How continuing advances in image-reading tech benefit you

The Internet is a vast source of information, an archive of data that grows far faster than we can catalogue it.
With unquantifiable amounts of data being input and uploaded every day, it’s impossible to keep up with, and without the proper tags to make it searchable, such information is effectively useless. 
But thanks to the independent work of developers at Google and Stanford University, image-reading software can be put to work identifying and labeling such data.
For many years “computer vision,” or the ability of a computer to read images, has been limited. 
Most technologies have only been capable of recognizing single objects, and their subsequent descriptions are often inaccurate and lacking in detail.
For example, if the computer were presented with an image of a cow grazing, it might give the caption “a cow,” which doesn’t give you as much a sense of the picture as the description “a cow grazing in a grassy field” might.
Researches have been working to overcome the limitations of computer vision by honing the computer’s neural networks and training them to recognize patterns. 
By focusing specifically on the networks for image recognition and language processing, those at Stanford and Google are creating software capable of reading images and videos and cataloguing them with captions that describe what they contain.
The better computers get at processing and sorting visual information, the more accurate the results we get from search engines become.
This improves research for all internet users. 
The amount of data available for searching will also continue to grow and improve. 
This is because, like many artificial intelligences these programs grow more accurate with continued use, as they rely on learning and experience to develop effective algorithms for interpreting data. If given time they could grow to be as perceptive as we are.
The applications for this software are limitless. Not only can it be used to inventory Internet data, but it could prove helpful in the development of technologies to assist the blind as computers capable of accurately reading visual information could learn to interpret and even navigate physical spaces.
It also carries implications for technologies like facial recognition or the collision prevention AI’s common in many new cars. 
The software is still being refined, and it may be years before we see it put to use anywhere outside image archiving, but the possibilities are undeniably exciting.