Keywords have been in widespread use for searching textual databases for more than half a century now. However, it is a wonder that, despite the technological advances in computing in the intervening years, it is still ubiquitous when it comes to search. So much so, Kalev Hannes Leetaru, an American internet entrepreneur and academic, ponders on why we are unable to replace keywords with something by which machines can understand the world.
More than half a century after the widespread introduction of the keyword search of large textual databases, it still reigns supreme. Of course, over the years there have been improvements. For instance, today, search engines leverage knowledge graphs and language models to reach beyond keywords and closer towards the meaning. Similarly, search histories, location tracking, search histories of others and other external knowledge can be brought to bear to help machines better understand what humans are looking for.
Yet, despite all this power, machines still operate at the level of words rather than ideas. Undoubtedly, synonyms and knowledge graphs can help connect slight word differences and dictionary relations. However, even the most powerful artificial intelligence (AI) systems cannot truly understand information in the way humans do, abstracting from words to the ideas, events, and emotions they convey. Even the much-vaunted smart home assistants with their voice interfaces can still only answer the most basic of questions that have been predefined and templated by their creators.
The idea of unnaturally expressing informational needs into the keyword language of machines has become so ingrained in society over half a century. Hence, despite having deep learning systems capable of understanding images and audios, they are used for converting audios and videos into text for facilitating keyword searches. That is why, as voice interfaces gradually supplant the keyboard, the keyword itself remains. So, one cannot help, but wonder what a keyword-less future might look like. What riches might a world hold in which machines can finally understand the world as humans do?
Click here to read the original article.
Please give your feedback on this article or share a similar story for publishing by clicking here.
