When a newborn baby enters the world as a clean slate He/She learn and develop to be an adult. Babies have an innate knowledge that helps them to voraciously learn and rapidly adapt. But that's not what AI does. In case of AI, it's machine learning. That's when a computer learns everything it needs to know from a giant dataset using trial and error. As a human or babies, there are just somethings you can learn from trial and error. But many computers scientists argue that most human skills are learned and AI could learn them too, without the need for pre-loaded rules. Still, a growing number of researchers are attempting to encode AI with a bit of common sense.
The current craze in AI are neural networks, collections of simple computing elements, loosely modelled as neurons in the brain, that adjust their connections as they encounter more data. They have produced incredible achievements in the past few years, from facial recognition to beating humans at chess or poker games. But neural networks require thousands of training examples to reliable from associations. Even then they can produce some embarrassing blunders. Compare this to a child who can see an image just once and after that instantly recognize it in other contexts.
Some AI can play classic Atari games with superhuman skill, but when you remove all the aliens but one the player inexplicably become a sitting duck. Different labs are categorizing human instincts and then they try to encode them into AI. These systems sit somewhere between pure machine learning and completely programmed.
One team developed an AI called interaction network. They have embedded the rule that such a thing as objects and relationships between those objects exist. This is like a babby's innate parsing of the world into objects. In tests, once the AI learns the specifics properties and relationships, it is able to predict the behaviour of falling strings and bouncing balls in a box. Another group's " neural physics engine" beats less structured neural networks for predicting ball collisions in containers. A lab created an AI which has an embedded rule to treat letters as objects and separate them from their background. This allowed it to solve CAPTCHAs better than other neural networks that were trained with 50,000 times more data.
We are far away from AI that can truly think like humans, but with these latest attempts to reproduce a common sense artificially. Researchers believe they will get closer to creating robots that can fully interact with the world the way we do.
Machines that start like a baby and learn like a child. What Do You All Think The Science Thinkers, How You Will Feel In A World With AI, Tell Us In The Comment Box.
Also Read:-Self-Driving Cars | What Kind Of Problem We Are Facing To Develop A Self-Driving Cars?
The current craze in AI are neural networks, collections of simple computing elements, loosely modelled as neurons in the brain, that adjust their connections as they encounter more data. They have produced incredible achievements in the past few years, from facial recognition to beating humans at chess or poker games. But neural networks require thousands of training examples to reliable from associations. Even then they can produce some embarrassing blunders. Compare this to a child who can see an image just once and after that instantly recognize it in other contexts.
Some AI can play classic Atari games with superhuman skill, but when you remove all the aliens but one the player inexplicably become a sitting duck. Different labs are categorizing human instincts and then they try to encode them into AI. These systems sit somewhere between pure machine learning and completely programmed.
One team developed an AI called interaction network. They have embedded the rule that such a thing as objects and relationships between those objects exist. This is like a babby's innate parsing of the world into objects. In tests, once the AI learns the specifics properties and relationships, it is able to predict the behaviour of falling strings and bouncing balls in a box. Another group's " neural physics engine" beats less structured neural networks for predicting ball collisions in containers. A lab created an AI which has an embedded rule to treat letters as objects and separate them from their background. This allowed it to solve CAPTCHAs better than other neural networks that were trained with 50,000 times more data.
We are far away from AI that can truly think like humans, but with these latest attempts to reproduce a common sense artificially. Researchers believe they will get closer to creating robots that can fully interact with the world the way we do.
Machines that start like a baby and learn like a child. What Do You All Think The Science Thinkers, How You Will Feel In A World With AI, Tell Us In The Comment Box.
Also Read:-Self-Driving Cars | What Kind Of Problem We Are Facing To Develop A Self-Driving Cars?
No comments:
Post a Comment