This specialization is born out of one of the most important insights I've gained about AI. If machines can learn, we can also teach them. In fact, that might be a defining characteristic for whether a system can learn. Can it be taught? Gardner sided in 2018, that there are approximately 10,000 data scientists in the world, let's approximate. This means that they are on the order of 10,000 AI experts who can design, and build autonomous AI from scratch using code. Most of these experts hold PhD degrees in areas related to AI. In contrast, they are on the order of 10 million software engineers in the world. Most of these developers specializing writing software applications, but are not specialists in AI. Their area of subject matter expertise is writing software, and they can do this across many diverse applications. Then there's 100 million or so process experts in the world. These mechanical engineers, chemical engineers, process engineers, controls engineers, supply chain analysts, logistics analysts, and many others, design and manage complex equipment, and processes, and they know these equipment, and processes inside, and out. This specialization is aimed at the 100 million process experts out there who manage, and seek to automate complex equipment, and processes. Whether you're an engineer, supply chain analyst, or an operator, there's no need for advanced degrees in machine learning or computer science to design, and build useful autonomous AI. Meet a former student of mine, Grant Bristow from Bell helicopter. I'm an aerospace engineer by education. I got my Bachelor's degree from the University of Texas in Austin, and went on to get my master's at Georgia Tech. I think it's really important for the students that are taking this class to understand that you're doing all this work on real autonomous AI here at Bell with no previous data science experience, no computer science degrees, you really an aerospace engineer? That's correct, and I've had to acquire the expertise in data science, artificial intelligence, and software development to augment my formal education so I can actually solve today's problems. It's that combination of domain expertise in aerospace as well as the tools that those technologies, and the knowledge in those technologies provide, we can tackle quite complex challenges in our industry. Process experts have subject matter expertise in their field, but their challenge is to translate that expert knowledge into a system an AI can work with. This requires a design, and a learning plan for the autonomous AI. Take a listen to Sean Eichenlaub from Pepsi as he talks about how this course helped him to do just that. Can you tell us about your experience with the course? I thought the course is very useful in terms of learning how to design a brain to control a system. How to make sure you understand everything about the system first, and how to break it down into the key components or concepts, the decisions that are made, and the strategies that are used when controlling that process. Then you wouldn't necessarily always use deep reinforcement learning for every one of those concepts, but use it for where it's necessary, and then use other types of control strategies for other concepts. Absolutely. Did you find the content accessible as an engineer with no prior data science training or would you recommend that people gets a familiarity with ML first? I think it was accessible, I had some experience with machine learning, I was somewhat familiar, but I think it is still very accessible for somebody without experience. What do you think is the value of the machine teaching framework? I think it's the ability, like I said, to combine concepts that are both deep reinforcement learning, and traditional controls. The fact that it's, becomes a little bit more explainable, and the way it behaves, the fact that you can have other people are going to have more trust in it if you're using strategies that human operators already use. That's right. After you design your AI, I'll show you how to use a low code cloud-based platform to build your AI. You don't need to be a data scientist, you don't even need to be a computer programmer. The platform you will use in this course has all the tools you'll need as a process expert to build great AI. By the end of this specialization, you'll be able to create a brain design document, a blueprint for your AI, and you'll be able to build, and train your AI brain from your brain design document. What if I'm not an engineer? If you're a data scientist or software engineer with an eye on developing useful AI, this specialization is also for you. Software engineers often need to build an integrated autonomous AI into their software applications. Data scientists are often asked to design, and build autonomous AI using only data. In this specialization, you'll learn how to work with subject matter experts to discover skills, and strategies that can be taught to machine learning systems. The combination of teaching machines skills, and strategies, and practicing on the real-world data is leading to better AI systems. Innovation executives also often take my course. This specialization will be very helpful to you as you lead your organization. You'll learn how to identify problems where autonomous AI can make a huge positive impact. You may think, wow, this is pretty technical, but in the area of industrial processes, especially manufacturing, and logistics, this level of technical detail is required for your job too, but you already know that. Designers, and builders are sometimes different. I need you to know that the people who design brains, and the people who build brains are often different, and that's okay. Course 1 is a foundational course for innovation executives, designers, and builders. Course 2, and 4 are more focused on design, and Course 3 is devoted to building, it's all about implementation. But we wrote this specialization with both designers, and builders in mind, and it's very important for designers, and builders to understand each other's work with autonomous AI, we encourage everyone to take all four courses. Make an impact. Finally, this specialization is for people who want to solve important problems in our world using autonomous AI. Sometimes people assume that autonomous AI is about dancing robots, but this technology in creative hands, can do much good for people. If you come to this specialization hoping to solve key problems in society to make a positive difference in the world, I encourage you to use our machine teaching framework, to teach AI, to do something truly useful.
