Anthropic engineer Felix Rieseberg stops using coding interviews, citing AI capabilities

Felix Rieseberg, an engineering manager at Anthropic, announced in a blog post on Tuesday that his team has ceased employing conventional coding interviews. He explained that while the capacity to quickly formulate problems and devise elegant solutions once served as a helpful, albeit minor, indicator of future team success, "Small coding challenges in real code did tell us something meaningful about candidates," he noted, but this is no longer the case. He firmly maintained, "There’s no more point in giving you a little React challenge than there is in quizzing people about assembly; Claude is now better at both," adding that "Our most impactful builders are successful regardless of how good they are at these tests."

Echoing this sentiment, Jaana Dogan, a principal engineer at Google, shared a similar perspective on social media, stating, "I’m not doing a ton of coding interviews nowadays but I’d be interested to interview how people use coding agents," because that usage itself provides more insight into a candidate's coding aptitude. Several months ago, Google reportedly initiated a review of its hiring methods regarding coding interviews and the qualities it seeks in applicants. While it would be surprising if similar discussions hadn't already taken place within other major tech firms where AI agents have demonstrated coding proficiency, the broader technology industry appears to be approaching this transition with greater caution than the fervent AI proponents. Nicole Bachaud, a labor economist at ZipRecruiter, stated in an email to The Register that "We are not seeing tech companies abandon skills assessments widely across the industry—at least not yet." Furthermore, according to their October 2026 Employer Survey of 1,500 employers, the information sector specifically stands out as the most inclined to require formal skill evaluations during initial applications, with 62.5 percent of employers doing so.

The nature of coding interviews at various tech behemoths, including Microsoft, Google, Amazon, and Facebook, differs based on what the hiring entity wishes to measure. These assessments might explore deep knowledge in a specific domain, proficiency in a certain programming language, general aptitude for problem-solving, or even collaborative behaviors. For instance, Educative describes Stripe's software engineer hiring sequence as involving an initial phone screening, a technical phone interview, and several subsequent on-site meetings. Rieseberg, having previously worked there, recounted how founder Patrick Collison encouraged new hires to observe anything that seemed unusual. He mentioned, "His bet was that newcomers would spot friction and opportunities that old-timers had long stopped noticing." Rieseberg seems to have adopted this philosophy in dismissing established experience, declaring, "Hunger beats tenure," although he may not be aware that technology corporations have faced legal issues for discarding valuable veteran staff in favor of younger, less costly professionals.

It is important to remember that technical interviews rarely focus solely on technical expertise. Many avenues exist for employees to contribute value to an organization beyond skills quantifiable through computer science tests. When personnel have access to AI models that surpass job candidates in answering technical queries, it becomes logical to examine the creativity and interpersonal abilities individuals bring to the table. Bachaud suggested that "The change at Anthropic might be reflecting a shift away from technical skills towards more human, soft skills." She posits that as AI achieves full coding capability, tech employers might pivot away from testing rote technical knowledge and focus heavily on human-centered soft skills. Although AI may write code competently, she maintains that it cannot replicate critical thinking. She pointed out that "A staggering 61.5 percent of Information employers now report that 'critical thinking' is a primary skill lacking in today's candidate pool."

Rieseberg firmly stated, "I don’t think the coding interview is going to come back," reflecting on the sentimentality surrounding the proliferation of "Algorithms and Data Structures" books purchased for career advancement, or viewing the end of the coding interview as another marker in software engineering's progression toward higher levels of abstraction. However, he admitted uncertainty regarding its replacement, writing, "I’m probably not the only manager trying to figure out how to interview for people who build well with AI." He added that he has not yet found a definitive method, but he is currently reviewing resumes and portfolios for small prototypes that display thoughtful execution, empathy, and principled reasoning, and he invited feedback: "If you have figured out something better, I’d love to hear how."

Bachaud contends that technical evaluations will evolve rather than disappear entirely, noting that 70 percent of employers in the Information sector have adopted proctored or in-person skill assessments to mitigate AI cheating. She concludes that "Ultimately, tech companies aren't necessarily testing less, but the way they are testing is changing." As AI automates many manual and technical job functions, work skills may become more universally applicable, potentially leading to increased job turnover once economic circumstances improve. In a recent survey of workers, they discovered that concerns related to AI are causing one-quarter of employees to begin contemplating their next professional move, and as evaluation methods and required skills shift, the roles available to workers may appear more distinct than ever before.

Why it matters

This signals a potential shift in hiring priorities across the industry, moving away from rote technical testing toward evaluating human-centric skills like creativity. Developers should watch how companies adapt their assessment methods as AI tools become more capable in coding tasks.