I haven’t started working at Palantir yet, but I did a lot of interviewing while I was at Amazon and based on the questions I was asked, I can tell you what NOT to do. Don’t try to memorize “common interview questions”. Study questions asked by companies like Microsoft, Google, and Amazon, but make sure you understand them, not simply memorize solutions. Learn the tricks to do common things like sorting, randomizing, searching, and calculating, but learn how to apply those tricks in different situations, because you never know exactly what question they are going to ask.

A good interviewer isn’t looking only for whether or not you get to the right answer, but how you get there. If you rattle off the answer as though you had memorized it, that doesn’t tell them much. Think out loud, talk about your strategy for designing a solution, talk about assumptions you are making and many alternative solutions. Ask questions! “Is memory size a problem? Is wall clock execution time more important, or is it more important to optimize memory usage? Will the usecase benefit from caching or a lookup table?”

Similarly, knowing design patterns is important, but understanding when to use them and being able to describe why they are good patterns is key.

Since understanding is the key, nothing I can tell you can really help, besides to know your stuff, and practice communicating out loud and thinking through a problem. I might write a blog entry about this soon, as many people have asked me this before.

-Carl

Thanks,
Hellome

Very exciting news, Palantir is a difficult place to get in, and it’s a compliment to your skills of a programmer to get in :)

-Ben

This is the doctrine of falsifiability…there is some reasonable evidence that this is not really how science operates. Check out the Kuhn for particular historical examples of how this has not historically been the case in chemistry physics, etc.

> It’s not about “moral high ground”, it’s about usefulness to society. Scientific theories yield useful predictions about our world,
> that is why we learn them in school. Religion yields no useful predictions about the world.

The point of science is to make predictions about the world. Just because English doesn’t yield useful predictions about the world doesn’t mean it shouldn’t be taught. Science on the other hand does not provide standards of ethical behavior, guidelines for interpersonal behavior, discussions of the meaning of life and death.

I’m sure if you stop and reflect a second Carl, your realize that someone on high deciding what things are “useful for society” and building a curriculum based on that is really an insanely dangerous proposition. I’m fully on-board with the importance of science and a declining role of religion in public life. But I think that can only be accomplished by convincing other people…including people who go to church and yet also think modern medicine is pretty neat. The idea that if you want the benefits of modern society you must give up your religious convictions is not fair. Not only is it fundamentally against ideas of free speech and religion, but it also ignores the many obvious contributions of religious individuals to modern society.

As for my audience, I honestly didn’t think much about my audience, outside of Steve. I’m sure depending on who I was talking to, I would color the conversation in very different ways, because despite my very strong feelings about religion, I don’t wish to actually insult someone or argue “in their face”. I would hope that any friend of mine could read this and understand my beliefs better for it.

Since the majority of the US is religious, and scientists are a minority, evoking pure democracy in curriculum decisions is dangerous also. Those who benefit from technology and medical science, however, most certainly are a majority, so maybe that is a fair way to do it. “Vote with your health”, if you use technology and see a modern doctor, then accept mainstream science, otherwise move to “religionstan”, and live like a sheep herder, and die of pneumonia at age 28, and don’t ruin my children’s education.

It’s not about “moral high ground”, it’s about usefulness to society. Scientific theories yield useful predictions about our world, that is why we learn them in school. Religion yields no useful predictions about the world. In fact, many predictions it makes are hurtful (such as the prediction that the world will end – I guess all our efforts to recycle are useless since the world will end any day now). Religion was useful when it was the sole supporter of music and art, the primary inspiration for the creation of masterpiece after masterpiece. The motivator of charity and universal code of morality. Humans have come a long way since then. Now we have the constitution, the endowment for the arts, secular charity and tax breaks to encourage donation, and a multitude of cultures around the world are now accessible thanks to improved communication which make accessible nearly limitless sources of artistic inspiration. The crutch is no longer needed.

Two comments:

1. I’m not sure I’d quite characterize science as the impartial truth producing machine you seem to be saying it is. Science is awesome and all, but it is very much a human endeavor neither free from bias nor without negative social consequences. You might check out “The Structure of Scientific Revolutions” for a few examples. Not my any means to day religion is pure…obviously not…but I wouldn’t say science can really claim the moral high ground.

2. Your post seems to be addressed to atheists, yet I’m not really sure what to make of it. It seems clear to me that to approach a religious friend and say “your belief is tantamount to the flying spaghetti monster and the result of confirmation bias” is not a good way to begin a productive conversation.

I’m especially wary of your radical 2+2=5 mischaracterization up above. While it may annoy you that some people don’t want to teach something like evolution in schools (not really the same as 2+2=5), I think it makes more sense to think about this issue as one of who gets to control a curriculum. It’s a very dicey issue with no particularly easy answers. If you noticed your child’s school was teaching math in a way that took all the interest and magic out of it for example – is it appropriate for you to get your local math teacher on board and design a different approach? If somebody said that this was not *real* math like mathematicians used and proscribed by no-child-left-behind, could you decide that because parents mostly fund the school with their property taxes the decision of someone in washington was not important?