Separating fact from opinion has always been a goal for machine learning usage in search quality development. Speculative journalism has it’s place as sharing opinions can be a good thing and can help people develop new ideas and improved systems for all kinds of purposes. It is important to isolate fact from fiction especially with YMYL (Your Money Your Life) searches, which can greatly affect peoples financial, physical, and mental health if bad information is given.
Recently John Mueller spoke on the extent to which machine learning is used in search and it was pretty interesting to learn this new information.
“There are always so many algorithms in play; some are more suitable for ML (Machine Learning) than others. Suitability also requires room to remove bias, allow debugging, allow critical corrections, etc. — in addition to delivering better results.” – John Mueller
If topics are easy to understand and a website has clear and concise content machine learning can do much of the work alone but when concepts get more complicated and searches are extremely open ended as far as intent goes sometimes machine learning, in it’s current state, has to take a back seat and let human intervention take over. Because of this, there is no real negative effect machine learning has on SEO and can actually makes the SEO process quicker by speeding things up if the content being SEO’d is simple and easy to understand.
Google offers a helpful glossary on machine learning and as someone who is reading up on machine learning regularly it is helpful.
Overall, it is a relief to hear John tell us that machine learning isn’t always used and it depends on the circumstances since it is one less thing that any SEO has to worry about as far as causing a ranking disturbance.