Artificial Intelligence has been a focal point of modern computer science and engineering, being researched by companies such as Elon Musk-sponsored OpenAI and Google's DeepMind. On smaller scales, people have created neural networks that generally accomplish specific tasks. Neural Networks are programs that are modeled on a humans' neural systems. They learn from a starting point of zero knowledge with a few guidelines about how something works. This can be used to accomplish tasks from identifying pictures of cats to driving a virtual car (source). In the video below, you can see the learning process of an AI learning to control a virtual car:
The AI controlling the above cars was given only the ability to turn and accelerate with no instructions on how a car is supposed to drive. Over the course of many runs, or generations, the AI learns how to drive a car from the starting point to the end successfully. This basic idea is the same in other AI. Given a task and a few basic rules, the AI repeatedly tries to complete the task, learning from its mistakes each time. Eventually, an AI can be able to drive, play chess, or even beat professional video gamers at their own game.
In the real world, AI is applicable to many tasks. Obviously, there are AI you can play against in games like chess that can be fun. They are also useful practically, as NVIDIA demonstrates with their self-driving car. This car has been taught to drive and use sensors to avoid obstacles. Filled with sensors, the platform has been tested on highways, in countryside, and in cities, and has been successful, and may soon be available to companies to purchase and use.
Other than discussions about human and machine progress, AI doesn't really have a place in the classroom. However, there are valuable lessons to be learned from it that can definitely be applied to a classroom setting. People are always told to learn from their mistakes, and that is exactly what an AI does. This lesson can be used in teaching, as some classes may benefit from an open-ended environment where students are given guidelines and an assignment. In some situations, it could certainly help for students to teach themselves and learn from their mistakes as they go. Obviously, this isn't always applicable, as students are not robots. Overall, its just a nice, useful, educational learning experience.
The AI controlling the above cars was given only the ability to turn and accelerate with no instructions on how a car is supposed to drive. Over the course of many runs, or generations, the AI learns how to drive a car from the starting point to the end successfully. This basic idea is the same in other AI. Given a task and a few basic rules, the AI repeatedly tries to complete the task, learning from its mistakes each time. Eventually, an AI can be able to drive, play chess, or even beat professional video gamers at their own game.
In the real world, AI is applicable to many tasks. Obviously, there are AI you can play against in games like chess that can be fun. They are also useful practically, as NVIDIA demonstrates with their self-driving car. This car has been taught to drive and use sensors to avoid obstacles. Filled with sensors, the platform has been tested on highways, in countryside, and in cities, and has been successful, and may soon be available to companies to purchase and use.
Other than discussions about human and machine progress, AI doesn't really have a place in the classroom. However, there are valuable lessons to be learned from it that can definitely be applied to a classroom setting. People are always told to learn from their mistakes, and that is exactly what an AI does. This lesson can be used in teaching, as some classes may benefit from an open-ended environment where students are given guidelines and an assignment. In some situations, it could certainly help for students to teach themselves and learn from their mistakes as they go. Obviously, this isn't always applicable, as students are not robots. Overall, its just a nice, useful, educational learning experience.
Comments
Post a Comment