Recently, artificial intelligence laboratories have begun to fix flaws in robots using a similar system to the “Alexa” personal device made by Amazon. One major problem the scientists have found is that the robots don’t understand the words “Pick it up”. For example, if a person placed a pencil on a desk and asked the robot to pick it up, it wouldn’t be able to. The robot does not understand what “it” is. Is “it”, the desk or the pencil? There are three major types of memory including semantic, episodic and declarative. The old system only understood semantic memory, which is fact based such as the grass is green.
Due to these problems, a new type of system was created by the MIT student computer lab called ‘’Com-Text’’. Com-Text stands for command in context. The Com-Text can handle situations similar to the pencil and desk scenario. If you command the “Com- Text” to pick the pencil up and place it down on the desk, it would be successful. Now you may be wondering, how does something like this work? Com-Text can observe a range of visuals stimuli and language cues, to decipher and learn about an object’s size, shape, position, type and even if it belongs to somebody. From this knowledge base, it can then reason, analyze and respond to commands. By experimenting with this line of research, many things are possible for the future of “Com-Text”. This could potentially enable better development of robots and even self-driving cars.
When I read this article, I had several ideas and concepts catch my attention. I was originally shocked to find out that even with such complex robots, there were basic issues that were a problem, like understanding what “it” meant. In addition, I found that with the robots more complete understanding of language, they were now able to understand new commands.
New doors are open with the development of these new robotic systems. One idea, is you can train robotic systems to teach other languages such as German, Japanese, French and Italian. This would open new opportunities for people learning new language. This is one of the several positive impacts of this new advancement. With voice recognition, robots can be used around the house.