AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

Felix Rieseberg, an Anthropic hiring manager who has led work on Claude Cowork and Claude Desktop, says his team has stopped asking software engineering candidates coding questions. The change applies to his team; the source does not establish that Anthropic has ended coding interviews across the company.

Felix Rieseberg, an Anthropic hiring manager who has led work on Claude Cowork and Claude Desktop, says his team has stopped asking candidates for software engineering jobs to answer coding questions. The team found that data-structures-and-algorithms interviews gave it “zero signal,” according to The Pragmatic Engineer, which reported the change. The decision is described as a team-level change, not a company-wide end to coding interviews.

The reported change concerns a familiar part of technical hiring: asking candidates to solve coding problems during interviews, often focused on data structures and algorithms. Rieseberg’s team has removed that step from its recruitment process because it judged the exercise uninformative for its hiring decisions. The source does not provide details of the team’s replacement interview format or how it evaluates candidates instead.

The claim comes from Rieseberg’s blog post, cited by The Pragmatic Engineer. The newsletter characterizes Anthropic as a leading frontier AI company and says other teams at Anthropic, as well as AI-native startups, could follow. That is a prediction, not confirmation that those organizations have changed their hiring practices.

The report does not name the number of candidates affected, say when the policy took effect, or present the evidence behind the team’s “zero signal” assessment. It also does not establish whether the decision covers every engineering role on Rieseberg’s team or all stages of its hiring process.

At a glance
reportWhen: Reported in The Pragmatic Engineer’s Pu…
The developmentAn Anthropic engineering team has removed coding questions from its software engineering recruitment process after concluding that data-structures-and-algorithms interviews provided no useful hiring signal.

How Engineering Hiring Could Change

The decision challenges the value of a widely used hiring screen. If a team concludes that coding puzzles do not help distinguish candidates who can do the job, continuing to use them may add time and frustration without improving selection. For applicants, removing that interview stage could change how they prepare and what skills employers ask them to demonstrate.

But one team’s experience does not show that coding interviews are ineffective everywhere. The report offers no comparative hiring results, such as how candidates selected without coding questions perform after joining. The practical significance is narrower: an Anthropic team has publicly described dropping the format, placing renewed attention on how software engineers should be evaluated as AI changes development work.

The decision may also prompt employers to reconsider whether interview exercises resemble the work their engineers actually do. That is a broader implication, rather than a reported policy shift across the industry. The source says other AI-focused companies could follow, but gives no examples of firms that have already done so.

Products Worth Considering

DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)

DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)

Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade

Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade

As an affiliate, we earn on qualifying purchases.

A Team-Level Hiring Decision

The report appears in The Pragmatic Engineer’s “The Pulse,” a series covering developments and trends in technology companies. Its item on Anthropic focuses on the hiring practice described by Rieseberg. It does not say Anthropic has issued a company-wide policy or that its overall engineering recruitment process has been redesigned.

The distinction matters because hiring methods can vary by team, role and interviewer. A change by a manager leading particular products may indicate a local assessment of interview usefulness, but it cannot establish how other Anthropic teams assess software engineering candidates. The source also does not specify what kinds of coding questions were removed or whether candidates still complete practical technical exercises in another format.

The newsletter places the item alongside other technology workplace developments, including AI spending and performance expectations at other companies. Those items provide industry context, but they do not show that Anthropic’s decision was caused by AI tools or by a change in software engineering work. No such explanation is supplied in the material.

““zero signal””

— The Pragmatic Engineer, summarizing the team’s finding

Scope and Evidence Still Unclear

The available report does not say whether Anthropic’s broader engineering organization has reviewed or adopted the team’s approach. It also does not identify the team’s replacement assessment, provide a timeline, or explain how it measured whether coding interviews produced useful hiring information.

There is no reported comparison of candidate quality or later job performance before and after the change. The phrase “zero signal” is attributed to the team’s conclusion, but the source material does not include the data or methodology supporting it. It remains unclear whether other companies will follow, and whether this is a temporary experiment or a lasting change.

Whether Other Teams Follow

The next meaningful development would be clarification from Anthropic on whether the change is limited to Rieseberg’s team or reflects a wider hiring policy. Further detail about the team’s alternative interview process and the evidence behind its decision would help assess what has changed for candidates.

The Pragmatic Engineer predicts other Anthropic teams and AI-native startups could adopt similar practices, but reports no confirmed follow-on decisions. Until companies or hiring managers announce such changes, the reported development remains a specific team’s approach rather than a broader industry shift.

Key Questions

Has Anthropic ended coding interviews company-wide?

No such company-wide change is confirmed in the report. It says Rieseberg’s team stopped asking coding questions for software engineering candidates.

Why did the team stop asking coding questions?

The team concluded that its data-structures-and-algorithms interviews provided “zero signal,” according to The Pragmatic Engineer’s account of Rieseberg’s blog post. The source does not include the supporting data.

What will replace the coding interview?

The supplied report does not describe a replacement interview or assessment method.

Are other companies making the same change?

The Pragmatic Engineer says other Anthropic teams and AI-native startups could follow. It does not report confirmed changes by those organizations.

Source: newsletter.pragmaticengineer.com

You May Also Like

Show HN: Bigwords.page – Turn Any Screen Into A Sign. The URL Is The App

Bigwords.page is a browser-based tool for displaying signs, messages, slides and countdowns from a link, without an account or app.

Why Malicious Claude Extensions Reached Thousands In A Supply Chain Attack

A WIRED headline reports a supply chain attack involving Claude extensions, but the affected tools, verified scale and user impact remain unclear.

Can Ironclad Help Move Computer Use Forward?

OpenAI has a page titled “Advancing computer use with Ironclad,” but the available information does not establish what the project is or what it does.

24 Ways To Explore Jev For AI Decision Support

Thorsten Meyer outlines 24 Jev use cases, with three live in his publishing operation and others rated strong, unproven or poor fits.