A big battle is going on in the technology world – The AI battle between tech giants Microsoft and Google.
In recent days, tech titans such as Google and Microsoft have engaged in a technological arms race for the future of search. Both companies are striving to revolutionize the search engine market of trillion dollars in value by integrating the AI chat-bots in their services.
However, Microsoft was one step ahead, by officially launching the new Bing search engine, with ChatGPT integration. The appearance of ChatGPT has made Microsoft’s Bing suddenly appreciated again, while Google is said to be falling into the FOMO (fear of missing out) syndrome. Now, let’s dive deep into the article to understand the differences between Google vs ChatGPT – which one is better?
The “strange” position of Google
Launched in May 2009, Bing was considered as an important rival to Google Search, but then fell far behind. According to SimilarWeb statistics over the last twelve months, Google has accounted for 91% of the global search market, while Bing has accounted for 3%. The figures are 87% and 9%, respectively, according to Statista.
The search engine battle is heating up with the emergence of ChatGPT. This chatbot was developed by the startup OpenAI, invested in by Microsoft with billions of dollars and quickly reached 100 million users just within two months. The Business Insider says that Google is in a “strange position” on the online search front: it dominates the market, but is also being surpassed by a start-up.
Despite the fact that ChatGPT has some weaknesses, it is viewed as having the potential to overthrow Google. As a result, Sam Altman, the CEO of OpenAI, speaks out: “It is impossible for a technology to totally eliminate a major firm.” He believes that large companies, like Google, will be able to develop “countermeasures” to fight back against ChatGPT.
So, can Bing combined with ChatGPT defeat the dominance of Google Search? To find out Google vs ChatGPT – which tool is smarter, Language firm Preply analyzed the performance of these two tools by asking them the same 40 questions. They include various subjects from technology, health, legal, politics, finance, entertainment, culture to business and so on. While some are frequently asked questions, others will require a certain level of intelligence. The Preply team wanted to challenge search tools, to provide quick, clear answers, with appropriate context and most importantly, useful to the users. And, here is what they found.
Google vs ChatGPT: Which one’s better?
At the basic level with 12 questions asked, Google has 7 superior answers, whereas ChatGPT has 4 superior answers, and one answer is equal.
When asked “How to tie a tie”, Google displayed an instructional video, while ChatGPT provided a lengthy text-based instruction that was not as visually appealing as the video. On the other hand, when asked “Is Santa Claus real?” ChatGPT firmly stated that Santa Claus is not real, while Google said that Santa Claus is real.
In the intermediate level, ChatGPT provides 15 better answers than Google. In the seven more challenging questions, ChatGPT gets 4 better answers than Google.
In total, with 40 questions, ChatGPT has 23 better answers, Google has 16 better answers, and one answer is a tie. It can be seen that, with the questions that are more difficult, ChatGPT gives better answers and has greater processing ability. Google, on the other hand, provides more intuitive and understandable results with questions that are simpler than those that are more difficult.
Google got into the FOMO mentality and launches AI product Bard to compete with ChatGPT
According to Jana Eggers, CEO of the AI firm Nara Logics, Google is being drawn in by the fear of missing out (FOMO). Google has been investigating AI for many years and has contributed to the development of several of the technological platforms used by ChatGPT. However, the user excitement for ChatGPT has frightened investors, and Google is also rushing to release its own AI chatbot.
On February 6, 2023, Google announced an AI chatbot technology called Bard and it will be launching in the next few weeks. Bard will directly compete with ChatGPT.
OpenAI is just a startup company, ready for the “nothing to lose” battle. Meanwhile, Google will face more risks in terms of reputation. According to experts, if Google is not careful in launching the product, they may be subjected to a series of investigations into privacy violations, copyright, and discrimination… the hidden potential in the answers of the chatbot.
“It is unknown how Google can avoid issues with Bard. Perhaps the company is accepting a higher degree of risk compared to its previous approach,” said Max Kreminski, an AI researcher at Santa Clara University.
According to Max Kreminski, users are accustomed to obtaining information through Google over the years. Therefore, if they seek advice from Bard, they might not notice that the AI might provide false information. “The story could end up in a hospital or court,” he says.
In agreement with the above idea, Emily Bender, a professor of computer linguistics at the University of Washington believes that putting a “friendly, fluid” AI interface on top of search engines does not correctly solve the problem. Instead, large technology companies should encourage people to use AI products safely and effectively.
ChatGPT vs Bard: What’s the difference?
As AI tools, ChatGPT and Bard offer similar services. They are designed to explain complex subjects and provide human-like responses to user queries.
However, ChatGPT and Bard are distinguished by their data sources. Because Bard pulls data from the internet, it is always up-to-date. Also, it has the ability to gather more information in real time, since it can access the most current research. On the other hand, ChatGPT’s data comes from research published after 2021, so it can only handle more current information. Moreover, because ChatGPT generates content in a single text prompt, it provides content in smaller chunks than Bard.
Bard is based on LaMDA, short for Language Model for Dialogue Applications. It is developed to allow software to better participate in a fluid and natural conversation. LaMDA is built on the Transformer architecture, similar to other language models such as BERT and GPT-3. However, due to its training, LaMDA can understand questions and dialogues with different shades of meaning on various topics.
OpenAI’s GPT, or Generative Pre-training Transformer, was first released in 2020 and the GPT-3.5 language model series completed its training process in early 2022, forming the backbone of ChatGPT. “ChatGPT sometimes writes answers that sound plausible but are inaccurate or meaningless,” OpenAI mentioned in a blog post.
However, Bard is still in the experimental phase and not available to the general public. Therefore, the result of the battle is still pending.