Overcoming Exhaustion, Rebuilding Resilience
I’ve been experimenting for four years to defeat the internet’s hot-start problem. Here are some learnings.
The second essay in a two-essay series.
“The readiness to submit to the Nazi regime seems to be due mainly to a state of inner tiredness and resignation, which … is characteristic of the individual in the present era, even in democratic countries.” — Erich Fromm, in 1941’s “Escape From Freedom”
“I can’t deal with the notifications 24/7, so I try to stay away from it.” — Woman, 20s, in a 2024 survey “Tuning Out: Americans on the Edge of Politics”
At the end of 2022 I learned something alarming about myself: the more news I read each day, the worse I slept that night.
The news-tracking model I created for myself showed that I slept 31 minutes less (-7.5%) after reading more than 40 articles a day than on days when I read under 10. The sleep deficit was even worse at the extremes: I slept 50 minutes less (-11.6%) on 40+ article nights than when I read no news at all.
In my previous essay I warned that the internet’s hot-start problem is being exacerbated not just by information overload and technological change but also deliberate efforts to create more cynicism and confusion that exhaust us. My hypothesis is that this exhaustion weakens human resilience for critical thinking and leads us to seek two very simple, low-entropy information patterns: 0 for outright avoidance, or 1 for an echochamber. These tune out states aren’t good for an informed society before an election or what might come in the months and years ahead.
For the past few years I’ve been studying my own behavior to better understand how humans can rebuild their resilience and move beyond low-entropy information patterns. Self-reported news exhaustion and news avoidance have been spiking in recent years, but there’s been little data-backed study into how news reading or information consumption affects sleep itself. I’ve been experimenting on myself not just for the practical necessity of wanting the best information to make decisions but also to prevent deliberate cynicism from exhausting me. What I’ve learned through these experiments informs what we’re creating with a new product called Frontpage.
The chart above shows how my sleep patterns changed in 2023. My overall sleep increased nine minutes to 6:46 a night last year, a modest gain overall. But my sleep on 40+ article days improved by a remarkable 24 minutes (6.33%); the sleep amount at 2023’s reading-behavior extremes — 6:44 a night — was the same.
This experiment wasn’t conducted in a controlled environment, and I don’t want to oversell its implications. But what this exercise demonstrated was that better access to data about ourselves can help us overcome news exhaustion when we’re aware of the problem. Reducing news exhaustion won’t end news avoidance on its own, but it’s a good place to start. I’ve identified three changes I made to my reading behavior that might have helped.
1. No news before bed.
I lost 42 minutes of sleep at night in 2022 if I read more than 20 articles after 6pm instead of not reading at all. I was surprised by that huge decline.
I cut back my reading in 2023 after seeing 2022’s results. That shift led to the following:
- I decreased my afternoon reading by 27% and my evening reading 36%.
- I cut my late-night news reading by 45% after 10pm and 53% after 11pm.
- I increased my nights with no news by 60%.
Those changes seemed to have helped: 2022’s 42-minute decline after 6pm dropped to a six-minute decline in 2023. This improvement has mostly held in 2024.
2. Taking longer breaks.
In 2023 I turned off all news notifications and took more breaks from reading the news. I increased my low-reading days (<10 articles) by 70% and my super-low-reading days (<5) by 200%.
When I did read the news I packed it into a few days each month, increasing my high-reading days (>40) by almost 40%. Most of that reading occurred twice a month on Tuesday or Thursday mornings.
3. Removing sensationalism; diversifying inputs.
2023’s biggest shift was changing what sources I read for information: I stopped reading three of 2022’s top-read sources — CNN, Fox News, and the New York Post — by at least 75% on my high-reading days, the biggest decline in my top-read sources. That change alone may have contributed to the 24-minute sleep improvement on high-reading days.
Instead of reading CNN or Fox I started reading more local news from California and global news from Asia, Europe, and South America. Diversifying what I read not only made me better informed but made me feel better informed: decreasing sensationalism and improving the feeling of accomplishment both likely improved the quality of my sleep.
