Hire people who give a shit. - by Alexandr Wang
Startups die when they become credentials instead of cults—here's how to screen for people who actually give a shit, not just resume builders chasing your brand.
Read Original Summary used for search
TLDR
• Screen for two things: caring about your specific mission AND having a history of obsessive work (if they've never been obsessed before, your company won't be the first)
• The "credential vs. cult" trap: as you scale, more people interview for the brand rather than substance—ROI of hires drops even as top-of-funnel increases
• Specific interview questions to detect obsession: hardest you've worked, hours per week, why you cared, most unmotivated period, proudest achievement, was it worth it (obsessed people always say yes)
• Big company 11am-4pm culture proves most people don't give a shit—you can't do meaningful work in 5 hours/day
• Culture IS the company, not the perks—flexible hours and office design tell you nothing about whether people care
In Detail
Scale AI's CEO argues that the primary hiring criterion should be "giving a shit"—specifically, caring about the company's mission AND having a proven history of obsessive dedication to work. The core danger as startups scale is becoming a "credential rather than a cult": people join for the brand signal (like Harvard) rather than the substance, creating a university-like churn of smart but uninvolved people who never dive deep enough to do meaningful work. This happens because growing companies miss that while top-of-funnel increases, the ROI of hires drops dramatically as more candidates interview for resume building rather than mission.
To combat this, the author uses specific interview questions to detect genuine obsession: What's the hardest you've ever worked? How many hours? Why did you care? When were you most unmotivated? What are you most proud of? Was it worth it? For truly obsessed people, it's always worth it. The proof is in past behavior—if someone has never been deeply obsessed about something before, betting they'll start with your company is a bad bet. He contrasts this with the "recruiting as courtship" approach (pursuing people for months, finding a spark) versus the "college admissions office" approach (sifting through homogeneous candidates), which produces only higher-credentialed homogeneous soup.
The broader implication is that culture equals the company itself, not cosmetics like perks or office design. The uncomfortable truth is most people don't give a shit—evidenced by big company engineers working 11am-4pm (you can't do meaningful work in 5 hours). Unless you actively fight against becoming a brand that attracts credential-seekers, you'll end up with a company full of people treating it as a 2-3 year resume line rather than a generational mission.