Share this
Social Media Responds to ChatGPT
by Henry Chapman on Jan 20, 2023 12:00:00 PM
Social listening insights on the democratization of artificial intelligence
OpenAI, an artificial intelligence startup, commanded attention on a global scale in November 2022 when it released its newest chatbot, ChatGPT. From its very launch, ChatGPT has been open to the public for free use. ChatGPT uses OpenAI’s GPT 3.5 engine to generate text based on prompts supplied by users. Since the launch, users have used ChatGPT to generate all kinds of writing, from recipes to academic papers.
As attention for the platform grew, conversation about it blew up online. The post volume graph below not only shows the surge, but also the sustained growth of “ChatGPT” in online conversations following the launch.
Figure 1: ChatGPT’s post volume since its initial launch; Infegy Atlas data.
ChatGPT as a catalyst for change?
While there has been plenty of debate and concern over the ethics of using ChatGPT as a replacement for traditional writing, there are two other areas of conversation that have surfaced on social media: online commentators speculate that ChatGPT will disrupt both traditional search and software engineering.
In this brief, we explore the potential impact of ChatGPT on the fields of search and programming, as well as share both positive and negative changes that consumers anticipate and are discussing online.
Narratives around ChatGPT
We used Infegy Narratives to summarize the conversation around the OpenAI ChatGPT launch. Infegy Narratives categorize and cluster related areas of conversation. Amongst the clusters of conversation, we found that social media users talked about using ChatGPT as a potential search engine (light green cluster). We also saw that users repeatedly discussed using ChatGPT as a potential programming aid (orange cluster).
Diving further into these two areas of conversation, we find that users identify both positive and negative potential futures associated with ChatGPT.
Figure 2: Narratives around ChatGPT; Infegy Atlas data.
1. ChatGPT as a potential Google replacement
One of the core areas of discussion around ChatGPT was how it could supplement or even replace traditional search engines. Unlike search engines, which rely on keywords and pre-existing web pages, ChatGPT can generate new text based on its understanding of the user's intent. This means it can provide more specific and accurate information, as well as answer follow-up questions or provide additional context.
This word cloud shows conversations relating to Google Search and ChatGPT. Words like “threat,” “concerns,” “respond,” and “disrupt” feature prominently.
Figure 3: Word cloud showing negative terms around ChatGPT’s threat to traditional search engines; Infegy Atlas data.
ChatGPT can also make the search process more user-friendly since it responds to natural language. This could lead to a shift in the way we access information and potentially decrease the reliance on search engines as a primary source of information.
The significance of this shift could include a decrease in traffic and advertising revenue for traditional search engines and an increase in demand for natural language processing and machine learning technologies. It could also lead to a change in the way businesses optimize their online presence: there would be a greater emphasis on understanding and catering to natural language search queries, as opposed to Google Analytics and SEO-driven strategies.
All this being said, ChatGPT queries are more computationally expensive than a traditional search engine. It’s possible that concerns that it may replace traditional search anytime soon seem overblown.
2. ChatGPT as a programming aid
Infegy’s Narratives show that social media users discuss how ChatGPT can be used as a programming aid. ChatGPT provides natural language interfaces for coding and troubleshooting. This can make the programming process more accessible and intuitive for those who may not be familiar with traditional coding languages. For example, a user could give ChatGPT the prompt, "write a Python script to scrape data from a website," and it would generate the necessary code.
This word cloud shows a high density of programming-related topics. Many of these benefits were discussed quite positively.
Figure 4: Word cloud showing positive and negative terms around ChatGPT’s potential use as a programming aid; Infegy Atlas data.
ChatGPT, like any machine-learning model, is not perfect and can generate incorrect code or solutions. These models are not often fine-tuned for a specific task or domain. Therefore, it may not generalize well to new situations and may generate code that is not applicable. Furthermore, the model may not be aware of the best practices, conventions, or standards of the programming language and may generate code that is not readable, efficient, or maintainable.
Stack Overflow, a Q&A website for programmers, was so alarmed at the influx of ChatGPT-related posts that it banned all automatically-generated responses. The ban attracted attention and was discussed profusely online, causing the large post-volume surge shown in the graph below.
Figure 5: StackOverflow’s post volume graph around ChatGPT’s launch; Infegy Atlas data.
Conclusion
The release of OpenAI's ChatGPT chatbot in November 2022 has sparked significant online conversations about its potential uses.
Using social media analytics, we discovered that users think ChatGPT has the potential to provide more specific and accurate information and a more conversational user experience than traditional search engines. This could eventually lead to a decrease in traffic and revenue for the existing search industry.
Users also determined that, as a programming aid, ChatGPT can make coding more accessible and intuitive, but it is not without its limitations. It is important to thoroughly test and verify any code generated by ChatGPT before using it in a production environment.
Overall, the impact of ChatGPT on the search economy and the field of programming is likely to be significant and multifaceted, with both positive and negative consequences.
Share this
- October 2024 (4)
- September 2024 (2)
- August 2024 (2)
- July 2024 (2)
- June 2024 (2)
- May 2024 (2)
- April 2024 (2)
- March 2024 (2)
- February 2024 (2)
- January 2024 (2)
- December 2023 (3)
- November 2023 (4)
- October 2023 (4)
- September 2023 (4)
- August 2023 (4)
- July 2023 (4)
- June 2023 (3)
- May 2023 (5)
- April 2023 (3)
- March 2023 (6)
- February 2023 (3)
- January 2023 (4)
- December 2022 (2)
- November 2022 (3)
- October 2022 (4)
- September 2022 (2)
- August 2022 (3)