DUE to the vast amount of data on the Internet, artificial intelligence (AI) has developed rapidly in recent years. In this AI era, what kind of search engines do we need to handle such massive data?
Let’s take a step back and look at the development of the Internet and search engines over the past thirty years.
Since the commercialisation of the Internet in 1994, the past 30 years have fostered four indispensable elements in our lives: ecommerce, search engines, social media, and the mobile Internet. Among these, search engines have become an essential tool in our digital lives amidst the vast sea of information.
Before Google was founded in 1998, if you are an older reader, you must have used search engines like Yahoo!, Infoseek, AltaVista, Excite, and others. It was an era of fierce competition.
After Google was established, it quickly attracted a large number of users with its superior search results and created the search advertising business model. Validated by the market, Google almost monopolised this crucial Internet business, rising to become one of today’s tech giants.
Over two decades have elapsed, and Google has consistently maintained its competitive edge. It has continuously enhanced its search engine capabilities to provide users with a better search experience.
Just as Google seemed poised to continue reaping the benefits of this market, the tech industry experienced rapid changes. After OpenAI launched ChatGPT on Nov 30, 2022, not only did it bring changes to many fields, but it also offered new possibilities for the search engine market.
Firstly, Microsoft, an investor in OpenAI, integrated ChatGPT technology into its search engine, Bing, providing users with a new search experience. Users no longer need to sift through a multitude of links to find the needed answers. Bing, or later Copilot, uses advanced AI technology to generate answers for users and provide relevant links, allowing users to complete their searches more quickly.
Additionally, this has spurred a new generation of AI-driven or Large Language Model (LLM)-driven search engines, such as Perplexity, rekindling interest in the search engine market that had been “quiet” for many years.
Based on the developments over the past year and a half, here are my five observations:
Bing/Copilot: Since integrating ChatGPT technology, Microsoft has been striving to carve out a new path in the search engine market amidst the AI wave. Personally, I hadn’t used Microsoft’s search engine for many years, but over the past year, I have started using it in various situations and enjoyed the benefits of answer-centric search results. From Bing to Copilot, Microsoft aims to steer the next generation of search engines towards “AI-driven personal assistants” (Microsoft called it “Your everyday AI companion”.) This seems like a smart business strategy, as it can create new value and services in the AI era rather than continuing to compete with Google in the search engine market. However, whether the industry players will develop new business models beyond advertising and subscription services in the future remains to be seen. Perplexity: Besides the big tech companies vying for the AI-era Q&A-based search engine market, new startups like Perplexity have also garnered much attention. Perplexity’s advantage lies in its startup nature, unencumbered by existing business models that might limit their creativity in developing new products and services. Its pro version even allows users to switch between different large language models like GPT-4 and Claude, which is very convenient. Notably, one of Perplexity’s major investors is Jeff Bezos, the founder of Amazon, making future collaborations between Perplexity and Amazon’s products and data worth watching. Google Search: Despite the new developments and new players in the AI-driven search engine market, Google’s search engine market share hasn’t been significantly impacted over the past year. Additionally, Google is vigorously promoting its large language model and new service, Gemini. As long as Google maintains innovation and quality in its product iterations, the impact of this wave on its leading position in the “traditional” search engine market might not be substantial. Moreover, Google should consider leveraging its market advantage to develop more creative AI products, positioning itself well in the AI era. Meta AI: Meta’s advantage lies in its social media content, which traditional search engines have struggled to index, such as on Facebook and Instagram. Meta’s launch of Meta AI in April 2024, based on its LLM Meta Llama 3, serves as Meta’s AI assistant and allows users to search Meta’s social media platforms. This not only brings more attention to Meta’s open-source LLM but also potentially opens a new market – social media search advertising. Ability to ask the right question: We’ve been trained from a young age to be good at finding answers. However, no era has emphasised the importance of the ability to ask the right questions as much as now (for example, the prompt you send to ChatGPT). Sometimes, the ability to ask good questions is even more important than the ability to find answers. In this era, if we possess the ability to ask good questions, we will find better answers more quickly – an ultimate combination of efficiency and effectiveness.
Today, we stand at the dawn of the AI era, where all established paradigms will be reshuffled. Whether it’s tech giants, startups, or ordinary users like us, we must embrace this new era, adapt to these changes, and master the ability to ask the right questions at the right time to find the most valuable answers for us.