TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
The Pragmatic Engineer published a podcast interview with Titus Winters, a former Google engineer and co-author of Software Engineering at Google, now at Adobe. The conversation covers Google’s large C++ refactor and engineering practices, as well as Winters’ views on AI’s effects on productivity and proposed changes to computer science education.
The Pragmatic Engineer has published a podcast interview with Titus Winters, a former Google engineer who helped lead the company’s code-readability program and co-authored Software Engineering at Google. The discussion examines how Google managed a large-scale C++ refactor and built engineering safeguards, alongside Winters’ views on AI’s impact on software work and the need to rethink computer science education.
Winters spent more than a decade at Google and is now a Senior Principal Scientist at Adobe, where the source says he focuses on developer experience and productivity. The interview covers his work on Abseil, Google’s open-source foundational C++ libraries, and a refactoring effort that the report describes as spanning a codebase of nearly 250 million lines of C++.
Winters recalled that a team of about a dozen people tackled the project after an internal call for volunteers. According to the report, engineers used language-parsing techniques to analyze build targets, identify exact patterns in abstract syntax trees and generate edits. The work also required the old and new APIs to coexist during migration, with changes distributed in batches. The report says the need for tools during the effort led to the creation of the clang-format toolkit.
The conversation also looks at Google’s use of automated tests and code review. The report says Google made a test mandatory with every change after its Search product struggled to maintain a monthly release cadence around 2007. It also describes a review system requiring sign-off from a project owner and a language-qualified readability reviewer. At the time Winters left Google, the source says, more than 10,000 engineers were enrolled in the readability program.
Engineering Safeguards and AI Trade-offs
The discussion connects testing, review and code migration to a practical question for software teams: how to make changes safely as codebases and output grow. Winters argues that reliable safeguards let engineers improve a system without fearing that a change will cause damage. That makes the interview relevant beyond Google, particularly for organizations dealing with growing code volumes and complex dependencies.
The AI portion highlights a tension between faster code production and operational reliability. The source reports Winters’ account of conversations with engineers at other companies: code output has risen sharply compared with a year earlier, while outages were said to be up about 30% and feature shipments remained static. Those figures are an attributed observation, not a measured industry-wide study in the material provided. They nevertheless point to why output alone may be a poor measure of whether AI is improving engineering productivity.
Winters’ comments on computer science education also matter because he helped work on updated ACM/IEEE undergraduate curriculum guidelines. The interview frames a challenge for educators and employers: prepare new engineers for current practice, including AI tools, without reducing software engineering to producing more code.
Products Worth Considering
Apple iPad Air (4th Generation, 10.9-inch, Wi-Fi, 256GB) – Space Gray
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Google’s C++ Migration to Curriculum
The interview draws on Winters’ work at Google, where the source says he joined around 2010 and later led the readability program. The C++ refactor was described as a broad effort to replace an old API across a large, actively used codebase without stopping development. The report compares the challenge to replacing the foundation of a building while it remains in use; that is Winters’ analogy, not a technical measurement.
The source also places Google’s testing mandate in the context of Search release delays. It says that around 2007 the product could no longer ship monthly and was being released quarterly. The company’s requirement that changes include tests was presented as a response to that difficulty. The report says the release cadence later improved, but supplies no exact dates or figures for the recovery.
Winters’ curriculum work is part of a broader conversation in the episode about what new engineers need to learn and how they can distinguish themselves. The supplied report introduces that topic but does not provide the full curriculum proposals or the complete set of points from the interview.
“We were replacing the foundation of the Empire State Building. We were not allowed to close the building or block traffic. Oh, and we discovered nuclear waste in the basement.”
— Titus Winters, recalling the Google C++ refactor, as quoted by The Pragmatic Engineer
Limits of the AI Productivity Figures
The source does not provide the interview’s publication date, a full transcript of the AI discussion or supporting data for the reported 30% rise in outages. The figure is described as Winters’ account of what he hears from engineers at other companies; the timeframe is compared with a year earlier, but the report does not identify a sample, methodology or baseline dataset. It should not be read as a verified industry-wide statistic.
The supplied material ends partway through the article’s list of takeaways, during a discussion of AI benefits. It does not include the full account of those benefits, Winters’ detailed curriculum recommendations, or any response from Google or other companies. The article also does not establish that Google’s testing mandate alone caused the improvement in Search’s release cadence.
The Full Interview and Education Debate
The interview is available through The Pragmatic Engineer’s podcast channels, including YouTube, Apple and Spotify, according to the source. Its transcript and episode timestamps are also listed on the original report page. Those materials are the next place to check for Winters’ full comments on AI, curriculum changes and how early-career engineers can stand out.
No policy change, new Google engineering program or formal curriculum adoption is announced in the supplied material. The next developments to watch are whether Winters’ proposals appear in the updated ACM/IEEE guidance and whether further evidence supports the reported relationship between increased AI-assisted code output, outages and feature delivery.
Key Questions
What is the news?
The Pragmatic Engineer published a podcast interview with Titus Winters about his Google engineering work, AI and computer science education. The source does not give a publication date.
Who is Titus Winters?
Winters spent more than a decade at Google, where he led its code-readability program and worked on Abseil. He is now a Senior Principal Scientist at Adobe, according to the report.
What happened in Google’s C++ refactoring project?
The report describes a team of about a dozen people using parsing and automated code-editing methods to update an API across a codebase of nearly 250 million lines of C++. The work was done while the codebase remained in active use.
Did the interview prove that AI is increasing software outages?
No. The report attributes to Winters an observation from conversations with engineers: outages were said to be up about 30%, while feature shipments were static and code output had risen sharply compared with a year earlier. It gives no sample or methodology, so this is not established as an industry-wide finding.
What does the interview say about computer science education?
It discusses Winters’ work on updating ACM/IEEE undergraduate computer science curriculum guidelines and the need to reconsider what new engineers should learn. The supplied source does not include his complete recommendations.
Source: newsletter.pragmaticengineer.com
