TLDR: A study analyzing 3.8 million tweets about ChatGPT’s launch found that people’s reactions and timing of engagement are significantly influenced by their occupational skills and cultural backgrounds. Technical professionals tended to be early, positive adopters, while writing-centric roles engaged later with more skepticism. Individualistic cultures engaged earlier but showed more negative sentiment, while high uncertainty avoidance cultures expressed less positive views. The overall shift to more critical sentiment was due to new, skeptical voices joining the conversation, not a change of heart among early users.
The public launch of Artificial Intelligence (AI) products, like ChatGPT, often acts as a major event that captures collective attention and reveals how societies react to new technologies. A recent working paper, titled Winning and losing with Artificial Intelligence: What public discourse about ChatGPT tells us about how societies make sense of technological change, delves into this phenomenon by analyzing public discourse on social media.
Authored by Adrian Rauchfleisch, Joshua Philip Suarez, Nikka Marie Sales, and Andreas Jungherr, the research explores how people’s economic interests and cultural values influence their engagement with and perceptions of technological change, specifically focusing on ChatGPT.
Understanding the Public Conversation Around ChatGPT
The researchers analyzed a massive dataset of 3.8 million tweets posted by 1.6 million users across 117 countries following ChatGPT’s public launch in late 2022. They aimed to understand who engaged in the discourse, when they engaged, and their overall stance towards the technology. To achieve this, they used advanced AI models to classify the sentiment of tweets, identify the users’ countries of origin, and determine their occupational backgrounds.
The Role of Your Job: Skills and Stance
The study found a clear link between a person’s occupational skills and their attitude towards ChatGPT. Individuals in roles requiring higher levels of mathematics skills, often indicative of more technical professions, were more likely to express positive views and less likely to post negative messages. This suggests that those who develop or apply AI systems tend to see ChatGPT as a beneficial tool.
Conversely, people in occupations that heavily rely on writing skills were less likely to express positive sentiment. This group, which includes writers, authors, editors, and other creative professionals, often perceived ChatGPT as a potential threat to their livelihoods or artistic integrity. A notable example cited in the paper was a widely shared tweet about a user publishing a children’s book co-written and illustrated by AI, which triggered a wave of negative responses from creative fields concerned about ethics and the devaluation of human labor.
Regarding the timing of engagement, users with higher programming skills tended to join the ChatGPT discussion earlier. This is understandable, as these professionals are often early adopters eager to test new technologies, especially those offering coding support. In contrast, those with high writing skills generally engaged later in the conversation.
Cultural Context: How Where You Live Shapes Your View
Beyond individual occupations, the study also examined the influence of national cultural orientations, using Hofstede’s cultural dimensions. The findings revealed that users from more individualistic countries were more likely to engage earlier in the ChatGPT discourse. However, surprisingly, they were also more likely to express negative sentiment and less likely to express positive sentiment.
One possible explanation for this is that individuals in highly individualistic cultures might view ChatGPT as a threat to personal autonomy, creativity, or job security. In contrast, people from more collectivist cultures might focus more on the potential societal or communal benefits of the technology.
Furthermore, users from cultures characterized by high uncertainty avoidance were less likely to express positive sentiment towards ChatGPT. This suggests a cautious approach to new technologies in cultures that prefer stability and predictability, possibly due to concerns about ethical risks, job displacement, or the reliability of AI-generated content.
Interestingly, the study found no significant effect of ‘power distance’ (related to egalitarianism) on either attitudes or the timing of engagement.
Also Read:
- Consumer Trust and AI: Why Hiding AI Use Damages Brands More Than Exaggeration
- Unpacking Cognitive Biases in AI-Generated Content: A New Study Reveals How LLMs Influence Users
A Shifting Conversation, Not Changing Minds
A crucial insight from the research is that the overall shift towards a more critical stance on ChatGPT over time was not primarily due to early adopters changing their opinions. Instead, it was largely driven by the entry of more skeptical voices into the public discourse. This means that the aggregate sentiment trends observed on platforms like Twitter reflect changes in the composition of participants rather than a genuine shift in attitudes among those already engaged.
In essence, the study highlights that digital platforms serve as dynamic arenas where people negotiate the meaning of technological change, with their responses deeply rooted in their economic interests and cultural backgrounds. This ongoing public negotiation offers valuable insights into societal concerns and potential future societal divisions related to AI.
