For example, the Apple Lisa was an interesting and useful piece of technology that demonstrated the usefulness of the GUI to business users who had never seen one before. It solved the need to make computers friendly.
However, it failed because it didn’t build a base of adherents. The computer simply failed to live up to the hype surrounding it. The next system that Apple built, the Macintosh, did live up to the hype a bit better — yet it built on the same technology that the Lisa used. The difference is that the Macintosh developed a considerable array of hard-core adherents and that is what machine learning needs.
Considering the role of machine learning in robots
A goal of machine learning today is to create useful, in-home robots. Now, you might be thinking of something along the lines of Rosie the robot found in The Jetsons. However, real-world robots need to solve practical and important problems to attract attention. To become viable and attract funding, a technology must also amass a group of followers, and to do that, it must provide both interaction and ownership.An example of a successful in-home robot is the Roomba from iRobot. You can actually buy a Roomba today; it serves a useful purpose; and it has attracted enough attention to make it a viable technology. The Roomba also shows what is doable at a commercial, in-home, and autonomous level today.
Yes, the Roomba is a fancy vacuum cleaner — one with built-in smarts based on simple but very effective algorithms. The Roomba can successfully navigate a home, which is a lot harder than you might think to accomplish. It can also spend more time on dirtier areas of the home. However, you still need to empty the Roomba when full; current robot technology does only so much.
You can find other real-world robots that people are using to perform specialized tasks, but you won’t find them in your home. With each of these cases, the robot has a specialized purpose and acts in a limited number of ways. Other sites present other robots, but you won’t find general-purpose uses in any of them. Before robots can enter a home and work as a generalized helper, machine learning needs to solve a wealth of problems, and the algorithms need to become both more generalized and deeper thinking.
Using machine learning in health care
An issue that is receiving a lot of attention is the matter of elder care. People are living longer, and a nursing home doesn’t seem like a good way to spend one’s twilight years. Robots will make it possible for people to remain at home yet also remain safe. Some countries are also facing a critical shortage of health care workers, and Japan is one. As a result, the country is spending considerable resources to solve the problems that robotics present.The closest that technology currently comes to the vision presented by an in-home nurse robot is the telepresence robot. In this case, the robot is an extension of a human doctor, so it’s not even close to what the Japanese hope to create in the near future. Like the Roomba, this robot can successfully navigate a home. It also allows the doctor to see and hear the patient. The robot partially solves the problem of too many patients in too large a geographical area and not enough doctors, but it’s still a long way off from an autonomous solution.
Creating smart systems for various needs
Many of the solutions you can expect to see that employ machine learning will be assistants to humans. They perform various tasks extremely well, but these tasks are mundane and repetitive in nature. For example, you might need to find a restaurant to satisfy the needs of an out-of-town guest. You can waste time looking for an appropriate restaurant yourself, or you can access an AI to do it in far less time, with greater accuracy and efficiency.Another such solution is Nara, an experimental AI that learns your particular likes and dislikes as you spend more time with it. Unlike Siri, which can answer basic questions, Nara goes a step further and makes recommendations.
Using machine learning in industrial settings
Machine learning is already playing an important part in industrial settings where the focus is on efficiency. Doing things faster, more accurately, and with fewer resources helps the bottom line and makes an organization more flexible with a higher profit margin. Fewer mistakes also help the humans working in an organization by reducing the frustration level. You can currently see machine learning at work in- Medical diagnosis
- Data mining
- Bioinformatics
- Speech and handwriting recognition
- Product categorization
- Inertial Measurement Unit (IMU) (such as motion capture technology)
- Information retrieval
- Analyzation: Determining what a user wants and why, and what sort of patterns (behaviors, associations, responses, and so on) the user exhibits when obtaining it.
- Enrichment: Adding ads, widgets, and other features to an environment so that the user and organization can obtain additional benefits, such as increased productivity or improved sales.
- Adaptation: Modifying a presentation so that it reflects user tastes and choice of enrichment. Each user ends up with a customized experience that reduces frustration and improves productivity.
- Optimization: Modifying the environment so that the presentation consumes fewer resources without diminishing the user experience.
- Control: Steering the user to a particular course of action based on inputs and the highest probability of success.
Understanding the role of updated processors and other hardware
There are five schools of thought (tribes) related to machine learning. Each of these schools of thought tell you that the current computer hardware isn’t quite up to the task of making machine learning work properly. For example, you might talk to one tribe whose members tell you of the need for larger amounts of system memory and the use of GPUs to provide faster computations.Another tribe might espouse the creation of new types of processors. Learning processors, those that mimic the human brain, are all the rage for the connectionists. The point is that everyone agrees that some sort of new hardware will make machine learning easier, but the precise form this hardware will take remains to be seen.