I am starting to review Introductory AI courses on this blog, while developing my own course that give the audience understanding on what is behind modern AI architectures, why AIs are making mistakes, what are economics of AI, why companies mostly loose money why introducing machine learning and AI into their businesses, and what are areas of research that can improve today’s data technologies and robotics.
Introductory AI courses from leading organisations like OpenAI, Microsoft or MIT share a set of core topics and learning objectives. These courses are designed for beginners and aim to provide foundational knowledge of AI and robotics and introduce practical skills for working with AI tools and concepts. None of these courses dive into the evolution of human intelligence – something that might help the audience to understand human vs. machine differences in learning, perception building, and forecasting.
Most online courses share modules and topics. They review definitions and the history of AI, talk about narrow AI and its real-world applications, and General AI, which currently does not exist. I consider it a valuable part to discuss several algorithms to illustrate how AI approaches problem-solving, e.g., search algorithms. Furthermore, the differences between symbolic AI and artificial neural networks (ANNs) are explained. The machine learning part addresses supervised, unsupervised, and reinforcement learning. Most courses introduce concepts of transformers and generative AI, educate about Natural Language Processing (NLP) and Computer Vision, and discuss AI applications and practical tools, including working with Python and popular AI libraries.
My personal favourite among introductory courses is by Andrew Ng (click here). You can enrol for free, and you don’t require any knowledge before starting. Andrew has an Introductory course on generative AI on the same platform. I will take it and report, but people I trust are praising the content.
I am cautious about recommending OpenAI courses (https://academy.openai.com/). There has always been a hidden agenda to emphasise costly ChatGPT and let you ‘forget’ that there might be alternatives.
If you have first experience with Python and a basic recollection of your school calculus, I recommend a YouTube tutorial by Andrej Karpathy, one of the best AI researchers and developers of our time. You will understand backpropagation and what is under the hood of a neural network.
I encourage everyone to experiment and try out a few courses, even if some content might appear repetitive. I understand you might claim you don’t have the time. If you pause and reflect for a bit, you might notice that it’s not about having time. It is about your agency.
Agency is the most critical feature of intelligent life. On my podcast and on this website, as well as in my Romy&Roby book series, there are numerous reflections and conversations about what ‘agency’ entails. Agency in philosophy refers to the capacity of an individual or entity to act intentionally and make choices within a given environment. It is distinct from mere reaction to external forces; it involves purposeful, goal-directed behaviour and the exercise of (free) will. You won’t become an AI geek by taking one introductory or even intermediate course. However, you will send a signal about your priorities and how you think about your future self, both at work and individually.