Writing

Our first recruitment marketing coordinator hire was a month out of college with a marketing degree and zero idea what recruitment marketing was. We had to teach her the basics before she could do anything useful. Within a few months she was pulling reports, interpreting our weekly dashboards, and asking me every other day if she was supposed to pause a campaign because a recruiter changed the req again. It was a lot of manual button-pushing and processing data. The kind of work an AI model handles in seconds now.

Last week was the biggest vertical week of my Zugspitze build, getting in just over 14,000 feet of climbing. This also means I had some extra time on trail to catch up on podcasts. One episode stuck with me: David Epstein on the Science of Running, talking about how the best coaches think less about piling on work and more about finding the single thing holding an athlete back.

Last week on a client call I heard framing from a client that felt like a bit of a leap forward. This company was already investing in optimizing their content for AI search engines, but their reasoning was what felt unique. This org is hoping to attract candidates with some baseline level of AI fluency. They figure candidates already using AI tools to job hunt are probably the same candidates who’ll use AI tools effectively inside the role itself. So if their jobs and content surface well in ChatGPT, Claude, etc., they’re putting their openings in front of exactly the type of people they want. A self-selecting filter - or a more modern way of advertising on ‘niche’ job boards.

I opened LinkedIn last Tuesday morning. The first two posts I saw were these.

The Quieter Conversation About AI Jobs

Yesterday was the start of transitioning from my Boston training season to Zugspitz ultra training. It’s a bit of weird timing, since the Boston Marathon is next week, but the hay is in the barn, as they say. I took advantage of a few drier days up in the PNW to sneak out on the trails with the dogs for an easy rip up and down the Chanterelle trail.

Subscribe now

I’m back into my Boston training block right now.

Subscribe now

Subscribe now

I presented at CultureCon this week. The talk was titled “AI Won’t Fix Your Culture Problems. It Will Expose Them.”

Back in 2019 at Unleash, Edward Snowden was the keynote speaker. I remember sitting in that session a bit surprised at the choice of guest. And I remember being fascinated by his perspective on privacy and how the rise of social channels was quickly killing our ability to think independently.

I read an open letter this week from a talent acquisition industry veteran to companies like Mercor and Scale AI who are building AI hiring tools. The message was clear: call us first. Consult TA experts. Let the people who’ve been recruiting for decades teach your models what good hiring looks like.

A few weeks ago, Deloitte agreed to refund money to the Australian government after delivering a $440,000 report that included fake academic citations, fabricated court quotes, and nonexistent references. The culprit? AI hallucinations that nobody caught before the report went out the door.

If you type “what are the best AI interview tools for recruiting” into ChatGPT, you’ll turn up a ‘consultant-like’ output that points you on your way.

I keep seeing the same headline: “AI agents will make recruiters more efficient.”

Many books have left lasting impressions on me throughout my journey growing Change State . As an avid trail runner and outdoor adventurer, Let My People Go Surfing and Shoe Dog stand as two of my all-time favorites. Both are highly recommended for anyone curious about entrepreneurship—you rarely get direct insight into the founding journey of behemoth companies like Nike and Patagonia straight from the founders themselves.