VN — VnExpress

Anthropic Manager Drops Coding Tests in Hiring as AI Outperforms Humans

AINovember 2, 20261 min read

Felix Rieseberg, a hiring manager at Anthropic, says the company has stopped giving coding tests to job candidates. He says AI models like Claude now write code better than most people, so these tests no longer show real skill.

What happened

Felix Rieseberg wrote on his blog that, starting in October, he stopped asking candidates to do traditional live coding interviews. He said the company's Claude AI models now write code better than humans, so a coding test does not measure much anymore. He is now looking for new ways to find and judge young engineers.

Key numbers and facts

Rieseberg stopped traditional coding interviews starting in October.
A Karat survey of about 400 technical leaders in the United States, India, and China, done in 2026, found that 71% said AI makes it harder to judge a candidate's technical skills.
A CoderPad survey found that skills like system design, debugging, and working with AI are now valued more.

Why coding tests are changing

A coding test asks a candidate to write computer code in a set time. Companies use these tests to check if someone can solve problems. Now that AI can write code quickly, many companies are asking different questions. They want to know if a candidate can define a problem, design a solution, and spot mistakes in code made by AI.

Why it matters

This change shows how AI is reshaping the tech job market. If AI can do much of the coding, companies must look for other skills. Young engineers may need to show what they can build, think clearly, and care about good design, not just memorize rules or write code under time pressure.

What to watch

Whether other tech companies drop coding tests too.
What new hiring methods replace them, such as project work or AI-assisted tasks.
How schools and young job seekers change their training to match these new skills.

Glossary

Anthropic: An AI company that makes the Claude AI models.
Claude: A family of AI models made by Anthropic that can read, write, and work with computer code.
Karat: A company that runs interviews and skill checks for technical hiring at other businesses.
CoderPad: A company whose tools let employers test and interview software engineers online.

SenseTime open-sources 9-billion-parameter AI model that reuses layers for deeper processing

SenseTime has released SenseNova Loop-9B-Preview, an open-source multimodal model announced October 30 that reuses computation to extend processing depth without duplicating weights. It accepts text and images and generates text for document understanding, coding, and tool-use tasks, running a shared block of 24 decoder layers twice from a model of about nine billion parameters. SenseTime has published the weights, inference instructions, and training guidance under the Apache 2.0 license.

AITechNodeNov 2

DeepSeek team open-sources proactive coding agent research prototype

DeepSeek's Harness team has open-sourced dsh-superagent, a research prototype that takes development requests from GitHub and optionally Feishu, works on each in its own session, and returns results such as pull requests to the original conversation. The project, configured around a single GitHub repository, converts authorized requests into work items and includes a web panel for matters needing human attention. It is released under the MIT license.

AIVnExpressNov 2

Speakers of Less Common Languages Pay More for AI

AI models consume more tokens when processing less common languages, putting those users at a disadvantage compared with English or other widely used language speakers. A CLEAR study cited in the article calls this the "hidden tokenizer tax," finding that among the 20 most common internet languages, the most expensive needed 1.5 to 5 times as many tokens as the cheapest, a gap that grew to a 17-fold difference across 200 languages.

AI전자신문Nov 2

Doubts Over AI Investment Returns Spread; Wall Street Hedges Risk with Put Options and CDS

Wall Street is moving to guard against losses from massive AI investment as doubts grow over whether returns will meet expectations. According to the Wall Street Journal, individual investors and tech executives are using put options and credit default swaps (CDS) to prepare for asset price declines, and demand for hedging against falling GPU rental fees is also rising. Put options on CoreWeave have surged, with trading concentrated in 2028-expiry, $35-strike contracts, while CDS premiums for five major tech companies rose over the past month to their highest levels since 2021.