Google issued a code red at the end of last year just because of the rise of ChatGPT. The code red is the highest level alert, which is issued when Google’s core business receives a very serious threat, and this is the first time that Google has issued a code red since its inception. For as long as I can remember, Google has been the world’s largest search engine service provider. Everyone is accustomed to using Google to search, and even calls the search action Googled.   “Don’t ask me, just ask Google!” This conversation often happens in my daily life. Every time we struggle with something we always use Google to find the answer. But in the age of artificial intelligence, one powerful AI chatbot has revolutionized our future. Can you imagine that maybe in the future, when you have any question you will use an AI chatbot to get the answer instead of Google? ChatGPT provides another way to search for information. The search engine retrieves content through keyword matching. It will retrieve a bunch of content that may be related to the information you want to know, but whether this information is what you want can only be judged by yourself, which means when you are Googling something, you have to click on every link to find the information that you want to use (Gao). But ChatGPT is different, you just ask it and it will tell you the answer immediately, no need to judge by yourself. I think it is possible for ChatGPT to replace Google because it is more humanized with high quality, it is almighty, there are no ads when you’re using ChatGPT, and its future development is unlimited.

To understand why Google issued the code red, we need to find out what ChatGPT can do first. ChatGPT is produced by OpenAI which is a company that specifically focuses on AI research and AI deployment. They want to use artificial general intelligence to benefit humanity (OpenAI). According to Katharina Burchholz, a data journalist, ChatGPT only took five days to reach one million users, for the same record, it took two years for Twitter, ten months for Facebook, and two and a half months for Instagram to get one million subscribers. From this we can see its influence. ChatGPT is actually very easy to understand, which is “conversation” and the operation is also very simple, it is just a chat room. You can use this chat room to ask questions and it will answer you. You can ask whatever you want and use any language that you are comfortable to use. Although this AI looks nothing special, its function is more powerful than you imagine. For example, ChatGPT is good at summarizing. We often see an article on the Internet that may be very long, and we may not find what we want after reading it for a long time. At this time, we can copy the entire article and paste on ChatGPT, it can help you summarize a short paragraph with the main points in a short time. It can also expand the outline, the only thing you need to do is tell ChatGPT your outline and set a work limitation, it can help you to come up with a well written essay. Two professional ethicists, Professor Gordijin and Professor Have, said that “One does only have to come up with a good prompt, which should even be feasible for someone suffering from writer’s block” which means that ChatGPT has the ability to write a great article if you give it some requirement, it can also help the people who have writer’s block.

First, ChatGPT is humanized. Continuing what I just said, ChatGPT can help you write an essay, some people began to doubt how high its level was. Christian Terwiesch, the economic professor of the Wharton University of Pennsylvania who is professional in MBA (Graduate Management Admission) teaching, used the MBA test question to test ChatGPT and grade its answer sheet to an uninformed marking teacher. The MBA test is used in admission to a graduate management program such as a Master of Business Administration program. At the end of this MBA test, ChatGPT got a B score, even though it is not a perfect A score, in this difficult test with a low pass rate it still can easily pass it. By this experiment, Professor Terwiesch found out that ChatGPT did a great job at analysis questions and the basic operations management, but ChatGPT made a few mistakes in some simple calculations at sixth-grade level math (Terwiesch). These simple mistakes let ChatGPT seem more like a human and the marking teacher didn’t notice that the test was not written by a human. This research shows it is an AI with highly education and also more humanized than Google.

Yesterday I asked ChatGPT about my favorite esport team T1 because I am really worried about them in their next competition, and I asked it because I was wondering what kind of answer it would give me and I have never asked it the question that is not about my homework before. The first question I asked was “who will be the champion in MSI, which is the competition’s name. ChatGPT gave me a very safe and boring answer, “As an AI language model, I do not have the ability to predict the future with certainty. The outcome of the MSI (Mid-Season Invitational) tournament will depend on various factors such as team performance, strategy, and luck…” (Graph 1) But it knew what game I was talking about even if I didn’t tell him the full name. Then I asked “Who is possible to win MSI” by asking this question, I expected that I would get some data analysis. ChatGPT gave me some possible team names such as RNG, MAD, and DAMWON KIA (Graph 1) but because the information ChatGPT receives is all before 2021 (OpenAI), the teams that it gave me are actually not fit for 2023. The last question I asked ChatGPT was “I am really worry about my team T1” I wanted to know how it evaluates my team or would it cheer me up. I was very satisfied with the answer that ChatGPT gave me (Graph 2). It encouraged me while also understanding what a fan would want to hear and the content was not as rigid as the AI we thought before. This personal experience also shows the humanization of ChatGPT.

