Many societies are facing the increasing threat of social upheaval's arising from factors such as political divisions, fallout from economic downturns, the covid‑19 pandemic, and environmental challenges. No obvious path to create more social stability exists. An enabling factor in social upheaval is open exchange on social-media platforms. In addition, marketing and media organizations thrive on division among consumers, promoting conflict to monetize differing identities, opinions, and lifestyles. Divisive targeting and segmentation on social media employ sophisticated personalized algorithms to create the illusion that one's preferences and opinions are simultaneously the same as those of some people ("us") but fundamentally different from and in harsh conflict with those of other people ("them"). This familiar us‑versus-them mentality, although profitable, appears to tear at the very fabric of society because it makes trust and conflict resolution seem elusive, which gives rise to a high frequency of disruptive events.

Too often, organizations' trust-building efforts are expensive and meaningless exercises in reputation management.

In 2020, journalists Jeff Horwitz and Deepa Seetharaman reported on internal research that social-media giant Facebook (Menlo Park, California) commissioned in 2018 to examine polarization among users. Crucially, Facebook executives, aware that the company's algorithms created division among users to garner attention and increase the amount of time users spend on the company's social-media platform, decided not to act on the research or to lessen the promotion of conflict. One conclusion among Facebook executives was that US citizens, in particular, had been drifting apart naturally since the 1960s. Late political scientist Samuel Huntington believed that social upheaval follows a normal curve of development in US society and that, historically, the United States has a "nervous breakdown" about every 60 years. Such breakdowns occurred in the United States in the 1770s (the US struggle for independence), in the 1830s (mass democracy challenges to the "aristocratic republic"), and in the late nineteenth century (industrialization and urbanization). According to this point of view, because the country witnessed its last breakdown in the 1960s, it is due for another one. Indeed, a recent article in the Week discusses Dr. Huntington's work in the context of the unprecedented invasion of the US Capitol Building in Washington, DC, on 6 January 2021.

Explaining much of the current social upheaval simply in terms of natural cycles is tempting. But a new research paper by an international team comprising researchers from the University of Washington (Seattle, Washington), the Delft University of Technology (Delft, Netherlands), and several other institutions suggests that social media itself could be a grave threat to humanity and that insufficient information exists to determine whether it is. The paper calls for the creation of an urgent, large-scale, interdisciplinary scientific collaboration to gather the data necessary to reach a conclusion about this matter so that policy makers have an opportunity to avoid a disaster driven by the mechanisms that Facebook observed in 2018. Similarly, the US National Science Foundation (NSF; Alexandria, Virginia) recently awarded the Prosocial Behaviors in the Digital Age project at Cornell University (Ithaca, New York) a multiyear $1.19 million grant to study how norms that deter negative behavior online emerge and how people maintain them. Project colead Drew Margolin, an associate professor in the university's Department of Communication, explains that "there's this intuition, 'If you see something, say something.' We all have that intuition.... But in social media, who are you saying it to? Do they share your view? Are they actually on the other side? Do you look like the bad guy? All these things are really hard to understand" ("NSF funds work on flagging bad online behavior," Cornell Chronicle, 8 July 2021; online). That the major social-media platforms are—at least in the short term—either unwilling or unable to regulate toxic discourse has become increasingly clear, and finding more organic ways to preserve civility online might be a fruitful effort.

The proliferation of human-behavior data—including data that could aid in the development of methods for detecting destructive trends and social uprisings on social media and other platforms (such as smartphones)—presents special challenges for researchers and policy makers. A recent editorial in the journal Nature discusses the proliferation of various forms of social data, including location data, device‑use and software‑use data, and spending data. These diverse forms of data are providing researchers with new tools for analyzing and predicting human behavior. Although such data might drive advances in social, political, and economic research, they may also introduce new ethical issues and privacy risks. In particular, the Nature editorial notes that managing and curating data are key challenges and that language barriers can further limit collaborative research. In addition, cultural diversity can make data sets more complex and more difficult to interpret with artificial intelligence and various analytics tools. Researchers are exploring various solutions to improve research quality, identify potential errors, and develop complementary research approaches.

Along these lines, Santa Fe Institute (Santa Fe, New Mexico) researchers and their colleagues from other institutions devised a new way to design surveys to address challenges in acquiring accurate data about social trends and sentiment. The new methodology may improve trend analysis in politics, consumer behavior, and other areas. In the study, the researchers found that asking participants who they thought their friends were going to vote for provided more accurate data than did asking participants who they were themselves going to vote for. The researchers anticipate that combining individuals' perceptions of friends and associates with new computational models for human social dynamics will enable better predictions about emerging trends and developments—particularly developments in politics and areas relating to health. Emerging techniques for accurate surveying may affect a range of areas beyond politics. In particular, new methods may drive advances in consumer-sentiment analysis and AI systems for product design and marketing. Imaginably, such techniques could apply to technology-foresight practices and improve the analysis of developing technology trends. In addition, such techniques might detect changes in public sentiment about disruptive technologies and related regulations that could inform various commercialization efforts.

Ultimately, most modern government, nongovernment, and business entities suffer fundamentally from a lack of trust among many stakeholders. Although many citizens once expressed relatively high levels of trust in most official organizations, various divisions and messaging on social media have largely eroded that trust. Too often, organizations' trust-building efforts are expensive and meaningless exercises in reputation management, suggesting the urgent need for organizations to understand, predict, and control trust-eroding trends and sentiments and to focus on achieving cooperation and harmony among a variety of stakeholders.