x Brown

AI Will Revolutionize Search Engines

AI Will Revolutionize Search Engines

SOURCE: TTI MarketEye blog

By Murray Slovick


AI is looking to significantly impact the future global search engine market as powerhouses Microsoft, using OpenAI ChatGPT, and Google, with LaMDA technology, compete to become the most dominant search service.

Unless you’ve spent the last year in a cave, and possibly even then, you know that tools such as ChatGPT, the AI–powered conversational interface, will dramatically change how developers write and consumers use software, especially in the upcoming battle to create the future dominant search engine.

ChatGPT, from OpenAI, is a free website where you can interact with the AI using a chat interface. ChatGPT’s dialogue format makes it possible for ChatGPT to answer questions, write high school-level English class essays and help individuals create the code for a new website. ChatGPT is based on GPT-3.5, a large language model released last year. It is a sibling model to InstructGPT, which is trained to follow an instruction in a prompt and provide a detailed response. ChatGPT has demonstrated that it is relatively good at finding mistakes or suggesting improvements to code. It can point code developers in the right direction. Users will be able to ask, “This code is not working like I expect — how do I fix it?” The chatbot uses machine learning to generate human-sounding responses to text prompts and can understand a variety of topics and contexts.

There are plenty of reasons to believe that AI like ChatGPT can take a big step forward this year. AI enables a search engine to have the potential to combine information across multiple websites and integrate it into a plain-spoken explanation. In the future, search engines will be required to answer questions in a more humanlike way rather than simply being a collection of related links. At the recent Consumer Electronics Show in Las Vegas, Panos Panay, Microsoft’s head of Windows and hardware, said AI was quite literally going to reinvent how you do everything on Windows.

Anyone who has used Google knows you have to sift through websites when looking for information. On the other hand, AI based search engines aim to offer quick and direct responses to specific problems, even if the questions are complex. More than just providing search results, upcoming search engines will be able to instantly offer such things as ingredient substitutions in recipes and add foot notes to search results on the search results page.

At a press event recently, Microsoft detailed some of its plans to bring OpenAI’s ChatGPT artificial intelligence-based chatbot to revamp its possibly once-again relevant search engine Bing so that it can handle more complicated queries. Its Azure Bot Service can provide an integrated development environment for bot building. Integration with Power Virtual Agents, a fully hosted low-code platform, enables developers of all technical abilities to build conversational AI bots—no code needed.

The AI that will be powering the Bing search engine uses technology that is faster and more advanced than what the public has experienced so far, according to Sam Altman, CEO of OpenAI. Microsoft’s goal is no less than to eventually dethrone Google as the No. 1 search service. ChatGPT relies on the powerful GPT-3.5 technology. GPT stands for Generative Pre-Trained Transformer, a complex neural network. Use it and you’ll get a well-constructed (though not always accurate) answer. I’ve added this parenthetical caveat as the large language models have well-documented tendencies to spew toxic content.

Google has responded by announcing its own plan to add AI to search. Google is working on a ChatGPT competitor named Bard, so called according to the company because it is a storyteller. Microsoft has optimized Bard for search. An AI chatbot was built using Google’s LaMDA technology, which is similar to OpenAI’s GPT series of AI language models that help power ChatGPT.

LaMDA’s conversational skills have been years in the making. it’s built on Transformer, a neural network architecture that Google Research invented and open-sourced in 2017. Transformer is a novel neural network architecture based on a self-attention mechanism that Google believes to be particularly well suited for language understanding. The architecture produces a model that can be trained to read many words, pay attention to how those words relate to one another and then predict what words will come next.

Google has had a stranglehold on the search engine business for as long as many of us can remember. By way of review, you may recall that Stanford graduates Jerry Yang and David Filo unleashed Yahoo in April 1994. Yahoo was initially a database of websites that was organized through a hierarchical system rather than a searchable index of pages. in 1998, two other Stanford PhD grad students, Larry Page and Sergey Brin, founded Google based on a ranking method of checking site links and this gave Google the edge over its competitors. Microsoft’s Bing descended from Windows Live Search and MSN Search in July 2009. Its first major update occurred in August 2011, and it was completely redesigned in 2015. In October 2020, “Bing” was rebranded to “Microsoft Bing,” complete with a new logo.

Microsoft’s aim cannot be overstated. Although Google still dominates the global search market, Microsoft has seen some incremental gains. Google has an estimated $149 billion search engine business and Google Search accounts for 58 percent of its revenue and about 61.5 percent of Google’s cash flow. According to Statcounter GlobalStats’s Search Engine Market Share Worldwide (January 2023), Google had a 92.9 percent share, Bing accounted for 3.03 percent and Yahoo!1.22 percent.

Many questions remain about how the chatbots will combat misinformation. Google has much more reputational “risk” in providing wrong information and thus is moving more cautiously than it normally would with regard to innovations that are challenging to their status quo. Because of the ubiquity of misinformation, none of the new entrants want, for example, to take on the question, “Was the 2020 presidential election stolen?”

Follow TTI, Inc. on LinkedIn for more news and market insights.

About The Author