AI Took My Job, but Then I Used the Tech to Level up My Search

Mark Quinn is the senior director of AI operations for Pearl, an AI search platform for professional services. In a prior role at a startup, the arrival of OpenAI’s GPT-4 meant artificial intelligence could do the work of a team he was building. The following has been edited for brevity and clarity.

In my last job, I was at a startup. Before that, I was leading engineering operations at Waymo. It was a 3,000-person organization, a rocket ship all its own. In other words, my career was fairly well established and going well by most indications. So, when the startup came along, it was about this bigger swing and this even bigger opportunity, potentially, to help this company unlock what they were going after.

My main role in that was to lead what was the primary human-in-the-loop operation responsible for supervising and curating the AI.

When I joined, it was already a 500-person strong organization, and I was hired to ramp it to thousands. By all indications, we were doing the job really well. Then GPT-4 came out. After playing with it for just a couple of months, we realized that the bulk of the operation that I was scaling, really the entirety of it, was no longer needed. The technology had simply outpaced itself and the human in the loop.

I then spent my last few months there ramping that operation down and setting up a couple of other AI-related agents to help with things like quality technical writing. Once that was in place, my skills simply weren’t needed there.

It was not a super fun moment, and it was very, very bumpy. We had hundreds of people doing this work globally, so we had to figure out how to ramp down those contracts as gracefully as possible to allow these folks to have time to hopefully get into other roles.

On my team of about 10 people, only one person stayed on.

‘Info workers beware’

In my career, I’d gone from a place where companies like Waymo, Apple, and Amazon were coming to hire me to being out of a job and unable to get the attention of any company.

This moment is making it such that these great companies now have way more capability and people than they may need. So, you’ve got a lot of great people that are now having to find their next play, but the next plays are dramatically changing.

When I was hired at the startup, I spent the next four months with my team working tirelessly on basically solving this case and figuring out the right staffing and management model. Then GPT-4 came out, and when I gave it the case, it was an epiphanal moment. It spit out the exact answer — the perfect answer — in 30 seconds, including what we thought were very clever adaptations that had taken us a week to identify.

It not only gave us the answer but also the methods. I just sat there with my jaw in my hand. That was the moment that I thought to myself, “Info workers beware.”

My own pivot

After we began to wind down the operation, I started looking around to figure out what my next play would be. I spent about five months conducting my job search the wrong way and getting nowhere.

I remember this moment sitting there, again, with my jaw in my hand, wondering, “What am I doing wrong?”

It came out of that moment of almost desperation, saying, “I’ve asked everybody else. AI, what do you got?” It came back with more nuance and appreciation than I ever could have imagined. That’s when I moved into collaboration mode with AI.

One example was with Google’s NotebookLM. When it came out, people had fun with the idea of putting their résumés into it and creating a podcast. I actually found incredible utility in doing that. It’s interesting to drop your résumé and your LinkedIn profile into NotebookLM and see what AI makes of your career. What does it call out as the highlights?

When I did this, I realized that there were great things about my background and experiences that I wasn’t telling people because I didn’t see or appreciate them, but this podcast called them out.

Before using AI, I wrote a nice cover letter, updated my résumé, and started looking around on LinkedIn and applying. I was using my network, casting the line. I wasn’t just in a corner quietly hoping something would come to me, but it was the traditional approach of, “Here’s the résumé that I made for every job. Here’s a cover letter with a few tweaks.” I was essentially cold applying and trying to hit people up on LinkedIn.

Most people have probably heard that you should tailor your résumé, cover letter, or communications for a role. But that’s hard when you’re in the grind and just trying to get a job. You’ve already applied to a bunch, and you’re tired and don’t want to stare at the same words again and again. This is where AI is extremely helpful.

I also created what I called JobHunt GPT. Now I’ve turned it into CareerBuddy GPT, but JobHunt GPT was what came out of all this exploration. In my case, it was a custom GPT that understood my background, where I was trying to go, and the history of the jobs I’d applied for. So, I was able to go to it and say, “Hey, here’s a new job. Can you assess my candidacy for this?”

The first thing I get is an objective review of how I mesh up against a role. Then, I can say, “Alright, pick apart my résumé. What do I need to adjust?” It can generate the updated résumé, focusing on the things that are important for the role. And it can write the cover letter and identify the key people for me to reach out to. It’s essentially like lead analysis and lead development.

My advice to anybody else would be, don’t wait five months to figure out the right way to do it. The world has changed. This applies to anything, but especially if you’re looking for a job, you have to leverage the most powerful tool available, which is AI.

Do you have a story to share about your job hunt? Contact this reporter at [email protected].