Kiddos

AI is not something that gives free will to machines

• Artificial intelligence is fundamentally a mathematical operation! If you feed it the same data, it will give you the same solution every time.
• Artificial intelligence is also not a magic solution to any problem. For example, artificial intelligence can’t grade your essay (yet).

• Artificial intelligence is a way to train a computer to do certain things without supervision
• In normal programming, we use rules to give us answers
• In artificial intelligence, we use answers to give us rules!
• This is like when your parents give you advice. You remember each thing they tell you.

• After getting lots of advice, we make the rules based on these answers to problems
• Someone that doesn’t know the rules guesses their own rules, and these result in answers in life situations that can be good or bad
• Would you prefer rules based around answers you know will be good or rules based on answers you don’t know yet?
• This is why we need artificial intelligence. It’s like advice to computers!

• The computer looks at something called training data. This data can be faces, song snippets, handwritten alphabetical/ numerical characters, or videos of behaviors.
• Each piece of data has a label to it. Every face has a name, every lyric has a song, every visual behavior has a word for it, every handwritten letter has a correct recognition
• After thousands of pieces of data like this, the computer can use math to understand what parts of each data piece is important and what it indicates

For example... let's say we're trying to recognize a handwritten number.

• The number 9 has a circle and a vertical line in it.
• Lots of numbers have circles and lines, like 0, 8, 1, 4, and 6
• After seeing thousands of 9’s and doing lots of “guess and check” repetitively really fast according to a formula with all the numbers, the computer eventually learns that a 9 has a vertical line and a circle touching the line in the upper part on the left side!

Wow! Now computers can read handwritten numbers!

• We can take this idea and apply it to video and sound too! Every lyric, every melody, every pixel has something to say to the computer!
• We can teach a computer to recognize anything if we have enough training data for it!
• It all comes down to learning the 'rules' of how to recognize the things we're interested in.

• That's a question you'll get to answer in college. It's a little too advanced for you!
• But if you think you can take a crack at understanding it, start learning a subject of math called linear algebra