II. Rebuilding Resilience
“But diversity, the property that makes resilience possible, is vulnerable to blows that are greater than natural perturbations. It can be eroded away fragment by fragment, and irreversibly so if the abnormal stress is unrelieved.” — E.O. Wilson, in 1992’s “The Diversity of Life”
I first started calling the internet’s problem a hot-start problem when I worked at Robinhood and watched young investors make poor investment decisions because of bad information choices; those who under-performed the market in 2021 were overly reliant on an echochamber of one or two sources for information. The internet’s hot-start problem has made low-entropy information patterns preferable in what feels like an increasingly chaotic world. My Robinhood experience strengthened my belief that improved resilience against echochambers is necessary for making optimal life choices.
Scientists have studied resilience in evolutionary biology and how species adapt over time, but there has been little study into how humans can improve resilience when faced with too much information. Strengthening resilience means considering different ideas both to improve critical thinking but also to increase our empathy for those different from us. Diversifying what we listen to, watch, or read is difficult but can lead to a stronger and more resilient mind.
The chart above is the one I use to assess my own resilience each year: it shows my cumulative news reading from the most to least read source. 50% of 2022’s reading came from my top 1% of sources, an overly concentrated amount. The line’s distribution largely follows the 80/20 rule — also known as the Pareto principle. The red dot marks the Pareto value (80%) on the line.
I measure improved resilience each year by moving the Pareto dot to the right. That’s accomplished by doing three things:
- Read less breaking news. (Make the line shorter.)
- Read more sources. (Make the line longer.)
- Better distribute reading across sources. (Make the line flatter.)
The chart above shows my progress since 2022: overall news consumption has dropped almost 20% while my total sources increased 7%. Meanwhile, 2024’s top 10 sources make up less than 40% of consumption, a 24% improvement from 2022. Taken together, these changes have increased the Pareto value by 81%.
Some of my improved resilience comes from finding new information: 40% of my news consumption now comes from sources I’ve discovered since 2021. Substack alone has contributed 112 new sources during that time.
But the biggest shift comes from reading more local and global news.
The above graphs show how my news consumption shifted from political, sports, and business news in 2020 to local and global news in 2024. Reading local and global news is not only less sensational and exhausting than what’s on social media; it also helps connect the dots on stories lost in social media’s noise. America’s manufacturing resurgence is one example of a story not discussed enough. So is the increased corporate ownership of homes across the United States. The internet’s shift to prioritize volume instead of proximity has led to the rise of deliberate misinformation and the death of local news but also to a prevailing wisdom-of-the-crowds approach that has cheapened truth and leaves us all less informed. Diversifying what I read to include more than 200 local and global news sources has made my mind more resilient and better informed, even if the process to discover that information is sometimes harder to do.
III. Frontpage
Improving informational resilience has broad applicability in how we do our jobs, make better financial or political decisions, and in our own mental and emotional health. More AI-generated content and political uncertainty will make resilience a necessity in the coming years. But what I’ve found over the last four years is that there’s no product or experience that makes me smarter and more resilient while also respecting my attention. That frustration led me to start creating Frontpage.
Our goal with Frontpage is to empower humans with deep expertise and exceptional taste to curate the internet for others. We’re creating software that enables curators to help readers strengthen their attention and build more resilient minds. Today we’re releasing two very-early Frontpage experiments that demonstrate how we’re thinking:
- Across America features the best local reporting about America all in one place.
- Multipolar features the best international reporting all in one place.
Each frontpage is designed to highlight quality journalism that exists already but doesn’t receive the praise it deserves. There’s no mysterious feed on Frontpage, just well-intentioned curation. I’m the only curator now, but we look forward to welcoming others soon.
If our mission to build more resilient minds resonates with you, we’d love to hear from you.
~~
Join Frontpage’s waitlist here.
You can send me comments via email or sign up to receive future essays in your inbox.