Second, ChatGPT is almighty. You don’t have to use it for only homework or ask some professional questions, you can also ask him to write stories for you. One of the Chinese YouTuber, Gao, asked ChatGPT to write a science fiction novel about 500 words and the main character is his dog. ChatGPT turned out a novel that perfectly met the requirements Gao gave it with 500 words. This shows that you can use ChatGPT to write novels with any topic you want and you don’t have to pay any money on it. There is a Japanese who used ChatGPT to write a manga outline and used other AI to draw the manga, and the final product is just like a real manga artist’s product (Shi3z). As a result, ChatGPT can help you be a novelist and create novels that best suit your hobbies and this is what Google can’t do.

Third, ChatGPT doesn’t have ads. Sofia Pitt, a technology reporter in CNBC who has posted many articles that related to AI, asked ChatGPT and Google the same question that “How can I keep my Fiddle Leaf Fern plant alive?” and compared these two answers. Pitt thought the winner was ChatGPT because ChatGPT’s answer is the same as what she received from the plant company, but when she Googled the same question, the ads pop-up first and she had to go to various different websites. Here comes another advantage of ChatGPT, you don’t have to worry about the answer being impacted by any ads. 80% of Google’s revenue comes from advertisements, and the place where most of its advertisements appear is in the first few lines of the search results (Gao).

Fourth, ChatGPT’s future is full of possibilities. According to the New York Times, GPT-4, the latest version of ChatGPT, is here. Its abilities are stronger than the old one (Roose). In addition to its ability to get high grades in various tests, it can also help you with tax returns and the overall situation is more stable. GPT-4 can even analyze pictures. Even though ChatGPT may not be perfect right now, it is still improving and its abilities will only get better in the future.

In 2019, Microsoft invested one billion dollars in OpenAI, and some functions of OpenAI can only be used on Microsoft Products. Microsoft also intends to iterate OpenAI’s product into all their products such as Office series to make them all intelligent. In February, after Microsoft added ChatGPT to Bing, the number of downloads of Google Chrome was overtaken by Bing. The collaboration of Microsoft and OpenAI brings a very big crisis to Google. Google’s response is to announce that they will also release a conversational AI service, Bard,  in the coming weeks. However, in the demo video of Bard that was posted by the CEO of Google, Sundar Pichai, was found an error about the James Webb Space Telescope that is given by Bard. This counterattack doesn’t seem to be very effective. Google looks like it’s losing this AI war.

However, after I used it many times, I found ChatGPT’s fatal flaw, which is I cannot tell if ChatGPT is lying or not because I can’t find the reference about the information that was given by ChatGPT. This week I asked ChatGPT to help me with my political essay which is talking about the death penalty. I asked about some examples of the death penalty deter crime, and it gave me a Singapore example (Graph 3). But when I want to know more about the example that it gave me, there comes a problem, ChatGPT gave me one study published in the Journal of Contemporary Criminal Justice in 2010, when I asked it for the link for this study, it couldn’t give me. So ChatGPT said that I can try to search “The Deterrent Effect of the Death Penalty: Evidence from British Commuted Sentences in Singapore” to find the study, but I just could not find this source with the keywords. As a result, I couldn’t use this study in my essay because I didn’t know if it was credible or not. In this situation, we cannot make sure if ChatGPT is telling the truth or not because we can’t even find the source. What if it just made up a fake study to answer my question?

In conclusion, ChatGPT is like a highly educated and almighty human, its future is unlimited. Even so, Google still has a place in the future because ChatGPT’s inability to provide accurate sources is a very big problem for me. I think we also don’t need to treat AI as an enemy, it is just a useful tool. AI is just like a knife, a knife is a very powerful tool, but we are never afraid of knives. What we need to be scared of is the person with the knife. Hence, learning how to use AI in the right way will be a very important topic for us in the future.