Is chatbot AI a competitor to Google search? | AI in business #40

How is chatbot AI changing how we browse?

AI chatbots, such as ChatGPT created by OpenAI and its close relatives from Microsoft – Bing, Google Bard, or Claude from Anthropic, available in some countries, are changing the approach to information search. Instead of typing keywords and browsing hundreds of links, we can simply ask a question.

For example, we can replace the keyword, “energy saving,” typed into Google’s search engine with a question and AI chat, “What are the best ways to lower your electricity bill?” and immediately get a specific answer generated by AI.

AI chatbots are not only able to acquire the knowledge they have learned in the blink of an eye but also:

• search for information on the Internet – that is, perform the task of a search engine,
• compare them with each other – for example, to confirm their reliability,
• summarize the most important facts – so that we can, for example, based on the recap, choose the article whose content interests us most.

Most notably, AI chatbots can have an interactive conversation, so you can refine the results by asking more questions and receiving personalized advice and ideas.

Feature comparison: chatbot AI vs. Google Search

Comparing specific chatbots with Google Search reveals both common features and significant differences between them.

The most popular ChatGPT in the free version and Claude can generate text answers to natural language questions and conduct simple conversation. However, they do not index websites or give direct links like Google.

Search issues are handled differently by Bing and Bard, two AI chatbots that have access to the Internet.

In Bing, you simply type in any query, from which the search engine “fishes out” words and key phrases and searches the Internet for the information you need. Based on this, it generates a text response with links to sources of information.

Source: Bing (https://www.bing.com/)

Google Bard first generates an answer with the knowledge it has. Only after clicking on the “G” button, Google it, placed under the AI chat statement does the search begin browsing. In some situations, Bard highlights in green the information in the answer that it managed to confirm during the search. In others, it prompts for search phrases, the clicking of which redirects to the Google search engine.

Google’s search engine itself, on the other hand, offers advanced searches for specific sites and online resources with options for filtering, location, etc. It also displays information in the form of graphics (Knowledge Graph) and allows searching for images and videos. While the way information is displayed is changing dynamically and Google Search is increasingly displaying answers to questions without the need to open links, it does not generate full-text answers like chatbots do.

Source: Google (www.google.pl)

Speed and availability of information: AI chatbots vs. Google

The main advantage of chatbots is instant access to information. This is because they have the potential to deliver information faster than Google Search. Just ask them a question to get an answer right away, without waiting for a page to load or browsing through multiple links.

Of course, Google’s knowledge base is much broader, as it indexes billions of web pages. However, chatbots have increasingly large databases and references, and their answers can be fully sufficient for many popular user queries. Minimal effort in reaching personalized information is their big advantage.

Personalization of search results by chatbot AI. Opportunities and threats

AI chatbots can analyze previous conversations and match results to a specific user. On the one hand, it’s a good chance:

• displaying answers to questions that fit in with even the questioner not fully spoken intention,
• Personalization of the form of the response, such as its length or tone, as well as its content – if I ask about restaurants in the area, and the AI chatbot knows I’m vegan, it will suggest restaurants offering vegan options,
• friendly form of interaction often laced with humor, as well as
• The ability to continue even a long thread on a single topic and easily return to the ongoing conversation.

Unfortunately, the personalization of search results by AI chatbots also poses risks:

• selective content filtering – for example, displaying news not only based on an analysis of our preferences, but also its alignment with a country’s policies,
• The risk of creating “information bubbles” – that is, the chatbot tells us what we already know, what it has told our friends, or what it thought we would like to hear,
• promoting selected products or content – based on combining our preferences with the advertiser’s target audience,
• limiting a variety of sources of information – that is, suggesting what we have already paid attention to.

For example, a chatbot advertising a vacation may show only offers from specific travel agencies, leaving out others. Such personalization can also make certain content completely disappear from the view of our business. For example, if we regularly visit our own company’s website, a chatbot may hide from us… the existence of competitors. Therefore, the transparency of their operation is an important topic.

How do chatbot AI handle complex and ambiguous queries?

AI chatbots sometimes have trouble answering more complex and ambiguous questions that require understanding the relationship between different concepts. They often fail to understand context or nuance, so their answers are sometimes incorrect or incomplete. Therefore, the more precise the question, the better the chatbot’s answer. However, they are getting better at it. For example, after typing a word with a typo, Bing asks an additional question:

• Question, “What is the aquarelle technique?”
• Bing’s answer: “Did you have a watercolor technique in mind?”

At the same time, if a question consists of several interconnected threads, Bing searches for them as separate phrases and then consolidates them into a textual answer.

The future of search: will AI chatbots dominate?

AI chatbots perform much better than search engines on issues that require subjective judgment, advice, or prediction of consequences.

Their development in this direction may bring the greatest benefits. Most likely, however, in the near future, AI chatbots will be a complement rather than a replacement for traditional search engines. We have included their comparison in the table below:

Function	Chatbot AI	Google Search
Understanding natural language	Yes	Yes, but to a limited extent
Personalized search results	Yes	Yes, but to a limited extent
Ability to refine the results obtained	Yes	Not
Search speed	Depends on the chatbot chosen, a few to several seconds	Quick, but sometimes it can take a few seconds
Speed of getting an accurate answer to a question	As soon as the search results appear	After independently finding the right answer in the results

The strengths of AI chatbots are speed, convenience and the ability to generate responses in natural language – written or spoken. But search engines will still be needed when looking for specific sites or online resources.

Key to the development of chatbots will be the reduction of hallucination, that is, generating responses that are plausible but false. Issues of transparency and control over their operation are also key.

If you like our content, join our busy bees community on Facebook, Twitter, LinkedIn, Instagram, YouTube, Pinterest, TikTok.