Jason Margolin’s Post

This is interesting. AI has its place in the job hiring process, but if it’s masking too much (including language proficiency), it can sometimes feel like getting catfished. In person interviews are great when possible.

> Reports suggest that over 50% of candidates are now suspected of using AI tools off-camera to solve real-time coding challenges during remote interviews. Wait... isn't this one of the companies pushing the use of AI for just this type of programming work? 🤔 I remember 20 years ago, my CTO at the time having a real panic about interviews and "what if they just google all the answers?". The aim of an interview isn't to create some artificial, sterile room to stress test a candidate (though you'd think it was...), it's about seeing how they use the tools available, the tools they'll use on the job, respond to challenge and approach problem solving. All of this can be done without worrying about the impact of AI, because AI isn't your problem, your problem is how you structure your interviews.

Going back to in-person interviews is understandable when you can't verify the process remotely. It's practical. But it's a retreat, not a solution. The real challenge isn't stopping candidates from using Ai-it's rethinking how we evaluate engineers in a world where AI is part of the workflow. If someone can effectively use AI to solve problems, isn't that the skill we actually need? We can't keep testing for yesterday's skills while hiring for tomorrow's roles. Instead of banning the tools, we should learn to assess how people think, decide, and work alongside them. Returning to the old ways buys us time. But adapting to the new reality is what we actually need to figure out-now, not later.

Google is going back to in-person interviews for software engineering roles! Reports suggest that over 50% of candidates are now suspected of using AI tools off-camera to solve real-time coding challenges during remote interviews. Some Google employees have even called on leadership to ban remote interviews altogether, arguing that they no longer reflect a candidate’s true fundamentals. This shift highlights an interesting dilemma... AI is transforming how engineers work, but it’s also disrupting how we evaluate engineers. Balancing fairness, authenticity and progress will be key for every company hiring in the AI era. https://lnkd.in/gPa7V7GB

Part 1: The New Reality of AI in Interviews 🚀 How to Be an Interviewer in 2025: Embracing AI in Hiring Hi community! 👋 I’m Denys Derkach, Principal Lead Software Engineer with 10+ years in software development. One of my passions is conducting interviews — understanding how people think, solve problems, and bring ideas to life. 💡 Key Shift: In 2025, AI in interviews is no longer science fiction ( especially considering remote work as a lifestyle now ). Candidates may: * Use AI responsibly as an assistant * Rely on AI to generate answers * Or not use AI at all 🔹 My Approach to Evaluating Candidates I categorize candidates into three groups: 1️⃣ AI-Augmented Candidates (responsible use) 2️⃣ AI-Dependent Candidates 3️⃣ Pure Human Candidates (no AI) 🔹 Formats to Assess Candidates 📚 Theory discussions 📝 Pre-defined technical assignments 💻 Live coding 🛠 Use-case exercises ( I personally found this to be a vital part ) 💬 Note for Part 1: These are just my thoughts 💡 and will be happy to hear your opinion on it in comments! In Part 2, I’ll dive into practical signs on evaluate candidate and see if we can marry them with a project! #EngineeringLeadership #TechHiring #Interviewing #AIAutomation #AI

🤯 The Real Gatekeeper in Tech Hiring Isn't the Code. Five years ago, I got this email from YouTube. It wasn't an offer, but the final, high-stakes step: Hiring Committee (HC) Review. It hit me: after months of grueling technical screens, system design, and behavioral interviews, the last stage was pure documentation and alignment. It was a different kind of pressure, and one developers rarely prep for. In 2024, where AI is streamlining initial screening, getting to the HC stage means your code is approved. Now, the challenge shifts from competence to package presentation. 💡 3 Non-Technical Hurdles to Clear the Final HC Review: 1. The "Why You" Narrative: HC members review your entire packet—interview notes, resume, and sometimes even social presence. You need a cohesive narrative proving you're the unique solution to their problem. Did your resume quantify your impact with $ or %? 2. The Transcript Trap: For early-career roles, they request final transcripts. This is a check on integrity and consistency. Lesson: Be transparent from day one; discrepancies are a red flag the HC will immediately flag. 3. System Design Clarity: Interviewers' notes must clearly articulate your architecture. You need to coach your interviewers (subtly!) by structuring your answers using established frameworks (like STAR for behavioral) to make their note-taking easy and HC-friendly. The best developers don't just solve the problem; they make the solution easy to sell to the ultimate decision-makers. The HC process is System Design for your career. What non-technical "final mile" challenge have you faced in a high-stakes tech interview? Share your strategy! 👇 #SystemDesign #TechHiring #DeveloperTools #AI #LearningJourney

When was the last time you heard candidates praising a technical interview process? A $500B+ publicly-traded company just adopted our next-gen hiring process, and candidates love the experience. Our next-gen hiring process includes evaluating • Fundamentals of software engineering without an AI assistant • A real-world task on a code repo with an AI assistant in a Cursor-like IDE experience • Reviewing code written by an AI agent When you interview developers in contexts that mirror real work, you get stronger signals and a better candidate experience.

You know, it’s been around for awhile lolol but now it’s becoming wide spread easier for people to stack ai tools to create the illusion they are actually taking the interview. Plus, it’s crazy how many companies allow the browser lockdown, that doesn’t work and is very undertested, I found bypasses within 2 minutes. You really need kernel level detections to stand a chance to block all 3rd party applications on a device, but the chances of all companies being able to have that level is probably near 0%. Plus go to Reddit and make a comment under any interview for x company post, wait till you see the amount of people trying to get insight. A couple guys even said they were handed the questions they would get in the final stage. Ultimately though in the rise of AI, which is only going to get better, the game has changed tbh, just not too sure how long companies can hold onto the processes pre-AI that clearly are getting broken easily by AI.

AI adoption disproportionally affects junior hiring. That's why you have to upgrade your career skills and crush interviews when you do get them. Socratify uses AI to build your judgement so your rapidly acquire the skills of more senior professionals

You’re Not Being Rejected by AI- You’re Being Rejected by Ambiguity If your LinkedIn says “U.S.” and your resume lists no real location, you’re not signaling flexibility, you’re signaling vagueness. Here’s the truth: most employers aren’t offering relocation. When they say hybrid, they don’t mean fly in for a few meetings. They mean you live close enough to show up when needed. If you’re on the West Coast, that’s a stretch-but, New York to Boston is potentially doable. And no- AI isn’t the reason you’re getting screened out. It’s frustrated humans who are tired of playing “guess the zip code" and hardcoding the immediate location. If you want to be taken seriously, make it easy for people to know where you are and where you are willing to travel to. Clarity wins interviews.