I recently stumbled across a very interesting free course on Artificial Intelligence (AI), taught by Stanford University’s Professor of Computer Science, Phillip Isola. It’s called: “Spam Free AI: Getting past the hype and finding the real AI edge” [1]. It covers all the major topics – from NLP (Natural Language Processing) to Deep Learning, and more. The course is 1 hour 31 minutes in length, and you can watch it on YT if you’d like, here:
I’ve been working with AI for 15 years now – and have been doing so professionally since 2009, when I co-founded Conversic.com [2], an AI service that improves email marketing ROI; we launched in 2010 as ‘Conversica’.
So I was very curious to see what sort of misconceptions people had about it. And this course does a great job of dispelling some myths about AI… As well as giving you some solid and practical advice on how best to use these technologies within your business right now.
One thing that struck me was something Phillip said at the start of his course; he clearly thought they were all overhyped – meaning they’re not quite as impressive as everyone thinks they are… So let me show you what he said when introducing this topic… And then help clear up that misconception for him! 🙂
NLP uses a network that progresses through something called training.
NLP uses a network that progresses through something called training. It can be used to find patterns and identify topics, helping machines understand the meaning behind words, so they can use them appropriately.
NLP is also referred to as computational linguistics (CL) or natural language processing (NLP).
NLP can be used to find patterns and identify topics.
NLP can be used to find patterns and identify topics. It can help machines understand the meaning behind words, so they can use them appropriately.
NLP is used in many different areas of business, such as customer service representatives who need to understand what customers are saying and how they’re feeling when they call in with problems or questions. In addition, machine learning models are being developed that will allow machines to analyze vast amounts of data by looking at relationships between things like words and images—a process known as deep learning (which we’ll discuss later).
It can help machines understand the meaning behind words, so they can use them appropriately.
NLP can help machines understand the meaning behind words. It’s not just about recognizing individual letters, but also about understanding what those letters mean.
For example, if you were asked to write down a list of animals that start with the letter P and then try to find a word that sounds like it in English (such as “parrot”), you might come up with “pigeon.” But if you were given an image of a pigeon and asked which animal starts with P? Then maybe your answer would be “parakeet,” because those two animals look similar—they both have feathers! This example shows how NLP can help machines use words appropriately when they don’t know them yet (like finding out which animal starts with “p”).
That’s usually done by using a corpus of text that is broken down into sentences and rules.
The corpus is a collection of text that has been broken down into sentences and rules. The sentence level, or lexical level, is then fed into a machine learning algorithm that determines how much each word contributes to the meaning of the sentence.
The next step is for you to train your own system on this new corpus using your own data: if you have some text in mind but don’t know how it works yet, look at other corpora from other languages or cultures (e.g., English texts).
Those sentences are then fed into a machine learning algorithm.
Machine learning is a type of AI that uses data to make predictions. It can be used to classify objects, understand language and even predict the future.
Those sentences are then fed into a machine learning algorithm, which trains an algorithm on what it has learned so far. The trained algorithm will then be able to tell you whether or not you should buy this new item based on how much work it would take for me to make (or if I already have).
The machine learns by identifying relationships among the words and determining which parts of speech apply to each sentence.
The machine learns by identifying relationships among the words and determining which parts of speech apply to each sentence.
For example, if a sentence contains a verb (“The dog bites”), then it must also have an object (the dog) and an action (biting). The machine applies those rules to make new sentences that also fall under the rules.
Then, it applies those rules to make new sentences that also fall under the rules.
Then, it applies those rules to make new sentences that also fall under the rules.
For example, if you have a rule that says “a noun is always followed by an adjective,” then any sentence where “this” and “that” are used together will pass this test. The machine learns by identifying relationships among the words and determining which parts of speech apply to each sentence. Then, with some help from its dictionary (which contains all possible combinations of words), it can use that list as an aid when generating new texts based on conditions previously identified in training sessions with human editors who helped teach the system how each phrase relates to others within your vocabulary list