 13sec

chevron_right

What Keith Is Reading Now

01:00:20 • 2min

chevron_right

Final Wrap-Up and Contact Info

• Mentioned in 23 episodes

Seven and a Half Lessons about the Brain

Lisa Feldman Barrett

In this book, Lisa Feldman Barrett demystifies the brain through seven short essays and a brief story on its evolution. She explains where brains came from, their structure, and how they function in tandem with other brains. The book dispels myths such as the 'lizard brain' and the battle between thoughts and emotions or nature and nurture. Barrett highlights the brain's primary role in controlling the body, discusses brain development, and emphasizes the social and cultural aspects of cognition and consciousness.

#1101

• Mentioned in 26 episodes

Tomorrow, and Tomorrow, and Tomorrow

Gabrielle Zevin

Spanning thirty years, 'Tomorrow, and Tomorrow, and Tomorrow' by Gabrielle Zevin is a sweeping narrative that explores the lives of Sam and Sadie, two friends united by their love of video games. From their childhood encounters to their adulthood as successful game developers, the book delves into themes of identity, creativity, disability, and the redemptive possibilities in play. It is a coming-of-age story that examines the multifaceted nature of human relationships and the need to connect, love, and be loved. The novel is praised for its mature and nuanced portrayal of characters and its ability to engage readers even without prior knowledge of the video game industry.

#5996

• Mentioned in 6 episodes

Your Code as a Crime Scene

Adam Tornhill

#626

• Mentioned in 39 episodes

A theory of justice

John Rawls

In 'A Theory of Justice,' John Rawls develops a moral theory of justice as fairness, which is an alternative to utilitarianism. The theory is based on the social contract approach and uses the concept of the original position and the veil of ignorance to derive two principles of justice. The first principle, the greatest equal liberty principle, ensures that each person has an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberty for all. The second principle, which includes the difference principle and the equal opportunity principle, states that social and economic inequalities are to be arranged so that they are to the greatest benefit of the least advantaged and attached to offices and positions open to all under conditions of fair equality of opportunity. Rawls's theory aims to reconcile liberty and equality in a well-ordered society.

In this episode of The New Stack Podcast, hosts Alex Williams and Frederic Lardinois spoke with Keith Ballinger, Vice President and General Manager of Google Cloud Platform Developer Experience (GPC), about the evolution of agentic coding tools and the future of programming. Ballinger, a hands-on executive who still codes, discussed Gemini CLI, Google’s response to tools like Claude Code, and his broader philosophy on how developers should work with AI. He emphasized that these tools are in their “first inning” and that developers must “slow down to speed up” by writing clear guides, focusing on architecture, and documenting intent—treating AI as a collaborative coworker rather than a one-shot solution.

Ballinger reflected on his early AI experiences, from Copilot at GitHub to modern agentic systems that automate tool use. He also explored the resurgence of the command line as an AI interface and predicted that programming will increasingly shift from writing code to expressing intent. Ultimately, he envisions a future where great programmers are great writers, focusing on clarity, problem decomposition, and design rather than syntax.

Learn more from The New Stack about the latest in Google AI development:

Why PyTorch Gets All the Love

Lightning AI Brings a PyTorch Copilot to Its Development Environment

Ray Comes to the PyTorch Foundation

Join our community of newsletter subscribers to stay on top of the news and at the top of your game.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Home Top podcasts Popular guests Top books