If you have been on the internet in the past four months, you couldn’t have missed the name ChatGPT. Designed by OpenAI and launched to the public in November 2022, ChatGPT is a language model chatbot that generates human-like answers. As soon as it became available to the public, conversations about ChatGPT spread like wildfire across traditional and social media.
As AI-driven technology gets integrated into our online lives, it is important to recognize the critical role that training data plays in its success. A model is only as good as the data it is trained on so it’s crucial to ensure that the data is diverse, representative, and of high quality. Data that contains bias or disinformation will result in AI-powered models that perpetuate discrimination, which is why high-tech companies should find a trusted data provider that works with structured datasets, free from unwanted types of speech.
ChatGPT’s sudden surge in popularity raised a lot of ethical debates, which in return gave rise to the following question: How did the perception of ChatGPT change since its launch in November?
To help answer that, we analysed over 200,000 English mentions of ChatGPT in the USA and UK markets between November 2022 – January 2023. We observed four interesting brand image shifts driven by the public’s reaction to ChatGPT throughout the monitored period.
1. Brand interest: volumes remained relatively stable up until developments in January
After an initial peak of mentions at the beginning of December, when netizens and online media shared their thoughts and experiences with ChatGPT, online interest in the chatbot remained relatively stable until January 2023, when we observed multiple peaks.
The top conversation drivers in January included news about New York City’s public schools banning access to ChatGPT over cheating concerns and Microsoft’s multibillion-dollar investment in OpenAI, only days after the company cut 10,000 human jobs.
2. Brand Sentiment: shifting between excitement and concern
Overall, 84% of the conversations involving ChatGPT were neutral as users shared news about the launching of the chatbot as well as Microsoft’s investment without including their personal opinions. Most of the positive mentions of ChatGPT came from netizens who excitedly recounted their experiences with it. Many called it “better than Google”, “the next big thing”, and one of the most impressive technological innovations.
Negativity arose after people voiced concerns about the chatbot passing some of the most difficult tests in the US education system as students used ChatGPT to cheat on their finals. Besides potentially harming academic integrity, ChatGPT was seen as a threat to the IT job sector, a cyber security risk, and a perpetrator of biased and discriminatory narratives. Moreover, its parent company was criticised for outsourcing work to underpaid Kenyan workers who worked to improve a tool meant to detect toxic content on ChatGPT.
3. Media Type Breakdown: Twitter as the main source of volume
Unsurprisingly, Twitter was the main volume source with 62% of all mentions. Any brand or product launch that attracts conflicting points of view gets frequently discussed on Twitter. Users actively shared news about ChatGPT: from its launch and the various ways it was used, to Microsoft’s major investments in the developers of the app, OpenAI.
While some only retweeted the content, others talked about their interactions with the bot. The most popular discussions revolved around ChatGPT’s ability to:
- pass extremely difficult university and licensing exams,
- write functional code with user prompts.
Read more about conversations on other media types, including News, Forums, Tumblr, Blogs, Reviews, etc.
4. Brand impact: ChatGPT’s potential to transform 3 sectors
Online conversations about ChatGPT reflected the potential brand impact of the chatbot on three major sectors:
Information Technology (IT):
Many users found the chatbot’s ability to successfully write codes within seconds impressive, despite sometimes delivering inconsistent results. While some people saw ChatGPT as a positive shift towards time efficiency, others voiced concerns about the tool replacing human jobs and being used to develop malware.
While some users thought ChatGPT could help students improve their creativity, writing, and learning skills, many saw it as a threat to academic integrity and an aid to plagiarism. The AI chatbot raised serious concerns when it passed some of the most challenging American professional exams, such as the US medical licensing exam, Wharton MBA and the bar exam.
Some students admitted to having already used ChatGPT during final exams, prompting some school districts and universities in the US, Australia, Ireland and other countries to ban or restrict the use of ChatGPT.
Media and Journalism
Many users discussed ChatGPT’s ability to create complex and detailed forms of written content (articles, reports, opinion pieces, poems, scripts and whole stories). This development was seen as a threat to journalists, writers, content creators and marketers because its use could automate content production, causing a reduction in the need for human workers and a decline in content quality.
Many feared that it could also be used to create fake news stories and generate spam human-like comments and social media posts that support a harmful agenda. However, some saw ChatGPT’s potential to benefit the media and journalism industry by saving hours of work.
An innovation or a threat?
Conversations recorded and analysed throughout the monitored period showed that the initial excitement for ChatGPT’s automation ability was soon coupled with concerns about its possible misuse and potential to replace human jobs. While it’s still too early to conclude whether ChatGPT will do more harm than good, it is evident that as a brand, ChatGPT is receiving mixed reactions.
Brands need to monitor media and respond on time to threats to their brand image. One way that OpenAI is trying to counter concerns in the education sector is by promising to create a solution that will assist teachers in recognising work written by AI.
Are you interested in media monitoring or training datasets for machine learning?
We do both! Our bespoke media analysis reports will help you understand how the public perceives your brand so you can make informed decisions when faced with a potential reputational crisis. Download our report to learn more and contact us for more information!