AI Broke Recruiting

For both job hunters and recruiters

AI Trumpets

Hail OpenAI and their ChatGPT which was publicly released on November 30, 2022. It was only 3 years ago! I haven’t even paid off my brand new auto loan yet from a car I purchase at that time. The business and societal impact has been gynormous!

We love it because it helped us write emails and posts quickly and more eloquently, generated images for such content without paying royalties and wrote our code for us.

We also hate it because it’s diminishing job opportunities especially for new grads and older folks.

Dark Side of AI

There is no end to the controversies of the use of AI. As the story continues to unfold including the copyright infringement of the training data on which it was trained, the danger of AI “leaking” your private and company proprietary data, the astronomical high carbon footprint to support all the AI work, and the displacement of the poor and disadvantaged people to build AI data centers, I’m sure there will be more to come. Not the topic of this post but if you’re passionate about AI, I recommend reading Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI by Karen Hao.

Issues with AI in Recruiting

As a Coach

One of my passions is to mentor people who want to grow their careers. You’ve probably all heard about that you’ll need to customize your resume for every job. It’s a sound advice. Don’t put in your Microsoft certifications and clutter your resume if the job is asking for AWS experience. Get good referenceable projects to build your portfolio and internships under your belt.

One of my friends, who is also a head of data, and I were lamenting that with over hundreds or thousands resumes, everyone looks so good. However when you talk to the person, it is obvious that a majority of those folks are unqualified or under-qualified. My friend’s search for a senior machine learning engineer who can do a bit of data architecture & infrastructure has been on-going for the past 8 months!

As a job seeker, how do you stand out in a crowded AI-generated resume landscape?

As a Hiring Manager

I run a data team for a growth company and am a big part of the hiring team. I’ve read probably tens of thousands of resumes at this point. Read is a loose concept I’m using here. No one has time to read so many resumes. When applying to jobs is so “Easy Apply” in LinkedIn, it’s hard to actually read the 2-5 pages of resumes per applicant. I have no choice but to use some AI automation and vetting myself.

AI-to-AI Process

So here we are AI (submission)-to-AI (vetting) end-to-end recruitment process. Do you love it? Do you like applying to hundreds of jobs and not hear back from anyone? Feel ghosted? From my perspective, I apologize in advance that I cannot get back to everyone but it’s galactically impossible to respond to everyone. As a comparison, quantum computers are possible, replying to every candidate is not.

Ghost Engineers

Due to AI filtering, I see more Ghost Engineers applying to jobs. This is a phenomenon where a person with a valid work authorization applies and gets the job but the person who actually logs in and do the job are different people. Beware that those real people are tagged in system so those folks will someday actually look for a job, well, sorry they won’t pass the filter. There are telltale signs for Ghost Engineers. They don’t put their LinkedIn profile in their resumes, has no references and their resume reads like they can do everything. Here is an actual job title “Senior Data Analyst Scientist and FullStack Developer” I’ve seen.

Github Code and Side Projects

Github code and side projects posted after 2023 are no longer a reliable source of the candidate’s ingenuity, creativity and perhaps problem solving skills. We have to believe they’re most likely done by AI vibe coding.

Take Home Evaluations & Live Coding Interviews

Take home projects gone. We used to give potential candidates a take-home project; let them have the time to think through a problem and write up a proposal at the comfort of their home and not under the scrutiny of the interviewer in real-time.

Now with AI coding, we don’t give take-home evaluations anymore. We perform live coding or timed coding platforms (i.e. LeetCode or HackerRank). I recognize some people don’t perform well under those conditions (myself included) but what can we do? AI broke the take-home project evaluation. Many rounds of interviews is the new standard to ensure you are a real human able to solve problems. It’s not ideal for the hiring manager or the job seeker. But that’s where we are.

AI Overlord & Conclusion

Once the genie is out of the bottle, there is no putting it back. AI is here to stay. I welcome you, my AI Overlord!

In my coaching, I highlight the human-to-human networking is still vitally important. Go to in-person meetups and local conferences to meet your peeps, your people, your tech kins. Reconnect with your alumni school, high school or college friends (or uni for our non-American readers).

Highlight your soft skills. I once gave this advice to someone and he said “Pfff, soft skills are for xxxxx” (sorry, I can’t say that here). Now that I’m CDO and run an organization myself, I mostly use my soft skills to persuade, to encourage and to suggest someone/team to do my bidding (evil laugh). I cannot tell people outright what to do. That’s what a micro-managers do and I’m not one so I have to flex my soft skills.

In my experience:

• Individual Contributor (80% tech skills + 20% soft skills)

• Manager (20% tech skills + 80% soft skills)

We just hired a new person for our data team. His resume stood out to me. I asked him “Which AI tool did you use to write your resume?” He said he wrote it himself after reading our job description. He didn’t use AI. Ah ha! I thought. That’s why it stood out. He really made an effort to create a representation of himself that’s real and not a bunch of AI slop.

For whatever it’s worth, write your own resume, give it a try.

Party like it’s 1999! (Prince)