Following the introduction of Bard, its artificial intelligence chat, and its integration into the search engine to help users find what they are looking for with a simple and clear answer in natural language instead of directing them directly to a web page, Google has detailed how will you order the results from now on within your website.
When a user performs a search, Google uses its positioning and search engine optimization (SEO) algorithms. to order all the results according to the originality criteria and that they come from a high-quality website. How do you determine quality? Through rankings of competence, experience, authority and reliability. “We focus exclusively on the content, not on how it is produced”, says the company. «Ten years ago, there was also concern about the rise of mass-created content by humans; but it never occurred to anyone that we should ban all of them in our search engine.
Google says that if 10 years ago it didn’t prioritize large traditional newspapers over blogs in its search engine, now it doesn’t have to prioritize human-generated content compared to which tools such as ChatGPT can generate.
Generating content automatically to climb positions in the SEO of your search engine has been penalized for years for violating its policies to fight against spam. The company clarifies that it has many years of experience combating these practices and points out that the consumer must take into account that not all content generated partially or totally by an AI has to be bad content or spam.
For Google SEO, the important thing is what is written and not who writes it
For example, some of the time, real-time match results or instant financial information can help journalists and readers deliver and consume well-written and in-depth content in record time.
For Google, AI is just one more way to empower people through technology.
However, despite what the company alleges, there are many users who have shared their frustrations with the search engine in recent years, pointing out how for SEO it is no longer a priority that the results contain exactly what you are looking for.
Popular is already the trick of adding “reddit” to all searches to find concise and human-created content instead of having to dive through dozens of results that are positioned according to Google’s SEO criteria and not precisely because they respond directly to the question. user query.
Will AI improve the user experience?
One possible solution is the inclusion, as they already plan, of answering what is being sought directly by means of an artificial intelligence assistant that, trained by reading millions of websites, can respond directly in a text. This would improve the user experience, but may pose long-term challenges for Google, which is a company whose income depends mainly on advertising. If the user who is looking for something is not directed to any website, the task of the search engine since it was created, but is answered from the main page, who will invest in advertising to be the first result in searches?
Google has a mutually beneficial relationship with the media: the latter create interesting content that the former links to via SEO. One obtains information, and the other visits that provide income through advertising or subscriptions if they have paywalls to access the content. With artificial intelligence involved and instant responses, that balance, which has always been a headache for the media, can become even more unbalanced.
From the user’s point of view, Google declares that “it is not a problem for its algorithms”, that no type of priority is given to content generated by a human to the detriment of that generated by a machine.
Different users, universal results?
Something that can be harmful and alienating, since there are billions of humans, and each one, even communicating the same ideas, introduces new details based on their training, life experience and even style. However, at the moment, there are very few artificial intelligences capable of generating content, and if all users generate content with them that are perfectly optimized to please Google’s SEO, those who come looking for information will always end up obtaining the same type of written responses. with the same tone, style and political bias.
An AI is programmed to give “safe” and “ethical” results, but this can directly harm the veracity of certain topics, even more so if the texts written by humans end up buried behind those generated perfectly to Google’s taste.
The opposite can also occur. Google SEO can offer as the first result information that is directly false or harmful to the user that has been generated by artificial intelligence that only knows how to calculate which word to put after the previous one through mathematical prediction operations.
If a tool like ChatGPT doesn’t know if what you’ve typed is true or not (it doesn’t distinguish between 2 + 2 is 4 or 2 + 2 is 5), how can Google be so sure that its AI will be able to discern between what is true and what is false?
They will continue to focus on critical issues using “consensus,” says Google. In other words, they will hide results that may pose a risk, according to their criteria, on a large scale, as occurred during the Covid-19 crisis.
Microsoft has already presented for the first time, after 20 years, a serious alternative to the Google search engine, integrating the latest and most powerful version of ChatGPT in its Bing search engine. It is more than likely that Apple will integrate a similar assistant into its iPhone, which could mean millions in losses for Google, which is the default web browser on Apple devices.