As technology develops, the way we engage with it changes. Even if we just look at search engines, it’s changing all the time! Google has gone from being a search engine with ten blue links, to something much more. Not only are there more features on top of those blue links, we use Google for navigation, reviews, email, cloud storage, and more. This alone demonstrates how our relationship with certain tech has changed.
Google’s search engine has remained consistently present in many peoples’ lives, however. That function is something which isn’t so easily replaced…Or is it?
Recently, people have started discussing the use of AI tools like ChatGPT as a potential “next evolution” of search. Will ChatGPT replace Google? And how are people starting to use generative AI as part of their search journeys?
What is generative AI?
Generative Artificial Intelligence is AI which is capable of producing content in response to prompts - that includes text, images, audio, and so on. It learns by pulling and comprehending vast amounts of data and resources, and engaging with human prompts. The more it engages, the more it learns about natural speech and language modeling.
The concept of generative AI has been around for a while now, but it really came into the spotlight thanks to OpenAI’s ChatGPT in late 2022. This advancement in generative AI made it accessible to not just those with an understanding of tech and AI, but the general public also. In early 2023, ChatGPT got an upgrade to GPT-4 hailed as the most advanced generative AI yet.
To recap on how it works, the user will enter a prompt. The AI will then interpret the prompt, and produce an answer. For example if you were to ask it to produce a blog outline based on a title and some basic info, it will produce that. The prompts are usually carried out using natural language, and the AI will respond using similar language.
Why would a search user choose generative AI over Google?
There are a number of ways in which generative AI like ChatGPT can be used instead of Google. Let’s say a user is researching a topic to help them make a purchase, for example they want to buy some new hiking shoes. The typical way they would do this is by inputting search terms into Google, and browsing through different links until they end up with a complete picture of the information they want. If they use ChatGPT, they can ask questions and get that same information presented more concisely and easily understood in much less time. They can then ask follow up questions, to better understand what it is they need.
They may then choose to do some research on search engines, but they equally might move to looking at brands on social media, reading reviews, and engaging with other types of content before they look to buy something.
Generative AI’s ability to summarize a topic and provide additional information and context immediately is something which gives it an edge over search engines. The user doesn’t have to read through multiple links and conduct multiple searches.
In what other ways can generative AI replace traditional search?
When it comes to ecommerce, there may be a number of search queries needed to find just one product. Especially if you’re unfamiliar with a product category, it could take you a while to find brands and products that are suitable. That’s why platforms like Shopify have started to develop AI powered shopping assistants like ShopAI. This takes the form of a chatbot that responds to prompts about what the user wants to buy. It can then ask follow up questions to get more specific, and allows users to set a budget. From there, the assistant offers suggestions for products and brands that may suit the needs of the user.
This is a step further than what search engines are able to offer. Typically, this kind of discovery journey would require several search queries. The user would then have to explore links themselves and find the best fit. Google might offer some links as part of the shopping snippet, or additional info on links through structured mark-up, but it’s not quite what the user will need. With a shopping assistant powered by generative AI, they can find a number of suitable suggestions with just a few prompts.
Is Google doing anything with generative AI?
Google in general has been incorporating AI into its algorithm and platform for many years. Language models like LAMDA and MUM are great examples of this. As for public use generative AI, they recently announced that they’d be adding a generative search component into their Search Labs platform. This is the first real glimpse we’ve had into what Google might do with generative AI. This would essentially provide users with more information about the topic they’re looking into.
What this does is help to reduce the number of queries it might take for a user to find an answer. After all, that’s kind of what ChatGPT is already able to do, but they can’t recommend products or sites the same way Google potentially could. It used to be that so long as Google could provide useful content in the form of links, users would stick around. Now it’s becoming how quickly they can provide all the potential useful content a user might need before they even enter a query.
As new technologies emerge, it’s inevitable that it will have an impact on how we engage with established technology like search. People want to find new applications for it, and discover new ways to make our lives online even easier than before. As generative AI continues to grow and become more advanced, we’ll start to see how it changes the shape of the search landscape.