CIO of FunCorp, a mobile entertainment app development company, oversees the growth of UGC platforms with AI/ML content feed aggregation.
We are seeing more and more criticism surrounding traditional social networks, especially from those who have profiles on Facebook. Specifically, this criticism is regarding the lack of personal data security and the unethical use of data in advertising and marketing.
According to a recent study by Acquia, 65% of respondents would stop using a brand that was dishonest about how it was using their data. For example, huge corporations often use it to create ideal consumer goods, which is not always a good thing, especially when it comes to products and services that are not healthy.
We have been living in such a world for more than 10 years. But in the last two or three years, there has been a growth in anonymous content products, allowing communication with a wide array of users. These are not yet products for the masses, like current social networks, but they are rapidly gaining popularity. Some of them you've used; some you've heard about in the media, and others you've never heard of — but that doesn't mean they don't exist.
In the U.S. and other tier-one countries, some of the leading players include Reddit, iFunny (Full disclosure: iFunny is a FunCorp brand.), Telegram and Discord. Their total audience may very well compete with the most significant social networks. So, why do some people know very little about them?
Currently, these products are attracting innovators/early adopters. But public dissatisfaction with traditional products and the change in the content consumption paradigm may significantly shift the power balance in this market.
Traditional social networks use the user's social graph, or network, as the basis for creating content feeds. In short, the feed wants you to be in one content bubble with your friends, colleagues and relatives. With anonymous social apps, even if your interests aren't always the same, the content that feeds into the app is typically generated based on the user's attention. In iFunny's case, we understand this only by using machine learning algorithms that learn from the actions performed within iFunny. We know absolutely nothing about the user's offline life, except the phone model and OS version.
Uniting users in the content bubble is based only on interests. In reality, these could be people who would never meet or communicate in ordinary life. We bring together not only physical people, but also their virtual avatars, erasing concepts such as gender, age, skin color, orientation, financial status and other distinctive features of a human being.
So, is this good or bad for society? Which content model makes people happier?
It's impossible to draw a definitive conclusion. Social networks entered our lives less than 10 years ago, but we felt their real impact within the past few years when the authority of content creators in social networks outdid traditional TV and media. The original idea of blurring the boundaries in society resulted in even greater social disintegration; the degree of negativity and hatred exceeded all possible limits. Jealousy, politics, propaganda, fake news — that's what fills the services that were supposed to bring the world closer.
This has undoubtedly contributed to the development of anonymous social networks. Such services as Discord, Reddit, iFunny and Telegram are experiencing audience growth, but it is still unclear whether they will repeat the fate of the groundbreakers or bring a more positive contribution to society.
It's difficult to predict because on the one hand, there are advantages to never being personally criticized or harassed. You can be more candid in communicating with like-minded people without the fear that your profile will be linked to you in the real world. But at the same time, it makes some users think that anonymity equals impunity. Context pre-moderation and post-moderation are some of the most serious challenges for developers because the specificity is the absence of defined characteristics; therefore, it is not always fully and successfully automated, and human labor in this matter never has a 100% success rate.
While such projects are at an early stage and have not yet reached the scale of their predecessors, it can already be concluded that they have the ability to be more balanced, less toxic, safer and less susceptible to political propaganda, but at the same time, they require additional moderation controls.