Read Time 4 mins | Written by: Ryo Hang
The Challenge of "Smart" Candidate
The rise of artificial intelligence (AI) has introduced a new challenge in the recruitment process: dealing with fake resumes and candidates.
We have been flooded with fake resumes and fake candidates every day. I wrote a blog post talking about how to filter resumes. With "Smart AI" joining the force, identifying authentic candidates becomes more challenging.
In this blog post, we will discuss effective strategies for filtering out inauthentic candidates and unmasking AI-assisted interview responses. Equipping your hiring team with these insights will enable them to make informed decisions and focus on genuine candidates.
Bottom line: The candidates become "smarter", and so should we.
About us:
ASCENDING is an AWS Certified Advanced Consulting Partner. Our IT staffing service was backed by our seasoned in-house engineers, who screen 100-200 candidates per week for our clients.
Is it worth reading?
Having served numerous hiring managers from enterprises and SMBs, we understand the challenges of talent recruitment in the tech industry. If you often deal with templated resumes or encounter fake candidates, this blog is a must-read. We aim to share our experiences and help you avoid unpleasant encounters.
Screening AI-Enabled Tech Candidates
After interviewing several candidates who utilized AI assistance, I felt compelled to write this blog. As frontline soldiers in the battle against AI-enabled candidates, we face increasing difficulties. As an interviewer, I usually ask questions relevant to a candidate's experiences, but I soon realized I needed different approaches to uncover the truth.
Here are three steps to follow, and remember to turn on the camera FIRST!
Stay Alert for AI response:
There are typically two ways to screen tech candidates:
Code Challenge:
AIs like ChatGPT can efficiently solve code challenges, providing answers within seconds if you ask the right questions. However, you should be wary of candidates who never rewrite their code to solve a problem. A genuine engineer would attempt multiple approaches, discarding previous solutions for better ones.
Project Experience:
Inquiring about project experience is crucial for assessing a candidate's ability to provide immediate value to an employer. If the candidate processes a similar experience, it can create a lot of value for the employer to complete work efficiently in the short term.
If the candidate only provides surface-level answers, it may indicate a lack of actual working experience. For instance, if you inquire about CI/CD experience, a genuine candidate would delve into specific tools, steps, and share detailed insights. Superficial responses should raise red flags, prompting further investigation.
Overload the Brain:
Using AI has become commonplace in enhancing productivity and efficiency. However, our brains require time to digest information and generate/select appropriate solutions. Within short intervals, our brains are constantly overloaded with information. That's the trait, my team attacks during the screening process. Here are a few ways, feel free to come up own.
- Pinpoint a Complex Code Section and Ask About Thought Process:
In a stressful interview environment, humans find it challenging to process new content quickly. It takes me no time to explain why I came up with a particular code, as opposed to understanding someone else's code. By attempting a few code challenges, you will observe how fake candidates struggle.
- Follow up on Test Scenarios for a Block of Code.
Similar to the previous strategy, when an engineer presents a block of code, they should have considered various scenarios, including edge cases. Asking about test scenarios becomes simple for someone who wrote the code themselves. However, fake candidates will need to review the code again and contemplate the test scenarios. Reading someone else's code is never an easy task. :)
- Dig Deep into Specific Project Experience:
Asking about project experience is my preferred method for filtering AI-assisted candidates. I once identified unqualified candidates within 7 minutes using this approach. It requires extensive knowledge of a subject to ask the right questions though.
For example, if we discussed CI/CD earlier, I would follow up with questions about how to package a Java application? How to run unit tests? How did you deploy an application into a microservice cluster. You can drill deeper into each question and evaluate a candidate's ability to breeze through them, leveraging their previous experiences to create maximum value in the short term.
- Ask Open but Obvious Questions
In addition to delving into project experience, occasionally ask open questions that should be fairly common challenges for candidates with relevant experience. For instance, inquire about the rollback process in case of deployment failure or how they monitor failed releases.
Be Lenient with Genuine Candidates, Make Fake Candidates Struggle
In general, we are lenient with genuinely honest candidates who accurately represent their skills on their resumes. Our team never uses a set of template questions to screen engineers since we believe that each talent is unique. As long as candidates are honest about their project experiences, we can evaluate their skills effectively.
The screening process is designed to filter out fake candidates and those who exaggerate their skill sets on their resumes, making them struggle.
Conclusion:
Identifying authentic candidates in the era of AI presents a unique challenge for hiring managers. By staying vigilant, employing specific strategies, and conducting thorough interviews, you can ensure that your team spends time with genuine candidates who possess the skills they present. With these insights, you can make informed hiring decisions and build a talented team.
