‘By far the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.’
Eliezer Yudkosky, Machine Intelligence Research Institute
Compared to the evolution of human intelligence, machine intelligence and AI have developed at breakneck speed. The arrival of machine learning, deep learning, and the neural networks that make them possible has rendered the technology – and its applications – exponentially more complex. For those that aren’t tech-minded, it can feel as if new hurdles to understanding AI are introduced on a regular basis.
All hope is not lost. A basic education in the technology and concepts driving AI’s development can provide newcomers (and seasoned experts) the tools and vocabulary they need to break down the most knotty ideas into their foundational elements. What’s more, that knowledge can now be had on-demand, and often free of charge; the great democratising power of online education means that anyone can now get a grounding in artificial intelligence.
Here are six online courses to teach yourself about AI and navigate those hurdles.
The basics, for free
Course length: 9 hours
This brand-new, free introductory course has everything the AI novice needs to get to grips with the subject’s big questions. Aside from a general survey of the subject, it covers:
- Common AI jargon
- AI’s capabilities
- Possible applications of the technology
- How to talk to AI developers, and what questions to ask
It’s a great option for an all-in-one foundational course, and it’s led by AI giant Andrew Ng, founder of Google Brain and former Chief Scientist at Baidu. There aren’t many more qualified professors out there.
Course length: 4 months
For a more in-depth beginner’s course, Udacity’s ‘Intro to Artificial Intelligence’ is another great (and free) option. The program covers everything from the fundamentals to real-world applications, including:
- The maths behind artificial intelligence
- An introduction to machine learning
- AI for computer vision
- Natural language processing
You should be comfortable with linear algebra (here’s a free course for that) and probability theory (and a course for that, too), but otherwise, it’s a great choice for beginners who want a more technical introduction.
Course length: 3 months, 8 hours per week.
This course takes the AI education experience a step further by introducing a thorough practical element. Students are taught the fundamentals of AI before applying the knowledge to create their own algorithms. It includes:
- The history of AI
- Real-world AI applications
- An introduction to machine learning algorithms
- Building ‘intelligent agents’
- Problem-solving using Python-powered AI
This one is free, but has an optional paid certification should students want to put this on their CV. A background in Python is recommended (and is extremely useful to have if you’re planning a career in AI), which you can get here.
Course length: 56 hours
If you wanted to learn with Andrew Ng but found the ‘AI for everyone’ course too basic (or if you’ve done that one and want a new challenge), this is the course for you. It’s another free program with an optional paid certification, and features practical algorithm-creating assignments. Amongst many topics, the course covers:
- Linear algebra
- Machine learning algorithms
- Artificial neural networks
- MatLab tutorials
- Computer vision
It’s a major step up from the basic courses, and a great way to continue your AI education. Beware, though – it’s not for the faint of heart. It’ll likely take longer than the advertised 56 hours, and requires serious commitment and a technical background to complete. If you’re looking for the biggest, most renowned challenge in online AI education, this is it.
Course length: 8 hours
If you’re already familiar with the AI basics and are looking to specialise, courses like Nvidia’s computer vision program are worth considering. It’s another practical one, in which students are taught:
- Common deep learning algorithms
- How to implement algorithms for image recognition
- How to refine neural networks and apply them to real-world use cases
It’s not free, but it’s a course presented by one of the leaders in AI for computer vision, and is well-reviewed. Just make sure you’re confident enough in your AI knowledge before signing up.
Course length: 18 hours
Here’s another course focusing in on a popular subset of AI development – self-driving cars. If you’re interested in specialising in the future of the automobile industry, you’ll learn:
- How to apply computer vision to self-driving scenarios
- How to create a deep learning algorithm that can distinguish between hazards, road signs and other cars
- How to use the Keras open-source platform
It’s another paid course, but can often be had for a heavy discount (often over 90 percent) during one of Udacity’s frequent sales.
Hit the books
The aforementioned courses are the tip of the iceberg when it comes to AI education, but each offers a significant grounding in the subject. It’s possible to work your way up from the basics to real proficiency using online resources, often without spending a penny. Once broken down to its fundamentals, AI isn’t too complicated a subject. It does, however, require some serious commitment. There’s very little hand-holding in online education, so be ready to self-motivate and do plenty of practical coding work outside of class. You’ll be well-rewarded for your perseverance.