Trust in scientists took on heightened urgency during the Covid pandemic, as the public turned to experts for accurate guidance in a rapidly evolving health crisis. According to a study published in Nature, among 68 countries surveyed, India placed second in the measure of trust in scientists among people. However, pseudoscience, conspiracies and misinformation have also thrived in the country, alongside informed scientific views often undermining public health and education efforts and other policy decisions.
Nowadays, information on frontier technologies like artificial intelligence (AI) is often exaggerated and hyped, with commercial interests in mind. You might have scrolled through your social media feeds and seen politicians, celebrities, tech leaders, or even comedians sharing their views on AI. How can scientists make themselves heard against all this cacophony around the topic?
According to a new study by researchers from the University of Georgia, University of Utah and University of Illinois Urbana-Champaign, humour could be a valuable tool for scientists communicating online. They explored a simple question: If scientists on social media use humour when talking about AI, does it change how the public sees them and the information they share?
Their focus was on Twitter (now called X), a platform known for quick, bite-sized updates. The researchers wanted to know if different types of humour could influence people’s opinions. They defined three types of humour for this exercise: anthropomorphism, giving AI human-like traits; satire, a playful but critical take on AI; or a combination of both. Specifically, they wanted to know how likeable people find the scientist and how legitimate they believe the scientist’s tweets are as a source of scientific information.
In October 2020, the researchers surveyed 2,212 people across the United States. They selected the group to roughly match the U.S. population in terms of age, gender, and other demographic factors. Each participant was shown a screenshot of a made-up Twitter conversation as if Dr. Jamie Devon, a fictional AI scientist, had posted it. The focus of the tweet was a cartoon about AI, specifically self-driving cars.
The experiment had four versions of the tweet: a no-humour control group; Anthropomorphism, where the cartoon or text gave the AI human-like qualities, such as suggesting the self-driving car had feelings or quirks; Satire, with light-hearted mockery or jokes about AI; and a combination of anthropomorphism and satire.
In addition to these four humour conditions, the researchers also adjusted the numbers of retweets and likes to see if social media popularity, in the form of likes and shares, affected people's reactions to the scientist. While these numbers turned out to be less critical for the main research question, they helped the researchers control for the impact of popularity on the results.
After seeing one of the four Tweet versions, participants were required to answer questions about three factors: perceived likability to see if the scientist came across as warm, friendly, and likeable; perceived source legitimacy to see if the participants thought the Twitter conversation was a legitimate and appropriate source of information about AI; and mirth to see how funny they found the tweet.
Analysis of the answers showed that participants exposed to any of the three humorous tweets laughed (or at least smiled) more than those who saw the no-humour tweet. Out of these three humour types, the combination of anthropomorphism and satire sparked the most amusement. That laughter or sense of fun, in turn, made participants like the scientist more, and they also found the tweet to be a more credible source of AI information.
The researchers believe anthropomorphism, when we treat non-human objects or systems as if they have human traits, can help people relate to complex topics. For instance, if a scientist posts a cartoon of a self-driving car wanting a coffee break, it gives an otherwise complicated technology a lighthearted personality. Satire, meanwhile, uses wit or irony to draw attention to certain aspects of a topic, sometimes by exaggerating them. So, if a scientist jokes about an AI smartwatch that complains about working 24/7, it highlights real concerns about AI ethics and workload in a funny, digestible way.
Particularly noteworthy is how this humour led to higher likability and a greater perception of the source as legitimate. Some might assume that making jokes about AI could reduce how seriously we take a scientist. After all, we’re often taught to see experts as serious and objective. However, this study suggests that a bit of laughter doesn’t necessarily dismantle credibility. Instead, participants appeared more comfortable trusting the information because the humour bridged the gap between the scientist and their online audience.
In India, scientists increasingly use humour to communicate science and comment on societal issues. Nirmalya Kajuri (@kajunut) and Manu Awasthi (@mnwsth) are the more popular among Indian scientists on X, with 23.9K and 12.1K followers, respectively, and often blend lighthearted content with scientific insights. In another merger of science and humour, Shashi Thutupalli, a scientist at the National Centre for Biological Sciences (NCBS), partnered with comedian Biswa Kalyan Rath for a comedic yet informative podcast series. On the global front, Tel Aviv University's Oded Rechavi (@odedrechavi), with 141.5K followers, has been an early adopter of integrating memes and satire to depict academic life and its challenges. Despite their penchant for humour, these scientists continued to be viewed as credible voices in an otherwise loud social media environment cluttered with misinformation.
That said, the researchers caution that the persuasive power of humour could be a double-edged sword. While it enhances scientists' credibility, it could empower non-experts to spread false or misleading information. In a time when impersonators and bots are rampant online, audiences must exercise due diligence to verify the source of the information and the identity of the person sharing it. Moreover, humour isn’t universal. This subjectivity means that jokes that resonate with some might offend others, making it imperative for scientists to carefully navigate its use, especially when the consequences of controversial humour can be severe.
The study's results should, however, be taken with a bit of caution. They are based on a sample population from the USA, and the context for the humour is also American, which might not translate well in different regions of the world. Also, the study is only a correlational one and not causational. Although the research design tried to minimise these, other factors might also play a role.
As technologies leap forward, we’ll likely see more information in the news and social media, some true, most false. In a media landscape filled with clickbait headlines, heated debates, exaggerated claims and cautionary tales, sharing a joke can do more than just spark a chuckle. In this environment, scientists and experts who communicate engagingly may be crucial for guiding public understanding and encouraging constructive conversations about policy and ethics. Suppose humour can break the tension and make complex ideas more accessible. In that case, it can be a valuable tool to capture attention, spark curiosity, and foster trust while helping us learn more.