Welcome to AI/ML Demystified Mini-Course!
Welcome to the AI Demystified mini-course. Over the next 5 days, we’ll help you through the buzzword-bingo that is Artificial Intelligence and Machine Learning.
I’ll be your guide, Gant Laborde. As a professional software developer for 20 years, I’ve seen code that ranges from near perfection to outrageous. I’ve seen people break through barriers and invent novel concepts, but nothing has compared to the breakthroughs and wonderment of recent advancements in AI. With all the exciting new content in AI, it’s easy to feel a bit flooded. Some content is real, and some isn't. Some applications are tangible, and some are shrouded in impenetrable buzzwords. For innovation to go wide, we need an approachable base of knowledge.
Fortunately, Infinite Red is experienced in explaining advanced concepts, and that’s what inspired us to make this free mini-course. As a consulting firm, our daily vision is to explain advanced concepts in everyday language. Our devs aren’t Ph.D. students or professors. We teach teams, speak at conferences, and write blogs. Our approach to explaining AI is by example.
As this course progresses, you’ll learn key terms and concepts about AI and get lots of concrete examples to reference. AI is a pretty far-reaching concept, so this course will dive into the most useful terms being used today, and that means focusing on a lot of Machine Learning.
With no further ado, let’s jump into day 1!
What is Artificial Intelligence?
Artificial Intelligence is simply the term that caught on for making computers perform intelligent and sometimes human-like operations that emulate living thought. This can be as simple as having a lookup chart for playing tic-tac-toe, or actively computing risks and rewards to beat world-masters in Chess and Go.
AI is a pretty far-reaching umbrella term.
ANI and AGI
AI has two major subcategories: AGI, which is Artificial General Intelligence, and ANI, which is Artificial Narrow Intelligence. Most of the cool new things people are inventing today fall under the category of ANI. This includes identifying people and things in photos, self-driving cars, and generating new photos of people or things that have never existed! While the concept of training a machine to recognize things it’s never seen before is impressive, it’s not going to start knowing what to do with those recognitions.
A good analogy is that ANI generally lacks context. Take a look at this fun Halloween decoration. While we can appreciate the fun in this display, ANI would likely identify the parts and completely miss the entertaining context.
No self-driving car is going to form a union or form opinions. Those broad concepts are left to AGI, which just happens to be the premise for just about every Sci-Fi technological horror. AGI is the concept of solving intelligence as a whole. While breakthroughs of ANI are changing the world, they do very little to progress us towards AGI.
ANI is where Machine Learning (ML) resides. It is artificial narrow intelligence. Machine Learning helps us use the enormous amounts of data at our disposal in the modern world. It also encompasses faster and faster machines to process that data as a teacher for algorithms. Machine Learning has its own set of subcategories, with each relying on data differently. Generally speaking, however, Machine Learning lets us leverage information to train and refine algorithms, rather than people, to do certain kinds of work.
If a person were challenged to write an algorithm that fixes all grammar and punctuation mistakes, they’d have to first master the English language and then find some way to encode that knowledge. Then, after years of work, it’s demanded that they also have to support Spanish punctuation and grammar. I’m sure that person would quit on the spot! But if we trained an algorithm to do the same thing, we have mountains of books upon mountains of articles to draw from. It's much faster and easier to have a machine chew through all that information and make the same tool in all languages.
That is the power of Machine Learning AI. Computers have always been impressive at handling tedious tasks that no person could surmount. In the age of data, using computers to teach, predict, and simplify algorithms is augmenting products beyond anything we've ever been before. We can use AI to detect spam, flag fraud, recognize handwriting or speech, and even secure our data using our face.
That’s why it’s important to understand and identify these advancements in the world around us. Once we see it, we can come up with applications for it. We’re all the creators now -- and what we create is up to us!
Click "Complete and continue" at the top to take the quiz