I applied to the job using Google's career website and sent my resume. I received a call from the recruiter shortly after my application. The recruiter's communication was good and prompt.

We scheduled a phone interview, and it took them 3-4 days to get back to me with the results. Unfortunately, I didn't pass the phone screen.

I was first asked a programming question: Design and implement a class that returns a random string value from a set with an arbitrary probability distribution given by an array of probabilities. Using an existing random number generator with a uniform distribution (e.g., Random.nextFloat()), return the string for the random float value based on the strings' probability distribution. This sounded hard due to the wording, but it was easy to implement.

After I completed the coding question, we had plenty of time, and I was asked a bunch of floating-point number-related questions:

1. Given an array of floats, you have a sum function that takes the float array as input, but it returns a different result every time you change the order of the float numbers in the array. Why?
2. How are floating-point numbers represented in binary?
3. How do you add two floating-point numbers in binary?

The answers to these questions lie in the IEEE Standard for Floating-Point Arithmetic (IEEE 754), and you can find sources online.

I got really frustrated with this series of questions because they had nothing to do with the coding interview. After struggling to find answers, I finally ran out of time. I was so relieved, though, that the interview was over because it really felt like torture in the end, with the interviewer beating a dead horse.