The “attention economy,” as it’s termed nowadays, has an opportunity to develop differentiated measures of what kinds of attention it is garnering and basing value on. In doing so, it can shift into more helpful types of interactions that individuals and communities share with each other.

To do so, it would take an evolution in incentive and disincentive structures that currently drive decisions of major players and their disruptors in the attention economy. The new structures must account for long-term effects, future generations, and bad actors.

The current landscape of the attention economy generally lumps all types of attention into a positive feedback loop that customizes others’ feeds that could potentially get any one user more potential attention on the platform. But not all attention is good attention. For example, attention towards preteen girls could come in the form of sexual predation (NYTimes: A Marketplace of Girl Influencers Managed by Moms and Stalked by Men). A differentiation of the intentions behind each instance of attention is not possible or utilizable, for the most part, on current forms of social media, which I’ll refer to here as algorithm-based media.

In my opinion, too much attention can be unhelpful, and the wrong kind of attention can be harmful.

• Too much attention - whether in the form of likes, comments, or views, in excess of what might be considered “human scale” - is overwhelming and addicting in ways that feel similar to texture/sugar/salt balance in addictive foods, salaries in morally-questionable careers, and UX design that works “with” your millennia-trained human instincts. Just manipulative enough to keep you locked in, and not obviously noticeable or alarming enough to look for ways out.

• The wrong kind of attention - from people or organizations of people who have malicious or maladaptive intentions, whether explicit or not, attention can feel threatening. Of course, there are the common instances of Internet bullying, blackmail, or disinformation. But there are less obvious manifestations as well. For example, I used to practice calligraphy and handlettering, streaming an overhead view of my desk on Twitch, and built a small online community around it. It was fun overall, but I could tell that some individuals watching and chatting felt a sort of attachment due to circumstances in their lives and were looking for a salve that I did not have the training nor the capacity to provide. It felt emotionally heavy at worst, uncomfortable at best.

These are only two examples of types of attention that might be differentiated from user attention/engagement as measured by digital platforms. Plenty has already been written about rage as an emotional ramp for such engagement already, but I think that taking a look at the other types of attention can help policymakers to adapt algorithm-based media towards positive social habits.

A wise friend recently told me that we may not always be able to read social cues, depending on the cultural setting, and we may even misinterpret them, but we can almost always read intention. I’d add: it’s much easier to intuit intention if our attention isn’t too scattered, and if we’re physically in the vicinity of the other person. His advice helped me to rest my nervousness about frequently “getting it wrong” when around Japanese people (I’m currently in Japan; he used to live in Okinawa). I have a hard time corresponding with new friends here over text and social media without the context of voice, facial expression, body language, environmental context, etc. I love them no less even when I don’t understand what they are trying to say, or when I have trouble getting across what I want to express. But it did, for a while, cause a certain kind of retreat within myself, feeling like I was always messing up. Now, I’m finding an awkward balance, reducing text-based communications to make room for what may be considered old skool at this point, but personally preferable: face-to-face visits and trips, time to wander/meander and even run errands together with friends and communities I deliberately choose to live next to, and a good old voice call.

My research training and experience leans more quantitative than qualitative, so I have previously had a tendency to discount stories and anecdotes. However, I’ve learned to trust that over a decade of observing social interactions in social gatherings, especially during the rapid adoption of digital social engagement on phones, and especially in Silicon Valley, I clocked changes that were both noticeable and alarming.

Now, there are more studies and articles on social media’s effect on social relationships and interactions. One good source to check is The Center for Humane Technology’s Ledger of Harms. Here are a few from the Social Relationships section:

• People who took photos to share on Facebook experienced less enjoyment and less engagement with the scene compared to those who took photos purely for their own pleasure. Closer analysis indicates that taking photos to share on social media increases a user's focus on their own self-identity and self-presentation, distracting them from connecting to the world around them.

  Source: Barasch, A., Zauberman, G., & Diehl, K., 2017. Journal of Consumer Research 
  (https://academic.oup.com/jcr/article/44/6/1220/4627834)

• When encountering someone with an opposing political viewpoint, people are more likely to judge them as warm and intelligent if they hear that person’s ideas spoken rather than written down. Unfortunately, many social media platforms are currently designed to focus on text, reducing the chances of genuine discussion and debate and increasing the possibility of polarization.

  Source: Schroeder, J., Karadas, M., & Epley, N., 2017. Psychological Science
  (https://julianaschroeder.com/publications/)

• We are so distracted by our phones that we often fail to see the most basic things, sometimes at great cost to ourselves and others. Security camera footage from San Francisco public transit reveals that a gunman was able to pull out his gun and openly handle it at length without anyone noticing, before he eventually shot a fellow passenger.

  Source: O’Connor, L., 2013. Huffington Post
  (https://www.huffpost.com/entry/san-francisco-train-shooting_n_4066930)

• 50% of parents reported that mobile devices typically interrupted the time they spent with their children 3 or more times each day; only 11% reported that mobile devices did NOT interrupt their time with their children.

  Source: McDaniel, B., & Radesky, J., 2017. Child Development
  (https://srcd.onlinelibrary.wiley.com/doi/abs/10.1111/cdev.12822)

• The mere presence of a mobile phone can disrupt the connection between two people, leading to reduced feelings of empathy, trust, and a sense of closeness. In a series of studies, researchers found that when pairs of strangers were asked to have meaningful conversations, their ability to connect emotionally was significantly reduced if a mobile phone was visible.

  Source: Przybylski, A. K., & Weinstein, N., 2013. Journal of Social and Personal Relationships 
  (https://journals.sagepub.com/doi/full/10.1177/0265407512453827)

• A person’s social media usage level significantly predicts their level of neuroticism/anxiety one year later, as shown by a long-term study of 11,000 people aged 20-97. In addition, levels of neuroticism/anxiety predicted later levels of social media use, leading researchers to suggest a possible negative downward spiral linking these two processes.

  Source: Andrews, N. P., Yogeeswaran, K., Wang, M.-J., Nash, K., Hawi, D. R., & Sibley, C. G., 2020. Cyberpsychology, Behavior, & Social Networking
  (https://www.liebertpub.com/doi/10.1089/cyber.2019.0744)

• Half of Americans report that their partner is often or sometimes distracted by their devices when they are trying to talk to them.

  Source: Vogels, E. A., & Anderson, M., 2020. Pew Research Center
  (https://www.pewresearch.org/internet/2020/05/08/dating-and-relationships-in-the-digital-age/)

Keep in mind that only some effects of algorithm-based media have been researched thus far, and research has been retrospective at best. It does not keep up with harms and 2nd-order, 3rd-order, etc. effects to us or future generations.

A clearer understanding of how digital vs in-person social engagement may contribute to social underdevelopment or maladaptation is warranted. In addition, because of the speed of algorithm/AI-based media platform development, more current research methods and even speculative studies would help guide us towards preferred social futures.

What could we do with this clearer understanding? Well, what if algorithm-based media that lowers health indicators are required to give a proportional amount of funds to support social infrastructure that offsets their harms, kind of like carbon offsets or cigarette taxes? And what if it was structured in a way that heavily disincentivizes harms later discovered, as well, making them incentivized to conduct the research themselves about the potential negative implications they could cause? Perhaps fewer of them would strip away their safety/ethics/content moderation teams, for one thing. Transparency would likely increase. And some of the levers that drive the speed of development would be adjusted so that we could create and use algorithm-based media in a humane way.

Just some ideas.