At a glance.
- Disinformation, adversarial design, and the weaponization of uncertainty.
- US undertakes an investigation of COVID-19's origins.
- Colombia's Defense Minister accuses Russia of fomenting unrest.
- Chatbots as a persuasive tool?
Facebook's threat report cites Russia as a leading source of disinformation.
Facebook on Wednesday published its strategic Threat Report: Influence Operations 2017-2020. While it's true that, as the Washington Post summarizes, Facebook calls out Russia as the leading source of disinformation, the report contains a particularly useful account of trends as disinformation has developed over the last few years. The researchers cite three overarching trends:
- "IO moving into grayer spaces. The scaled techniques we saw in 2016 are now harder to pull off, more expensive, and less likely to succeed. As threat actors evade enforcement by co-opting witting and unwitting people to blur the lines between authentic domestic discourse and manipulation, it will get harder to discern what is and isn’t part of a deceptive influence campaign. Going forward, as more domestic campaigns push the boundaries of enforcement across platforms, we should expect policy calls to get harder."
- "Increased actor diversity While state-run operations will continue to persist, it’s important to remain open and flexible in our detection and response efforts because we know that not all IO is state-sponsored. More domestic, non-state, commercial actors are using the same tactics to influence public debate in their strategic interests. This will lead to more challenging attribution, with more layers of obfuscation between the operators and the ultimate benefactor. Additionally, the spectrum of inauthentic behaviors is wide. They aren’t always focused on politically-motivated campaigns. Many engage in lower-sophistication, higher volume, financially-motivated campaigns, like the clickbait content farms from Macedonia that leveraged real people’s content about protests in the US to sell merchandise or drive people to ad farms."
- “'Weaponization' of uncertainty. We expect to see IO actors continue to attempt to weaponize moments of uncertainty, elevate conflicting voices and drive division around the world, including around major crises like the COVID-19 pandemic, critical elections, and civic protests. Historically though, when we investigate IO targeting these defining moments, authentic voices typically outweigh inauthentic attempts to manipulate public debate."
Facebook argues that disinformation flourishes in the absence of information, which might be taken as a classically liberal endorsement of the efficacy of a marketplace of ideas. The social network also makes a strong case for what it calls "adversarial design," defined as "the process of thinking like a threat actor in circumventing our own defenses so we can make our policies, scaled detection, and products more resilient to manipulation techniques, in addition to finding ways to provide people with more context about what information they see on our platform." Much of that adversarial design seems to be realized in techniques to increase transparency and reduce the scope of possible inauthentic manipulation of opinion.
US Intelligence Community to investigate COVID-19 origins.
US President Biden has directed the Intelligence Community to investigate the possibility that the COVID-19 pandemic could have originated in a Wuhan, China, biological research facility. CNBC quotes the President as saying, "As of today, the U.S. Intelligence Community has ‘coalesced around two likely scenarios’ but has not reached a definitive conclusion on this question. Here is their current position: ’while two elements in the IC leans toward the [human contact] scenario and one leans more toward the [lab leak scenario] – each with low or moderate confidence – the majority of elements do not believe there is sufficient information to assess one to be more likely than the other.” So he would like some answers.
The Washington Post has a timeline of how what was once widely dismissed as a crackpot conspiracy theory has quickly gained traction in what one might as well call establishment circles. The Post complains that President Trump's statements about the pandemic were too often mixed with harsh words for China, and that this led people to dismiss the possibility that there was anything to the notion that the Wuhan Institute of Virology (WIV) had been the initial source of infection. Among the publications to dismiss some version of the WIV origin theory as claptrap were essays in the Guardian, the New Republic, and Scientific American, and they were far from alone in doing so.
It's important to distinguish at least four versions of the hypothesis that COVID-19 had its origins in WIV. From most to least plausible, they are:
- WIV was studying the virus, and it accidentally escaped.
- WIV had engineered the virus, and it accidentally escaped.
- WIV was studying the virus, and deliberately released it.
- WIV had engineered the virus, and deliberately released it.
The first question, of course, is whether the virus emerged from the lab, or whether it just happened to make the jump to the human population in Wuhan. So we shall have the opportunity to see whether a US investigation succeeds in bringing some clarity to the origins of COVID-19. Beijing has throughout the pandemic been less transparent than critics would have wished.
Colombian Defense Minister accuses Russia of fomenting unrest.
Colombia's Defense Minister, Diego Molano, is reported to have said that Russian cyberattacks, presumably disinformation campaigns disseminated in social media campaigns, were responsible for fomenting unrest in Colombia. The Russian embassy called the accusations unfounded and unsupported. Bogota City Paper has a summary of troll accounts active in inciting disturbances. Yahoo reports that Russia's Foreign Ministry has summoned the Colombian ambassador for explanations of the Defense Minister's remarks.
Chatbot persuasion.
The New York Times last week ran an op-ed on persuasion, specifically a tutorial on how to convince reluctant people that they should receive vaccination against COVID-19. The authors are an academic pediatrician and a psychiatrist who leads a political group designed to persuade the unconvinced to embrace progressive doctrine. Their fundamental approach is expressed with a simple chatbot that responds to vaccine skepticism with a series of reassuring questions. It's an interesting concept, but while chatbots can be programmed to dispense information, they provide only a simulacrum of that reassuring empathy and understanding. Thus there would be a fundamental dishonesty involved in actually using the chatbot, but the point is to let it teach you, the fully persuaded persuader, a style of communication the authors believe more effective than either dialectic or denunciation.
The techniques recommended—displaying empathy, and giving your interlocutors a sense of ownership of the knowledge being presented—are of course neutral with respect to truth or falsehood. That's the case of any persuasive technique: it's a way of inducing conviction, not a way of discovering truth. (Indeed, the op-ed itself needed a correction the day it was published. Its original version said that COVID-19 had killed more people than died in both World Wars, which is of course absurd by about two orders of magnitude. Zeal for persuasion usually does outrun commitment to truth, which is a potent contributor to the spread of misinformation.)