One of my core memories is about, of all things, values. I was standing at a crowded all-hands meeting, listening to a senior leader talk to us about doing work that matters. They mentioned a line that’s stuck with me all these years later. It’s a mantra I keep returning to: you can only know what your values really are when they are hard to uphold.
My daughter started a “Save Planet Earth” club at her elementary school a couple of years ago. They’d meet up during recess and talk about all the ways to raise awareness, fight pollution, and defeat global warming. The real education started when we talked about what she was willing to do. What she was prepared to give up on. Less take-out food, less toys, less … stuff. Without introducing some pain, some tradeoff, she would never learn what her actual priorities are.
It’s the same thing in job interviews. Your stories need to cost you something. If your answers are perfect STAR formatted stories that skip on your internal struggles, you’re not informing the interviewer of anything important. With the big AI labs (OpenAI and Anthropic) this will lead to a No Hire.
I’ve started coaching a few clients again, now that the 8:32 Club is opening up to new members, so this is straight from the trenches.
The problem
Anthropic and OpenAI have different views on what AI Safety means for them. They will not extend an offer if your answer of what AI Safety means for you doesn’t align with their view. This is a detail that most candidates who seriously prepare for their interviews know about. Then candidates do exactly the wrong thing with it.
A source familiar told Axios that Anthropic CEO Dario Amodei has expressed concern about new talent coming to the firm for the money rather than the mission.
Most candidates read company pages describing charters, principles, or purpose. The smart ones ask AI to help them contrast and refine their understanding of the company’s culture and worldview. From this, they write their answer to the question “Why OpenAI?” or “What gets you excited about Artificial General Intelligence (AGI)?”.
Through thorough research and hard work you’ll craft an answer that is genuinely generic. A boring answer. The same answer that all the candidates waiting in line behind and ahead of you rehearsed.
Worse, it will be as inauthentic a piece of garbage as the values printed on the walls of most soulless companies. The problem is that this is 100% about the company and 0% about you. It’s worth reminding yourself that the company is interviewing you, specifically.
The question
The path to a better answer starts, as it usually does, with a question.
Start with your well researched and generic answer. Then ask yourself, “what about this answer can only come from me?” Look back at your experience and find trade-offs you, personally, had to make between speed and safety. Or think about what AGI breakthrough will make your life, or the life of people you care deeply about, richer? Maybe there’s a migration project where you risked your reputation to advocate for delaying the cut-over? Or maybe a family member has a rare disease that can’t get enough funding for research, but AGI can break the cost calculus for the better? It doesn’t need to be a life-or-death situation; it can be the long-tail of physical and mental ailments that afflicts our species.
You have two goals in crafting your answer. The first is to align with the company’s values. The second is to tell a story that no one else in the world could tell, a story that is uniquely you. Do both, and you’re past the top of the bell curve, on the side of the top candidates.
“
tell a story that no one else in the world could tell
Further Reading