AI “Cheating” Lands Engineering Internships!
Chungin “Roy” Lee, a Korean-American student (hereafter referred to as Roy), successfully secured internship opportunities at Amazon, Meta, and TikTok. However, he was in no rush to choose which company to intern at—because he had no intention of joining any of them. His participation in the interviews was merely to demonstrate that the interview methods used by tech companies are outdated.
He is now organizing a startup, recruiting team members to sell the tool that helped him land these tech internships: an AI application that assists in solving problems during remote interviews.
Roy used this AI tool, “Interview Coder,” which he created in less than a week, to cheat his way into receiving internship offers from multiple tech companies.
I cheated my way to an Amazon offer and it was the easiest shit ever.
Leetcode interviews are dumb and have wasted millions of developer hours.
Help us get rid of them: https://t.co/d98c3qNiJ9
How Does Interview Coder Help Applicants “Cheat” and Land Tech Company Offers?
The operation of Interview Coder is simple. Essentially, it’s like taking a picture and then asking ChatGPT, “Can you solve this problem?” Roy explains that the programs used by tech companies to monitor applicants’ screens cannot detect Interview Coder. He used this tool to successfully secure multiple tech internship offers. “I recorded the interview process with Amazon as a final demonstration of the product to show that this method works, and that the recruiting process has been cracked,” Roy said. The video documenting the interview process garnered over 100,000 views before being taken down by YouTube due to a copyright claim.
Roy once commented on X, stating that he had spent over 600 hours practicing Leetcode, calling it the most painful part of his learning to code.
Amazon Furious and Pressures Columbia University
After Amazon learned the truth, they were furious and pressured Columbia University, where Roy is enrolled, to discipline him. As a result, Roy was suspended for a year. However, he took this in stride, even claiming that he might have left school on his own accord earlier. Roy mentioned that he had never seriously considered working at Amazon, Meta, or TikTok, and that his goal was merely to prove that the technical tests used by tech companies were outdated and caused unnecessary suffering for applicants.
Roy now offers Interview Coder as a subscription service for $60 per month and has enlisted influencers to market this AI interview tool. He expects to reach a $1 million annual recurring revenue (ARR) by mid-May this year.
Why Did Leetcode, a Programming Test, Become a Nightmare for Engineers?
Roy had once dreamed of working for a tech company, but his views changed after spending a staggering 600 hours practicing on Leetcode. “This was the most painful experience in my learning to program,” he said. The process of solving problems on Leetcode was grueling, and “technical interviews have been this way for the past 20 years, it’s absurd.” Leetcode is a software used by many tech companies to evaluate the technical skills of engineers, providing a large number of problems for engineers to practice online. Companies like Meta, Google, and Amazon all use Leetcode to filter applicants. Due to Roy’s aversion to Leetcode, the homepage of Interview Coder’s official website features the bold statement “F*ck Leetcode.”
Interview Coder’s Official Website Features the Bold “F*ck Leetcode.”
Roy pointed out that many of Leetcode’s problems are excessively difficult and disconnected from the actual tasks engineers face in their day-to-day work. “The key is whether you’ve seen the problem before, remembered the solution, and can act as though it’s your first time seeing the problem,” he emphasized. “Many problems are algorithmic, which doesn’t represent the work an engineer does.” In fact, there are many engineers who share Roy’s negative view of Leetcode. According to a report by CNBC, Meta engineer Ryan Peterman wrote in a Substack article last December, “Every time I bring up interviews, I hear frustrating comments about Leetcode.” He attempted to explain to readers why Leetcode is so despised, yet still widely used by tech companies.
Interview Coder is not the first tool to provide “cheating” services for interviews. Leetcode Wizard’s marketing director, Isabel De Vries, said, “Our product originated from the frustrations many users encountered with the same problems.”
The Rise of Leetcode and the Current State of Engineering Interviews
However, services like Interview Coder can be effective because many tech companies transitioned to online interviews during the pandemic, a practice that has continued due to its time and cost efficiency and its ability to reach a wider pool of candidates. Initially, this approach was convenient and effective, but with the advent of large language models, AI now has the ability to assist engineers in generating code.
AI has already become a valuable assistant for engineers. According to the 2024 Stack Overflow survey, over 60% of engineers use AI during the development process, and another 14% plan to use it in the future. François Fleuret, a research scientist at Meta’s FAIR AI team, revealed that when he asked software companies how much of their code was generated by AI, the answer was “rarely less than 40%, and for some younger engineers, over 90%.”
AI-Assisted Interviews: Is This the Future?
In fact, many tech companies have already openly embraced AI technology. Sundar Pichai, CEO of Google, disclosed in last year’s earnings call that over 25% of the new code at the company is generated by AI. Brian Armstrong, CEO of cryptocurrency exchange Coinbase, also stated that all of the engineers at the company are using AI development tool Cursor.
In such a widespread use of AI, Roy believes it makes no sense to conduct interviews without considering AI usage. “Tech companies should adjust their interview processes to account for the widespread use of AI, instead of punishing applicants who use AI,” he stated.
Roy argues that Leetcode is a burden for engineers. “Tech giants have no incentive to change. Leetcode is a poor system that benefits them, but it has a huge negative impact on the global developer ecosystem,” Roy emphasized. “This attempt to standardize problem-solving tests is outdated in today’s world.”
Google Considers Returning to In-Person Interviews to Avoid AI Impact
AI has already become a common assistant in the job application process. Henry Kirk, founder of entrepreneurial solutions company Studio.init, mentioned that in June last year, he hosted an online programming competition to recruit engineers, receiving 700 applications. However, he discovered that more than half of the participants had cheated using methods, including AI.
The impact of AI has even led Google to consider returning to in-person interviews. CNBC reported that “whether to resume in-person interviews” became one of the discussion topics at Google’s employee meeting in February. At that time, recruitment VP Brian Ong mentioned that the company was working to address the impact of AI on interviews, a concern “being watched by all competitors.” Pichai suggested switching to a hybrid interview format, “Since we have a hybrid office, I think we could consider making some interviews in person, which would also help candidates understand Google’s culture, benefiting both sides.”
Ironically, this year, Claude’s series of models developer, Anthropic, implemented an “AI policy” for applicants, asking them not to use AI technology during the job application process. “While we encourage people to use AI systems to improve work efficiency, please do not use AI assistants during the application process. We want to understand your interest in Anthropic and your communication skills without AI interference.”
Job applications are not as simple as exams in school, making it difficult to reach consensus on whether it is wrong to use AI tools to enhance performance. Regardless of whether this is a shortcut that undermines corporate trust or an approach to drive innovation in tech industry interview processes, the impact of AI on the job application process is bound to continue.
Article republished from: Digital Era
Sources: CNBC, Gizmodo, Ubergizmo