Do I ever get worried that people, I mean, they're going to create something like Skynet in the Terminator movies? My short answer is I never get worried. Hi, my name is Kurt Gachidza and I'm a senior data scientist. I'm here answering some questions about natural language processing from around the internet. Let's get down to it.
What are the most interesting use cases of natural language processing? Natural language process is when we give computers the ability to understand natural language and natural language in this case is something that is similar or more akin to what humans use when they're communicating with each other. When you go on Instagram, if someone has written something in another language, you can press the translate button and then it translates to another language. So that's a common thing.
Yorkshire Wisdom says, why does predictive text seem so aggressively assertive? When you're trying to type something, it doesn't lie. It has kind of memorized your behavior with the way that you type certain things, right? It's going to assume that what you type frequently is what you're trying to type right now, even though you might be trying to type something else. So it just looks at the frequency of the type of sentences that you type and it tries to predict that based on your behavior.
Someone on Quora asks, what is the difference between natural language processing and speech recognition? Well, there's really no difference between natural language processing and speech recognition. Speech recognition is one of the subfields of natural language processing, just like natural language processing is a subfield of AI.
Someone also on Quora asked, do chatbots pass the Turing test? No. The Turing test is basically a test where you're supposed to be unable to distinguish between a human and a machine, but do not be deceived. You can go on Wikipedia and then you can take all of the stuff that is on Wikipedia. You can basically just iterate over the knowledge graph on Wikipedia of all the articles. All the text that is there, you can feed it into a machine learning system. It can memorize all that stuff, and then it can be able to respond to questions. However, they can spend maybe five minutes being on topic, but as time goes on, with an increase in the length of time, they can start blabbering, basically.
YouBarbecueK69 on Reddit asks, why should I pay for a language course if there is Google Translate? There's a difference. Google Translate does not help you master another language. It just helps you translate stuff from another language. And remember, Google Translate does not capture all the nuances of language, which you actually get in a language course. It just captures what the data that's fed into it, and those things, you can actually know them from a language course because you're being taught by a native speaker of that language, so they make you aware of that. Google Translate doesn't do that. I mean, you can still get very far, depending on the interest level, but it's going to be an impediment in the long run.
Bridget on Twitter asks, why does my child say please and thank you to Alexa, but not to you? Sorry, forgive me for laughing. I'm laughing because I'm a dad and I have an experience with that. It's nothing to do with you, Bridget. It's just the way things are when it comes to raising up children. And believe me, all the stuff that we have talked about, it's not even, just from a philosophical standpoint, it's not even the tip of the iceberg. There's so much more that goes into play with NLP, which is why I'm so fascinated by it.