During COVID19 and subsequent increased social media consumption, the correlation between mental health and social media use becomes increasingly important to address. Particular focus is warranted on the machine learning or algorithms used on these platforms to keep us engaged and keep them in business.
When discussing any sort of systemic problem, It’s important to first acknowledge agency - in this case, the agency of users.
To this point, I spoke with Nir Eyal, author of Hooked: How To Build Habit Forming Products and more recently, Indistractable: How to Control Your Attention and Choose Your Life. Eyal believes that when empowered by understanding exactly how social media algorithms work, we can use social media tools for good and use them to build healthy habits in our lives.
“I wanted to figure out my own problem with distraction, and also share with others why social media is not something that we're powerless to,” says Eyal.
“The key is to sort out exactly where it hurts us, and where it helps us. We certainly know that along with potentially harmful side effects, they're also potentially very beneficial effects.”
Sorting this out is difficult, but not impossible. The key to this lies in 1) examining how our data sets are currently used in machine learning algorithms on social, 2) understanding the possibilities and limitations of our own willpower vs. these algorithms, and 3) where these limits exist, start designing technologies in ways that address the needs of users and improve their mental health.
So let’s start with the facts about data and algorithms on social media.
The business model is the message. From that business model flows company culture and beliefs, strategies for success, end-user experiences, and, ultimately, the very shape of society.
It is commonly understood that our private information is what makes money for most digital media companies. In fact, 98% of Facebook and Instagram’s revenue comes from advertising sources.
While many of us begrudge this model, we as a society have come to accept it as the norm.
Social Media User
However, during a time when we are online more than ever and large tech companies are being confronted for their business and privacy practices - perhaps now is the perfect time to take a second look at the data/money model.
Large corporations are paying them large sums of money to bombard our feeds with ads, targeted specifically for us based on 1) data they collect from us and sell with the bare minimum of our consent and 2) judgments they make about who we are and what we might be inclined to buy, which come from predictive algorithms carrying loads of biases and prejudices.
Unfortunately, many features of the platforms we use today are guilty of enabling discrimination, harassment, and ultimately violence that is, of course, mostly targeted towards these marginalized groups.
This is not to say that social media dictates how we think - again, we have agency. “This idea that it's somehow addicting all of us and hijacking our brain,” says Eyal, “is not only scientifically not true, it's actually hurtful.”
But what are the limits of our agency in platforms such as Facebook? Especially when we are forced to choose between online socialization or data privacy. Especially when machine learning algorithms are typically built by folks who are maximizing profit, not considering the impact of certain ads on those who are more likely to be negatively influenced by a particular ad, say a teen with an eating disorder.
At some point, we have to reimagine the data-for-sale, and discriminatory algorithm model. While this process will be slow (if not impossible) for larger corporations, there are social media startups such as Vigyaa.io who are setting the example that profit and health don’t need to be mutually exclusive.
“Algorithms are written for a very clear objective,” says Vigyaa.io Cofounder & CEO, Jitin Tuli.
“These objectives may result in some unanticipated/unplanned results. The big tech did not create these algorithms with any evil objective, but the side effects of those objectives turns out have some devastating effects on mental health.”
Being anonymous is at the core of Vigyaa’s post-sharing product. The platform even goes as far as not to use 3rd party tracking tools like google analytics, which trace users ip, but have developed their own analytics around platform usage.
“We are committed to protecting our users personal information and that's why we took the bold call to not have an email id attached to the account.,” says Vigyaa.io Cofounder & CEO, Jitin Tuli, “This gives them instant confidence and 100% privacy.”
The mission of the platform is especially suited for a world in lockdown with increasing rates of anxiety, as it provides a space where the user can safely vent, confess a dark secret or just be themselves.
Other social media platforms and interventions that understand and are executing on the relationship between data privacy and mental health include Live for Tomorrow, Trill and 2Swim.
Early testing shows that a healthier way to utilize user data may be Sentiment Analysis machine learning, which uses behavioral data (as opposed to personal data) to generate positive mental health benefits, instead of only generating profit.
“Our team is utilizing keyword analysis based on emotion,” says Tuli. “It is a huge contrast you see of a person sharing their posts when they do not carry a digital identity versus what you see on other social media platforms, they truly express themselves in an unbiased way without the fear of being judged. So, the first output of our still being built NLP/ML system gives a deeper insight of what is happening to individuals at a very emotional and personal level.”
Facebook Likes
The goal is to track the emotions of a user anonymously, and help them go from negative to neutral or positive emotion using discovered content. “That someone came into the platform with a very negative mindset, but when they got out of the platform, they were feeling a little better. Right? That's the win situation for us.”
Vigyaa.io has already gained praise from users who note that the platform’s community has been a great help, stating that people replying to each other's posts anonymously has helped them heal and not feel so alone. The site has aggregated 23000+ posts and 350k + users in only a few months.
Privacy assurances on anonymous social platforms add to the mental health benefits, allowing users the ability to express themselves without fear of this data being used for monetary gain.
Again, coming back to agency, it’s important to note that social media use is not a defining factor in a person’s mental health. There are many factors at play.
As Nir Eyal explains, “ If technology is causing the problem, then wouldn't you expect wherever the technology is used? That's where you should see the increase in teen suicide? Well, that's not what's occurring.”
Many countries including Norway and Japan have seen a significant drop in teen suicide, indicating that there is a quality of life issue - not a technology issue at play. However, this being said, social media can be a contributing factor in mental health. And since it is such an integral part of our lives, why not use it more consciously for good?
“We can use the technology for good, and help build healthy habits in people's lives,” says Eyal.
This starts with disrupting the myth that there is only one way to build and grow a profitable social media company.
“We believe that there are revenue models that can be created without a user profile,” states Vigyaa co-founder Gaurav Goel, “We want to change the data privacy debate by demonstrating an alternate approach.”
On a larger scale, the anonymous, behavioral analysis model could become the new normal for digital media companies as Big Tech continues to face scrutiny for their abuse of user data.
As Eyal eloquently states, “We can get the best of technology without letting it get the best of us.”
