Google Search Generative Experience: how to improve search with generative AI

Artificial Intelligence is booming. More and more companies are opting for it to improve their services and Google was not going to be less. The company is developing Google Search Generative Experience, a new search experience that facilitates the entire information gathering process thanks to AI. If you are interested, in this article we tell you in depth everything it can do for you.

What is Google Search Generative Experience?

Google Search Generative Experience (SGE) or generative search experience is Google’s commitment to include Artificial Intelligence in its search results. With this new addition, the search engine’s first answer will be generated by AI and will respond directly to the question asked by the user. In this way, the public will not have to enter the different URLs shown to them, being able to see the information they need (and other extra elements such as the sources used) quickly.

In short, SGE is a new search experience that integrates Artificial Intelligence and helps users to make their research work faster, easier and more efficient. Thanks to this new functionality, the work of searching for information and gathering it will no longer be done on our own, but Google will help us.

For now, SGE’s response is labeled as “Generative AI is experimental”, as it is not yet in its final version.

How to access Google Search Generative Experience

SGE falls under Search Labs, a Google experiment program that is not yet open to the public and can only be accessed by certain accounts after signing up for this waiting list. However, at the time of writing, it is not available in Spain. For now it is only available to U.S. residents and in English, so if you want to access it you must use a VPN. However, they will be opening it up to new countries until it is implemented by Google in the future.

Those lucky enough to sign up and be accepted will receive an email informing them that they can start testing the different Search Labs functions.

How does Google Search Generative Experience work?

As reported by Google, its new AI search experience uses a variety of large language models (large language models or LLM) including PaLM2, an advanced language model created by Google, and MUM (Multitask United Model), Google’s latest AI-based algorithmic version of search.

The company has also reported that SGE has been trained to be able to perform specific search tasks, including identifying high-quality internet results that verify the data presented as a result.

On the other hand, according to Liz Reid, Google’s vice president of search, with this new experience the company intends Google to limit its response capacity to some areas in order to make queries more and more secure. Thus, some questions related to medical, financial, dangerous or explicit topics will not be answered.

Likewise, when there are data gaps or unreliable information for a question, Google will not generate an AI response either.

Why doesn’t SGE allow conversation as in ChatGPT?

Another of the company’s objectives in creating this new AI search experience is to offer answers in which objectivity prevails over fluidity, despite the fact that users prefer fluid answers. This has been done completely intentionally by the organization to keep the answers in as neutral a sphere as possible.

The basics of Google Search Generative Experience

With all the information we have already given you about this new search experience, we can establish what the 5 basic pillars of this new Google addition are:

• Security: Google does not want to give first-person answers or create conversations between the AI and the user, as users are more likely to trust the answers they are given and the company cannot verify that all its answers are 100% true.
• Quality: the organization intends its AI answers to be of high quality, so it wants to refrain from answering questions in some sensitive areas (already mentioned above).
• Agility: AI reduces the number of steps users have to take to find the answer they are looking for.
• Relevance: in this case we are talking about advertising. This AI also aims to show more relevant ads to users and in the best possible way.
• Ecosystem: although the AI gives an answer to the user’s question, Google does not want to stop encouraging the public to go deeper into the topic and delve into the content created by other users. Therefore, when AI-generated content is displayed, the content sources on which it was based will appear next to it.

Are SGE and Bard the same thing?

Although both functionalities are underpinned by AI, they are very different. Bard is a Google chatbot that is completely independent of the search engine, while SGE has been trained according to result ranking criteria for a specific search usage. As a result, the quality of SGE’s answers is higher and it is less prone to errors. This has also been possible because it has received human feedback during its training.

Another thing the company wanted to improve with SGE over Bard is the lack of citations and references. With this new search experience, sources appear and the ecosystem of creators receives much more support.

Real-world examples of what generative AI offers

Better searches

This is the clearest example of what SGE can do to improve your searches and prevent you from searching for information in different places, with the amount of time that entails.

In this example we see how, in addition to answering the user’s question, SGE shows on the right side the different sources from which it has collected the information and at the bottom it suggests related questions.

Shopping support

SGE promises to be a great ally when it comes to online shopping. One of the things we liked most about this new update is that it gives advice on the product the user is looking for to make their purchase the best possible.