Join Doug Rose for an in-depth discussion in this video The Internet of Things, part of Artificial Intelligence Foundations: Thinking Machines.
- Recently it's become a lot less expensive to add connectivity to everyday things. Nowadays there are thermostats and smart watches that connect with each other and the world. There are also cheap sensors that check room temperature, their location, and some can even upload your medical data. This new way of connecting is commonly called the Internet of Things, sometimes you'll see it as IoT for short. These IoT devices can be a massive new source of data.
They'll give AI assistants a new way to interact with you. Some of the classic examples are embedding small devices in food packaging so your system can tell you when you're nearly out. Maybe your smart watch will tell your thermostat when you're on your way home. IoT also allows these devices to communicate with each other. You could unlock your doors as you approach or have your computer turn on when you sit down. All of your different things could be buzzing around you like mosquitoes on a humid day.
This flurry of activity will create even more data for machines to process. In some ways, these new devices will make it much easier to create data than it will be for humans to analyze the data. That's why many IoT companies also invest heavily in AI programs. Not only will AI allow organizations to see new patterns, it will also give them the ability to quickly react.
One area that has a lot of potential is IoT medical devices. You can now purchase an electrocardiogram, or EKG sensor, that's nearly as accurate as the ones in your doctor's office. These devices can check your heart's electrical activity. They're now inexpensive enough that many companies are embedding them into smartphone cases, there are even watches. These companies can also find patterns in the data using unsupervised machine learning on an artificial neural network.
The network could go through the EKG data of thousands or maybe even millions of different participants. They could use this data to find patterns that might accurately predict if you're having an upcoming health issue. They can use regression analysis to compare their EKG readings to many other independent variables. It might show that an increased heart rate in the morning is a warning sign for different health issues. The challenge with this approach is that your neural network may only find patterns, but not necessarily breakthrough cures.
It would simply notice that people who had your EKG patterns were more likely to run into trouble, then it would encourage you to see your doctor. That might be an obstacle if you're looking to prevent illnesses. Remember that machine learning and artificial neural networks create a black box. No one except for the network really knows how the machine finds patterns. You could react and predict, but humans don't necessarily understand.
It's certainly helpful to have machines find predictable patterns, maybe they find something like people are more likely to have health events on a Wednesday, but good doctors should always look for causes. That's why some companies might decide to use symbolic reasoning instead of machine learning. Here, doctors could clearly map out patterns of different EKG results. They could tell machines if they see these patterns, then the user should probably see their doctor.
You've already seen the Internet of Things having an impact on physical fitness. It's very common to see people with fitness bands tracking their heart rate and other key metrics. Once these IoT sensors become more sophisticated, then we'll have a lot more data for our AI programs.
This course will introduce you to some of the key concepts behind artificial intelligence, including the differences between "strong" and "weak" AI. You'll see how AI has created questions around what it means to be intelligent and how much trust we should put in machines. Instructor Doug Rose explains the different approaches to AI, including machine learning and deep learning, and the practical uses for new AI-enhanced technologies. Plus, learn how to integrate AI with other technology, such as big data, and avoid some common pitfalls associated with programming AI.
- The history of AI
- Machine learning
- Technical approaches to AI
- AI in robotics
- Integrating AI with big data
- Avoiding pitfalls