Ask Me Anything

I'm not going to be one of these people who will blindly respond to you by saying you can definitely get a job at a more elite company if you just keep trying. I couldn't possibly know that for sure without knowing your actual skills.

That said, here are a few thoughts:

• When I worked at OpenAI, and also when I worked at early Facebook (when it was only 500 engineers), the number of false negatives in interviews, where great candidates were rejected, was very high. This makes sense when a company is very desirable to work at, since it favors them to bias toward false negatives (which are so much better for them than false positives). This means a rejection doesn't necessarily mean you can't do it. It also means you should be rejected from quite a few elite places, the more the market skews toward going elite.

• Over the past few years, the market has skewed heavily toward the most elite workplaces, both because compensation has skewed that way and because there's not the same sense of plenty as there was in the dot-com days. In the dot-com days, you could work for elite firms, but there were also a boatload of startups, and many of the startups were compensating quite well. In fact, in the dot-com days, it was considered most attractive to work for startups because people felt like big companies just didn't get it. This increased competition squeezes far more people in through a tight little door.

• There is, of course, also the question of whether you indeed are the right material for the ultimate job you want. I have no idea whether this is the case. As a frame of reference, I've worked on three different teams at OpenAI, and on each one of those teams I was the slowest engineer. This, in my 28 years of work experience, had never happened before, not even close. On most teams in other companies, I was considered a pretty strong developer. Are you a stronger engineer than I am? I have no idea whether that's the case. Can you get objective feedback about whether you are? If you're lucky, once in a while an interviewer will go off script and be willing to give you true feedback about your interview performance after the fact. Sometimes you'll even meet a recruiter willing to do that. Barring those opportunities, I'd encourage finding parallel routes to getting that all-important feedback, because my advice would bifurcate for you depending on the answer. If you truly are elite, my answer will be: take more shots at the goal and keep going. There are no shortcuts to that. If instead you're tilting at windmills, my advice would be to seek the best job you can obtain right now and make the most of it.

Wishing you all the best. It's a very tough market out there, and I completely empathize with the difficulty of getting some jobs. I was flat out rejected the first time I applied to OpenAI.

The fact that you're already working in ML is a huge plus to not being outmoded by the rapid development of AI worldwide. In fact, one would argue that if Jevons Paradox does turn out to be the case in our industry, people like you will become even more valuable as AI adoption throughout all sorts of industries requires more and more people who actually know what they're doing.

That said, I completely hear you on the FOMO. In 2004, when Google opened its first office in Seattle, one of my skip-level reports went over there. He was very nice to work with, but he was an average developer, not exceptional. He ended up doing much better financially than I did, even though I was several levels above him and had a lot more scope for growth in my future. It can be hard to watch this sort of thing go on if we focus on it too much.

Here are a few different thoughts along different axes:

• There's a question of intrinsic reward. Annie Dillard was famous for saying, "How we spend our days is, in the end, how we spend our lives." There's a strong argument for optimizing for the substance of your days themselves being very enjoyable. It sounds to me like your current role and job are very enjoyable to you. This, of course, doesn't mean you couldn't have the same level of enjoyment at a frontier lab. All I mean to say is to encourage properly valuing the situation you're already in (which you obviously do, given what you've said).

• There's a sense in which, if our eyes rove the land far enough, we will always find reasons for FOMO. There will always be someone who is less deserving than you, someone who has it better than you, who makes more money than you, who looks better than you, who is younger than you, or who is smarter than you. It's easy for me to say this, but if I'm honest, I've had a very hard time in my life grappling with this, given that I've been an overachiever most of my life. The wisdom is in understanding that if our happiness depends on those seeming inequities not existing, then we shall resign ourselves to be unhappy forever.

• A big part of the decision about whether to go to a frontier lab also probably has to do with your personality type according to the Big Five. I, for instance, rate very high on "openness to new experience." A personality like mine should switch and try new things often. There are many people who rank low on openness to new experience, preferring stability and predictability. Those sorts of people should stay in the already-good role you're in. There is intrinsic value simply in switching what we do: for the experience of the switch itself, even if some switches end up disastrous. But the amount of that value depends a lot on one's personality.

My intuition is that you will do very well regardless of your choice here, because the way you describe your situation does not take your existing circumstances for granted, seeing only green grass on the other side. Your personality is likely to find things to be grateful for regardless of circumstance, and that serves you well.