Sundar Pichai, the CEO of Google, announced a few AI-model updates that also include Google MUM at Google’s annual conference. These steps are taken by Google to make users’ work easier while interacting with technology.
A recent post by Pandu Nayak, Vice President for search at Google, explained MUM and their goal to provide simplified results for complex queries.
Let’s first try to understand Google MUM update.
What Is The Google Mum Update?
Google MUM stands for Multitask Unified Model, which is a new AI-based technology by Google to understand and answer complex search queries.
Like the other powerful language models, such as LaMDA, GPT-3, and BERT, Google MUM is based on the transformer architecture. Transformer architecture is a neural network model that transforms a sequential input to a sequential output, such as speech recognition and text-to-speech transformation.
To understand it better, Google quoted an example. Suppose, you want to hike Mt. Fuji next fall after you have hiked Mt. Adams. Now, you want to prepare for it and understand what to do differently. You may need to search factors, such as the elevation of the mountain, the temperature during the fall season, the right equipment, or challenges one faces while hiking Mt. Fuji. In today’s scenario, it is easy for you to find an answer to all these questions. But you may need to perform a lot of different search queries to gather answers for all these questions.
Related: Google Launches Menu Icons For Search Results
Google says that this is a complex search query as you are not reaching the right answer with fewer searches.
As per Google, users conduct eight query searches on average for any complex task.
Thus, the main aim of the new Google MUM update is to help users easily find answers to their questions by performing a fewer number of searches.
Is BERT Different From MUM?
BERT stands for Bidirectional Encoder Representations. It is built on transformer architecture. BERT considers both the left and the right context to understand the nature of the word in the search query.
For example, “bank” as an individual word refers to a riverbank, a financial institution, or a high mass. BERT makes it easier for the search engine to distinguish between the words based on the other words used in a search.
Thus, it decodes the meaning of an ambiguous text by using the other words in the search query and helps generate more relevant results. Basically, it helps Google to understand the intent of a user.
Related: Core Web Vitals For Google SEO
Similar to BERT, MUM is also built on transformer architecture. However, MUM is 1000 times more powerful than BERT.
Some of the features of Google MUM, making it different from BERT are:
- It is trained in 75 different languages which helps the search engine to have a comprehensive understanding of the information.
- As the name suggests, Google MUM is built to perform multiple tasks at the same time to gather information for users.
- Google MUM is multimodal, which means it also understands the information available in the form of text, images, videos, or audio.
Is MUM A Ranking Algorithm Update?
In the recent past, Google announced the BERT update that changed how the search engine interprets the context of words in a sentence. During that time, Google maintained its answer to a question about how to optimize for BERT. As per Google, there is nothing to optimize for. It simply means, you cannot optimize for BERT, but you can optimize for users.
Similar is the case with the upcoming update MUM. It will help you find answers for any complex query, remove language barriers, and understand information across different mediums. Thus, there may be a significant impact on the concept of SEO.
Related: Complete SEO Guide
However, you have to wait to find out how exactly it will impact SEO. But you can focus on the following actions to improve SEO as per the new update:
- Offer high-value content to your users.
- Provide comprehensive and comparison studies to your users.
- You may no longer need to worry about translating the language to be shown on the result pages.
- Optimize images, video, and audio.
Related: WordPress Plugins For Image Optimization
How Will Google Handle Complex Search Queries With MUM?
Google is trying to integrate multiple tasks to create one search experience for its users. If the new MUM update works well, then it will allow users to perform complicated searches easily.
Related: Page Experience Report In Google Search Console [Guide]
Here is how Google will handle complex search queries with the help of MUM:
- Remove language barriers: As mentioned above, Google MUM is trained in 75 different languages. Thus, it will help the search engine to transfer knowledge across languages. For example, there is relevant information for your search in Chinese. But today, it may not appear in your search results if you do not type the query in Chinese. However, with Google MUM, it will be possible to find relevant information in any language irrespective of the language you choose to search with.
- Help understand information across different types formats: Google MUM is multimodal. Therefore, in the future, if you use any content format to ask a query, the search engine will show you the relevant results. Continuing with its last example, Google mentioned that you will be able to show a picture of your hiking shoes and ask, “Can I use these to hike Mt. Fuji during the fall season?”.
- Apply advanced AI to search: Google carefully tests every step that it takes forward with AI to improve the user accessibility to relevant information. To avoid any bias in its systems, it evaluates any pattern that may indicate bias in ML.
When Will The MUM Update Go Live?
Google has not yet announced the date when the MUM update will go live. However, it is running internal pilot testing of MUM.
According to Google’s announcement about the MUM update, one can see the MUM going live in the coming months and years.
Final Thoughts
The Google MUM update is still in its initial phase of testing. However, once it is live, information and content will be easily available in any language and medium. It will enable users to find relevant answers to their complex queries easily and faster.
These steps taken by Google are always incredible. As we will see changes in the search results once the MUM update goes live, we may also see SEO enter another stage of transformation.
