March 2026
Two weeks into March Madness and I'm already rethinking our hiring plans.
If you missed the last two weeks, I shut down normal operations at both companies and forced everyone to spend March learning and using Claude. The results have been staggering. At Pensford, we trained a chatbot on our core products, turned years of white papers into structured training programs, and connected our systems so that even I can pull insights from our database without bugging the team. At LoanBoss, the engineering team automated code reviews, deployment checklists, and ticket writing. A doc organization project that used to take a full day now takes five seconds.
I expected improvement. I did not expect this.
And it's forced me to confront an uncomfortable question: how does this impact hiring plans?
The bar for me to hire a new college grad right now is extremely high. Any hiring we do in the near term is most likely to reallocate existing people toward AI implementation and then backfill their old roles.
We are in the early stages of this experiment, but the potential is high enough that it would force me to hold off on hiring an analyst class from this year’s graduates.
I think we’re seeing this dynamic play out in the labor data each month. NFP has fallen off a cliff but the unemployment rate hasn’t budged. Companies aren’t firing, but they aren’t hiring, either.
A major study out of Stanford (Brynjolfsson, Chandar & Chen) used ADP payroll data covering millions of workers and found that employment for workers aged 22-25 in the most AI-exposed occupations dropped 16% relative to older peers between late 2022 and September 2025. Not because companies were firing them - because companies stopped hiring them. The adjustment mechanism isn't layoffs. It's a hiring freeze that nobody announced.
Entry-level job postings have fallen more than 35% since January 2023. Goldman Sachs flagged that unemployment among 20-to-30-year-olds in tech-exposed fields has climbed 3% in the last year alone.
Here's what makes this so tricky. The macro labor market looks fine. Yale's Budget Lab concluded that broad occupational shifts since ChatGPT launched are basically in linewith prior technology transitions. Over 80% of firms surveyed say AI has had no measurable impact on employment or productivity. The average executive uses AI a whopping 1.5 hours per week.
So at the macro level, everything looks normal. But zoom in on early-career workers in AI-exposed roles - finance, tech, professional services - and the picture is very different.
For the unlucky few that have lost their jobs, AI might just be a convenient excuse. A Harvard Business Review survey found that 39% of companies made low-to-moderate cuts in anticipation of AI's impact, not because AI had already delivered. Only 2% of large headcount reductions were attributed to actual AI implementation. Companies are cutting jobs based on a bet, not a result.
But there’s a perverse incentive to cite AI during layoffs, which might really just be about masking underlying weakness. It’s a double whammy - reduce expenses and megaphone AI…markets love you! Stonks to the moon!
Jack Dorsey (Block and Twitter founder) is the poster child. He cut 4k employees (nearly 40% of Block's workforce) and explicitly tied it to AI. But I don’t believe him. An Oxford Economics report found many layoffs that CEOs called AI-related were actually the result of past overhiring. Block had massively overhired during COVID and Dorsey used AI as the excuse.
But if you didn’t overhire, then the potential for the current team just exploded. They know the company, the industry, the workflows, etc. A lot of time and money has already been invested in them, why let them go now? AI is like giving a professional athlete steroids - it makes them better at something they were already great at.
There is a nationwide hiring freeze while companies figure out how much their existing team can achieve with AI instead of additional headcount. Just how big can Barry Bonds’ head get if we keep injecting him with AI?
