Google’s artificial intelligence algorithm is well known by now by the SEO industry; we have an understanding that RankBrain is supposed to significantly improve the user’s search experience, but what we seem to be having trouble understanding is how quickly this new AI is changing the SEO landscape.
First, it’s important to understand how AI works under the three classifications.
Different Types of AI
- Artificial Narrow Intelligence is programmed only to learn and excel at one particular task. Tech Crunch uses the example of supercomputers learning to beat world champion chess players.
- Artificial General Intelligence is the equivalent of how a human brain functions and processes everything from sensory stimulation to language, empathy and decision making.
- Artificial Superintelligence is going above and beyond the capabilities of the human mind. Supercomputers such as IMB’s Watson, and Google’s RankBrain strive for this level of processing and learning.
RankBrain’s creators believe that knowledge is scripted in connections between neurons — also known as learning. RankBrain is also known as a deep learning algorithm that strives to make connections and build pathways as a human neuron would — except it is apply this knowledge for search engines and users (a narrow field).
How Quickly is SEO Progressing?
Now that we know RankBrain to be an Artificial Narrow Intelligence but with the capabilities of Artificial Superintelligence, it’s time to discuss how quickly SEO is changing because of this technology.
Whenever Google announces another large algorithm tweak, many SEOs assess months or even years of previous data and try to piece together how rankings have shifted across all websites. However, it’s time to look at what type of website has been affected by a particular algorithm update in order to figure out how Google works.
RankBrain exists to understand all of Google’s core algorithms. Depending on the mixture of these algorithms, RankBrain learns what type of results to display. For instance, for certain types of websites or industries, Page Titles and Meta Descriptions might be more important than internal links or H-tags, therefore it will pull websites into search results with the best META titles. In other instances, H-tags may be the deciding factor on what websites Google pulls up — it all depends on the search and the algorithm mixture.
Note that Google’s RankBrain can also learn what “good” and “bad” sites look like, and it can further classify good and bad sites within a single CRM. There are certain structures of data that it looks for and if one website provides more important data in a specifically ordered way than another, then this creates a hierarchy.
The SEO industry in the next decade is set to become very technical now that we know RankBrain is closer to human functionality (if not already on par) than we once believed. Hopefully this will allow SEO to get easier as we start creating content and websites more for people, not bots.