How do machine
learning and AI work?
1. Motivation

- Our daily life without computers, the newest technologies, and handy gadgets is nowadays hard to imagine. There seems to be more and more of them around us every single minute and you really need to put quite an effort to actually harness all the available "technological powers". Modern humanity is currently experiencing a particularly rapid technological progress.

2. Topic introduction

- The age of artificial intelligence is undoubtedly upon us this very moment. The majority of our population is utilizing this unique technology every day, even though not many can grasp what it is exactly. What do you think? What is artificial intelligence (AI)? (Students' thoughts) It is a vast field of computer science, making machines work similarly to human intellect. However, programming computer to drive a car and obey the traffic laws is not nearly enough to call it AI. It also needs to be aware of other signals on his way, such as human irritability.

- Nevertheless, artificial intelligence is undeniably going to feel a bit less artificial and much more intelligent very soon.

3. Knowledge update

- Growing popularity of AI in recent years is a direct result of several critical factors. First of all, a tremendous amount of data is generated every minute. Interestingly, 90% of it has been created in the last two years. Now, thanks to advances in processing speed, computers can actually "digest" all this information much faster. That is exactly why tech giants are able to continuously pour money in AI development and saturate the market with a countless number of new apps.

AI is not a robot. The robot is only a shell concealing precisely what the technology is feeding off. Where do you think we utilize AI? (Students' thoughts)

- We have introduced countless AI systems into many aspects of our lives: some of them can be found in cars; many are underlying the work of Google search engine and Facebook news feed. Modern anti-spam filters in our email and autopilot aircraft systems, even our smartphones and a number of games are also a direct result of AI developments. Complex systems of weak artificial intelligence (narrow AI) are used on exchanges, in manufacturing, finance and military fields.

4. Understanding/ comprehension of the knowledge

1) Types/categories of artificial intelligence.

- Weak artificial intelligence is the first level of artificial consciousness specialized in a narrow task of a particular field: for example, it can beat the world chess champion, but only that and nothing more.

- Artificial general intelligence (AGI) is an artificial intelligence of a second level reaching and exceeding the usual human consciousness: it can solve math and logical problems, think abstractly, compare and master complex ideas, learn fast- including from its own experience.

- Artificial superintelligence (so far hypothetical) is a third level of artificial intelligence technology development, on which it becomes smarter than all humanity combined, at first a little and then later, as a result of self-learning, trillion times more.

2) How do machines learn?

For instance, even though a human being can easily distinguish a cat from a dog, the task is not so straightforward for a computer.

It is clear, that if we only consider the appearance, the difference between a dog and a cat is not really so obvious. You can of course assume cats have pointy ears and dogs floppy. However, these rules are not always universal.

With respect to the length of tail, the texture of fur and its color variations we get many more options. Meaning someone will have to program countless tedious algorithms to help the computer figure out the difference.

But remember: machine learning gives computers the ability to learn the same way humans do, from their own experience.

During the process, they analyze thousands of examples to create an algorithm. Subsequently, the algorithm setting will vary, depending on whether the objectives have been achieved; so over time, the computers are indeed getting smarter.

That is exactly how machines, such as IBM Watson can diagnose cancer or compose classical symphonies. Some programs can even imitate a human brain structure, which together with neural networks helped us come to the conclusion that machines can indeed solve problems.
AutoDraw is a new kind of drawing tool that pairs the magic of machine learning with drawings from talented artists to help everyone create anything visual, fast.
Alternative activity
Achieving the AGI level of intelligence, equal or better than human brain now seems truly feasible in the near future. The robots are already putting many people out of work, performing their jobs faster and more efficient.

Let's define the parameters making AI much more efficient than us, humans. (Students' thoughts) Speed, memory, productivity, ability to collect and analyze data.

In terms of data processing capabilities, AI is an undisputed winner, it helps us accomplish more in less time by assuming repetitive tasks in our daily lives allowing us to develop better strategies and build successful relationships.

Did you like this lesson?
Thank you for using our lessons!
We will be grateful if you could spare a couple minutes to answer a few questions.

Your response is going to help us improve the project and report to our partners and investors where exactly our resources are being used. With your help, we can continue to create new lessons and offer them free of charge to schools.
School name
Respectfully, Alex Lisovskyi,
Founder of Futurio

Made on