At their first AI day, Salesforce announced a slew of new content on Trailhead for learning AI, including a Trailmix and numerous modules. If you’re like me and have gone through them, you may have thought to yourself: “So… am I a data scientist now?”
I decided I wasn’t, and I wanted more. To be fair to my friends at Trailhead, that’s not what the Trailmix says on the tin, nor does the new AI Associate Certification guarantee expertise. The good news is there are so many resources out there to learn AI, and there is no reason you need to remain in the dark.
In this article, I’m going to share four learning resources that have helped me expand my knowledge of AI. These include three free platforms and a fourth one that costs a small monthly fee. Each of these resources contains both lecture and practical exercises, usually in the form of Jupyter Notebooks – an interactive platform for running code, including Python, which is the most common language for AI engineering.
1. Coursera: Machine Learning Specialization
The Coursera course, Machine Learning Specialization, is probably the most popular learning resource to build your machine learning knowledge from the ground up.
The author, Andrew Ng, created the first version of this course in 2012 with the launch of the Coursera MOOC platform. He has reportedly taught around 8M people about machine learning through this and other curricula. I am currently in progress with this course and have really enjoyed it.
What I especially like about it is how it approaches the math behind machine learning. Machine learning is very, very mathy. For each new concept, it gradually builds up your understanding of the required math, providing just enough for you to be successful. For someone like me who did very little math in university, this is key.
2. Jon Krohn
Jon Krohn is a data scientist, AI educator, and podcaster. Before he started hosting the Super Data Science podcast, he created a number of tutorial series on different data science topics. Where Andrew Ng’s course provides just enough math to solve specific machine learning problems, Jon Krohn courses approach specific machine learning skills and spend time taking you deeper into those.
If you imagine the entire knowledge of machine learning, Andrew Ng’s courses slice the loaf horizontally, layer by layer, from top to bottom. On the other hand, Jon Krohn’s slices are more traditional, taking you through the three core mathematical disciplines of machine learning: statistics, linear algebra, and calculus. I’m planning on interspersing Krohn’s training as I tackle the Andrew Ng Coursera course.
The Machine Learning Specialization and Jon Krohn’s machine learning series are very focused on the core skills and knowledge of machine learning. However, everyone knows that the cool stuff to play with is generative AI and large language models. The next two resources are focused on tackling these most recent developments in machine learning.
3. Cohere: LLM University
Cohere is an AI company that offers a great set of learning resources, like LLM University, which includes a set of lectures. The Cohere team has come up with some great technical training that walks you step-by-step through core principles and knowledge of generative AI. Most importantly, “Module 1: What are Large Language Models?”, is as good a high-level explanation of how LLMs work as you will find.
Cohere aims to be an AI platform that any developer can use to create AI applications. Consequently, much of the rest of the training looks at LLMs through the Cohere lens. However, much of what you learn can be applied to other models and platforms.
4. Andrew Ng: DeepLearning.AI
Andrew Ng is not one to rest on his laurels. To say he’s prolific in the space of AI is a complete understatement, and he has done the same with learning AI.
His learning platform, DeepLearning.AI, has a broad range of AI courses. There are a number of longer paid courses, but there are also a load of free short courses. These are great for picking up and learning some of the very cool AI skills, frameworks, and platforms out there right now. The Prompt Engineering for Developers course helped me start to understand how powerful LLMs were.
Curious about how diffusion models work for images? You can take the How Diffusion Models Work course. Or, if you just want to gain a better basic understanding of artificial intelligence today, check out the AI for Everyone course.
Each course is usually four hours of learning or less, and you’ll walk away having had a taste of building something and making it work.
Hopefully, you’ve found some useful information and plenty of resources to continue your journey into AI. If you enjoyed my latest blog about learning AI, be sure to stay tuned for upcoming posts!