Stochastic

For most intents and purposes, this is synonymous with “random,” though it’s a rabbit hole of a word – you can go very deep on what it means to various disciplines and in various combinations. The meaning can be very subtle or technical depending on what discipline or practice it’s applied.

Most commonly, you hear it in the context of a “stochastic process,” which is to say some sequence of events that involves a random variable, making the outcome of the process itself random.

You would rarely hear of the word used to describe a single thing – like, “a stochastic variable.” More often, it’s mean to describe a series of interrelated things, the outcome of which can’t be predicted. A random variable will produce a stochastic process.

The “ch” is hard – “STOW-kass-tick.”

Why I Looked It Up

I just kept seeing the word. My first few attempts to figure it out where stymied by the “rabbit hole” I mentioned above – I kept going down into specific disciplines – but then I realized I was just over-thinking it.

I finally took another look when The New Yorker published a visual essay entitled Is My Toddler a Stochastic Parrot? which refers to a concept in AI:

In machine learning, a stochastic parrot is a large language model that is good at generating convincing language, but does not actually understand the meaning of the language it is processing.

So, an AI just randomly stitches words together based on what it’s heard, until it finds something that makes sense, then learns from that. It doesn’t know what it’s saying, but just has some fitness function at the end that tells it whether or not the word combination means something. (Much like a toddler, I suppose – if they randomly make a noise that results in them getting a bite of Mac and Cheese, they will learn to do that again.)