CHATGPT WINS THE GAME OF CHATBOT BATTLE

Since both were introduced and became accessible to the public, ChatGPT and Perplexity have continued to compete closely to this day. Between the two, which one is superior? In this test, the same prompt was given to both of these popular AI giants. It’s important to note that the assessments given to each competing AI service in this article are purely subjective.

Technical Comparison

Below is a technical comparison of the two AIs discussed in this article.

Basic Version

Both technologies share some similarities, including premium or pro pricing, starting at $20 or around IDR 315,000 per month per user. Comparing their basic versions is challenging because the standard versions of both are almost identical.

Paid Version

For the paid versions, ChatGPT offers more pricing options with various tiers and additional features for businesses. The Team package provides access to ChatGPT's Advanced Data Analytics starting at $25 per user per month if paid annually. The Enterprise package is priced at $9,000 per month for 150 users, with stronger security and collaboration features suitable for large business needs. Perplexity AI, on the other hand, offers a higher-priced package called Perplexity Enterprise Pro, with a self-serve option at $40 per month or $400 annually per user. They also offer a custom Enterprise package, though its pricing is not publicly disclosed.

Information Accuracy

On this point, the author agrees with several AI reviewers who have previously chosen Perplexity as the more accurate AI. One key factor that gives ChatGPT a disadvantage is its limited ability to provide information that extends only to the last one to two years. In other words, OpenAI’s chatbot cannot provide the latest information. Meanwhile, its rival is able to offer very accurate, up-to-date information, accompanied by relevant, traceable sources.

User Interface

In terms of interface, ChatGPT appears more concise, simple, and spacious. On the other hand, its competitor displays a lot of content in a single frame, making the screen on a laptop or PC feel crowded. This is understandable, though, as Perplexity AI is a search engine integrated with a chatbot. The visual presentation resembles Yahoo’s layout from a few decades ago, which lost its competition with Google. Research has shown that simple and clean websites tend to be more comfortable to use psychologically, as evidenced by Google’s minimalist design that continues to attract users due to the lack of distracting elements around the main content.

Basic Features

ChatGPT clearly excels in terms of feature variety. Meanwhile, its competitor stands out in references, data sources, and follow-up question suggestions provided after responding to user prompts. ChatGPT allows its users to access a multitude of featured tools through Explore GPTs, such as humanize.ai for generating AI content that resembles human writing, Copywriter GPT for copywriting, CV writing for AI-assisted curriculum vitae writing, Scholar GPT for references and citations, Presentation Slides GPT for creating automatic slides, and many others.

Independent Testing

Here are some everyday usage scenarios where people might interact with chatbots. Keep in mind that the results of this independent testing are subjective, meaning the conclusions drawn are purely the author’s personal preferences.

Research Data Analysis Techniques

In this test, the same prompt was given to both AI chatbots.

"I am conducting research on part-time workers' experiences in two different industries: the creative industry and the technology industry. I have in-depth interview data with part-time workers from both industries. I want to understand the differences in work experiences between the two groups, especially regarding time management and income. What qualitative analysis technique should I use to compare these two groups? Is comparative thematic analysis suitable, or is there another method to explore differences between these groups?"

The results showed that Perplexity provided a more convincing answer because it included valid, traceable references. There were eight recommended sources, all of which were journals or academic articles. Meanwhile, ChatGPT offered only a narrative without clear sources. While the narrative provided was understandable and reasonable, it left additional work, namely, finding supporting references for the given statements.

Teaching Module Creations

In this second test, both chatbots were asked to create a teaching module. Here, it must be acknowledged that Perplexity clearly outperformed its rival because of two factors: first, the provision of valid, traceable references; second, Perplexity provided more up-to-date information than its competitor.

Creative Content Creations

In this scenario, ChatGPT performed significantly better than its competitor. OpenAI’s chatbot stood out for creative content creation, delivering responses that felt natural and communicative. Meanwhile, its competitor’s responses tended to be rigid. From this, it can be concluded that Perplexity is suitable for academic purposes, while ChatGPT excels in creative scenarios, such as business and marketing content, classroom activity ideas, and other creative needs.

Translations

Because of Explore GPTs, it can be concluded that OpenAI's chatbot clearly won in this test. In Explore GPTs, there are more than seven services specifically designed and developed for translation needs, particularly from English to Indonesian and/or vice versa. Meanwhile, its competitor still lacks accuracy in this area, with some translation lines that can be difficult to understand.

Article Paraphrasing

ChatGPT excels in article paraphrasing. As stated in previous points, ChatGPT is highly skilled in scenarios requiring creativity. ChatGPT offers an advantage in paraphrasing due to its deep contextual understanding and ability to generate word variations that match unique styles and tones. OpenAI’s chatbot, in several tests, was found to identify multiple meanings of words in different sentence contexts, so the resulting paraphrasing remains true to the original meaning but is presented with different diction and style. On the other hand, Perplexity also has the ability to change sentence structures and provide relevant information but is limited in adjusting language style and sentence complexity. Perplexity AI appears to focus more on providing information and is less optimized for generating unique, creative paraphrasing.