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
A Pragmatic Engineer podcast episode features Titus Winters discussing Google’s engineering practices, including a large-scale C++ refactor, mandatory tests and its readability program. Winters also discusses AI’s uneven effects on productivity and argues that computer science education needs updating; the supplied report does not include the full details of his curriculum proposals.
The Pragmatic Engineer has published an interview with Titus Winters, a former Google engineer and lead author of Software Engineering at Google, about large-scale refactoring, testing, AI and computer science education. The discussion offers a practitioner’s account of how Google managed software changes across a large codebase and raises questions about whether AI-generated code is improving engineering outcomes.
Winters spent more than a decade at Google, where he led its readability program and helped build Abseil, Google’s open-source C++ libraries. The report says he is now a Senior Principal Scientist at Adobe, focused on developer experience and productivity. The podcast’s host describes him as a contributor to the ACM/IEEE undergraduate computer science curriculum guidelines, which the interview also addresses.
One central example is a refactoring effort at Google in 2010, when the company had close to 250 million lines of code edited by about 10,000 engineers, according to the report. A dozen-person team used language-parsing techniques to identify and change old C++ patterns across the codebase. The work included parsing build targets into abstract syntax trees, generating edits, making old and new APIs coexist, and distributing changes in batches. The report says the team needed tooling that led to the creation of clang-format.
Winters also described Google’s decision to require a test with every change after Search struggled to maintain a monthly release cadence, according to the report. The article says the product had fallen to quarterly releases and later regained a faster cadence. It also describes a review system requiring approval from both a project or file owner and a language-qualified readability reviewer; more than 10,000 engineers were enrolled in the program when Winters left Google, the report says.
Engineering Practices at Google
The examples in the interview show how testing, code review and automated tooling can serve as infrastructure for changing software safely, rather than as checks added only after coding. At a company operating across a large shared codebase, a change can affect more than its immediate team. The report’s account of Google’s testing mandate links reliability practices to the ability to ship more often.
The interview’s AI discussion adds a current productivity question: whether faster code production leads to better software or simply more work to maintain. Winters relays observations from engineers he speaks with, rather than presenting a measured industry-wide study. That distinction matters for readers weighing claims about AI’s effect on engineering teams.
The education discussion is relevant to students, educators and employers because it concerns what undergraduate computing programs should teach as software development practices change. The supplied report introduces Winters’ call to overhaul much of the curriculum but does not provide enough detail to assess specific proposed changes.
Products Worth Considering
Samsung Galaxy Tab S10 FE 128GB Wi-Fi Android Tablet, Blue
Apple AirPods Pro 3 Wireless Earbuds
Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade
As an affiliate, we earn on qualifying purchases.
From Google’s Codebase to Classrooms
The podcast’s central thread is the challenge of maintaining software while it continues to change. At Google, the refactor described in the report involved code written over many years and used by thousands of engineers. The team could not simply pause development, Winters said, so it used program analysis and staged replacements to manage the work alongside ongoing changes.
The report also links code review to education inside the company. Google’s readability program trained engineers to review changes for language-specific conventions, while project owners supplied local knowledge. The host characterizes the effort as a large educational program, but that description is not independently quantified beyond the report’s figure of more than 10,000 enrolled engineers.
Winters’ comments on education draw on his work on ACM/IEEE undergraduate curriculum guidance. The source frames the apparent contradiction between helping shape curriculum guidelines and later arguing that much of the curriculum should be discarded as a topic of the conversation. It does not reproduce the guidelines or explain which parts Winters believes should change.
“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 Google’s 2010 refactoring effort, as quoted in The Pragmatic Engineer
Limits of the AI Evidence
The report says Winters has heard from engineers at other companies that code output is up substantially from a year earlier, outages are up about 30%, and feature output is static. It does not identify the companies, define how “code output” or outages were measured, provide the period behind the outage figure, or specify a comparison baseline. The observation should not be treated as a representative industry statistic or proof that AI caused the outages.
The supplied source ends partway through its summary of the interview. It does not give Winters’ full account of AI’s potential benefits, the specific curriculum changes he recommends, or the complete podcast transcript. It is also unclear from the excerpt when the episode was published.
The Full Interview’s Open Questions
The podcast episode is available on YouTube, Apple and Spotify, according to the report, and its transcript and timestamps are listed on the episode page. Those materials may provide the fuller explanation of Winters’ views on AI, curriculum reform and how new engineers can distinguish themselves.
For now, the supplied report supports a narrower takeaway: Winters describes concrete engineering systems used at Google and raises concerns about how software teams measure the results of AI-assisted development. Further evidence would be needed to establish whether reported changes in outages or productivity apply broadly, and the detailed case for revising CS education remains to be examined in the full discussion.
Key Questions
What is the news development?
The Pragmatic Engineer published a podcast interview with Titus Winters on software engineering at Google, AI and computer science education.
Who is Titus Winters?
Winters spent more than a decade at Google, where he led the readability program and helped build Abseil. The report says he is now a Senior Principal Scientist at Adobe.
What did the Google refactoring project involve?
The report describes a 2010 effort to update old C++ patterns across a codebase of close to 250 million lines. A team used parsing, generated edits and staged changes while the systems remained in use.
Does the report prove AI is increasing outages?
No. It relays Winters’ account of what engineers at other companies have told him, including an outage increase of about 30%. The report gives no measurement method, named sample or clear comparison window, so it does not establish a broad trend or causation.
What changes to CS education does Winters recommend?
The supplied report says Winters argues that much of the existing curriculum should be overhauled, but it does not detail his specific recommendations. The full interview or transcript would be needed to identify them.
Source: newsletter.pragmaticengineer.com
