This new AI tool from Google could change the way we search online in 2021October 8, 2021
Tech tools have provided numerous learning and earning opportunities. They have enabled subscription-based companies to introduce new services like Cox bundles. And there’s no denying that technology will continue to define the way we search online. Google is a prime example of that.
The ubiquitous search engine has used different technologies over the years to cater to the browsing experience of its users. Nowadays, it is developing new technology to understand the subtlety of human language on a large scale. Before we discuss the new tech, let’s take a brief look at the history of Google.
Origins of Google Search
Google originally began as a research project called “BackRub” in 1996. The first version of Google officially arrived on the market in August 1996. Ph.D. students Sergey Brin and Larry Page were the authors of the said research project. Google, as we know it today, arrived on the scene in 1998. It’s was designed to map the web in addition to organizing (or indexing) all the content on it.
If you want to understand the Google search index, you will have to think of it as the index on the last pages of a back. An index is an alphabetical list of alphabets or items concerning the pages on which they are mentioned. However, there are several major differences between the index of a book and that of Google. A book might have thousands of pages but the web has trillions of pages.
Another difference is that the web shows you a combination of words and not just one word at a time. Google receives and answers billions of queries every day. And not every query is typed with incorrect spellings. There are typos and spelling mistakes. Google tries to understand the errors to provide users with accurate results possible.
Google Search Ranking Mechanism
Just like every other search engine, Google uses an algorithm to provide results relevant to your query. It determines the relevancy of the content using factors like language, the freshness of the content, and location. Language understanding is perhaps the most important factor. To provide answers, Google has to understand the meaning of the query.
Now, on to the tricky part. Because language is complex, algorithms can’t comprehend it the way humans do. Every language has its nuances and subtleties. But programmers have developed ways to make software understand these complexities to some extent. A prime example is the synonym system. The first version of the synonym system arrived on Google some 16 years ago.
The New Artificial Intelligence Technology
The future of search is about to change, thanks to a new AI tool called MUM. The full form of MUM is Multitask Unified Model. The technology is programmed to understand the complexities of human language. Through MUM, Google intends to answer even the most abstract questions. The search engine has already used the technology to provide results about COVID-19.
MUM and the COVID Dilemma
The pandemic brought panic and fear with it. Being the world’s largest search engine, Google was confronted with the task of understanding the many names of COVID-19, in different languages. The search team at Google, headed by Pandu Nayak, worked hard and created MUM. Using the technology allowed them to “generate over 800 names for 17 different vaccines in 50 different languages”.
However, Google has only deployed MUM to handle COVID-related queries so far. According to Nayak, much work needs to be done to understand the human language. He says Google will use the technology to improve language classification factors.
The Future of MUM
Google sees MUM as the next face of machine learning. The tech giant has been working to create a new and improved version of machine learning for over a decade. According to Google reps, MUM can train multiple languages at once. It also can acquire complex knowledge. MUM presents the next evolution of machine learning. It has already shown tremendous capabilities.
From a user’s perspective, MUM is that virtual friend that you go to for advice. Why? Because it can understand your tone and can perceive your intent. Suppose you ask Google, “I have jogged for several miles today. What exercise should I do next to improve my cardiovascular strength?” This is the question you would typically ask your trainer. But MUM has shown that it can answer such questions.
Tech experts at Google are already working to improve MUM. The medium-term goal is to make the software understand the relationship between words and images. Of course, the long-term goal is to create a program that can understand the nuisances and subtleties of the language and even generate it. Google wants to provide a platform where people would be able to converse with the AI like the way they talk to their friends.
We can only wait to find out what Google has in store for us.