Made with ❤️ by Liam Hinzman
Why You Should Do a Focus in AI 🤖
Right now artificial intelligence is being used to detect diseases, enable cars to drive themselves, and even beat humans at video games.
AI is a exponential technology with tons of applications, and to build projects that can solve real world problems all you need is a laptop!
Understand how AI Works (25 hours)
Don't worry about understanding all the math 100%, this can come later. But you should have the basics of Python.
Do this Udacity course, it teaches you all the basics, and will give you experience with building some AI projects, like turning your cat into a painting 🐱 → 🎨
Or you can check out the courses on Made with ML (see "machine learning basics course").
Watch this Artificial Intelligence Full Course.
Learn an AI framework (8 hours)
I recommend PyTorch over TensorFlow. Each framework has very good tutorials.
The general consensus is that PyTorch is easier to use than TensorFlow.
Do the PyTorch blitz (2 hours), then do one of their other official tutorials (I recommend the GANs tutorial which will take 6 hours).
Replicate #2 (6 hours)
Building is the best way to learn. Find something interesting that AI can do and build it. If you replicate someone else's project, give credit!
Examples of good AI replicates:
Liam: Creating Fake People and Cats with GANs
Mayank: Writing Fairy Tales with RNNs
Aadil: Beating Video Games with Reinforcement Learning
Review your Linear Algebra (4-15 hours)
Linear algebra isn't needed to build AI projects, but you'll need it to understand the theory behind what's going on in your projects and research papers.
Check out this Udemy course → Machine Learning A-Z™: Hands-On Python & R In Data Science (it's not free, but it has good content).
Start Reading Research Papers (25+ hours)
Once you can read ML research papers and understand 80% of it in 2 hours, you’re legit 👌
Pick what interests you. Read at least 7, and replicate 1.
This YouTube channel highlights awesome papers
Create! (20 Hours)
Use what you've learned to solve a real world problem with AI!
Examples of good AI creates
Liam: Zilic: Detecting Any Disease in a Medical Image
Ananya: Pikachai: Your Best Friend is Now a Talking Pokemon
Shagun: Darth VAEdar: Finding Genes that Contribute to Cancer