The Hot-Start Problem
The internet’s deliberate chaos is weakening human resilience to seek out quality information. Enter Frontpage.
The first essay in a two-essay series.
“The amount of information was increasing much more rapidly [in the 15th century] than our understanding of what to do with it … We face danger whenever information growth outpaces our understanding of how to process it.” — Nate Silver, The Signal and the Noise
“The way to deal with them [the media] is to flood the zone with shit.” — Steve Bannon
Imagine needing to board a high-speed train that has no conductor and refuses to wait or slow down for its passengers. I use that metaphor to describe what it feels like today to consume news or information online: a runaway train that’s impossible to board.
I call it a hot-start problem.
A hot-start problem is when we face too much new information too quickly to start understanding it, creating gaps in our knowledge and cognitive dissonance in our thinking.
Hot starts occur on a personal scale whenever we change jobs or move to a new city. And they happen on a global scale whenever technology makes it cheaper to distribute information.
We’re currently living through the internet’s hot-start period right now: we create enough information to fill 3.14 billion iPhones everyday, a 75x increase from 2010. Meanwhile, the average lifespan of a Facebook or X post is less than one hour, a 70% decrease from 2011. Keeping up with our fast-moving feeds, notifications, and emails is comically impossible, leaving us trapped inside an informational fog. Everyday we live this hot-start problem together even if we don’t know it.
What distinguishes the internet’s hot start from its 20th-century predecessors is that its problem has moved beyond technology and into politics: some actors are deliberately making it worse.
The internet’s 25-year consolidation around a few engagement-based information patterns has not only transformed human attention into the internet’s primary currency but also stripped information from its original context, creating an arbitrary flatness that prioritizes volume over proximity and rewards noise. That change allows anyone from Russian trolls to American politicians to start a fresh controversy with a few posts on social media, exploiting the internet’s speed with manufactured lies and distraction to create confusion. These distractions — the demagogic attack on migrants across Western countries is perhaps the best example — inspire both narratives and counter-narratives that perpetuate the hot-start’s cycle and exacerbate the noise. And that is the point: the constant flooding of the zone with shit attacks our attention and makes everything appear more negative and chaotic than it is in reality.
The internet’s allowance for manufactured chaos has destroyed our trust in anything except aesthetic authenticity — brands or vibes — while also accelerating a collapse in our critical thinking, making hagiography’s numbing simplicity more attractive than what’s required with detached discernment. Our political media has not only failed to respond to this moment with help but instead distracts us further with “live” news that inhibits any opportunity for sustained attention. Meanwhile, the 21st-century shift from editors to algorithms leaves us to navigate the noise on our own, a futile exercise without the right education or tools to get started.
What worries me is that the internet’s deliberate chaos is both exhausting us and weakening our resilience to seek out quality information when we need it most. My hypothesis is that the hot start leads humans to want very simple, low-entropy information patterns that reduce to two states, 0 or 1: 0 for outright avoidance, or 1 for an isolated echo-chamber.
The phrases news fatigue and news overload have both spiked at least 375% in usage the last 10 years, but those jumps pale when compared to news avoidance’s 1700% increase during the same time. That change in language mirrors a change in human preferences: almost 40% of Americans now say they avoid the news. Meanwhile, those who do consume news are seeking out fractured yet solitary echo-chambers for their information. It’s why a few podcasts are winners but also why some Americans only watch MSNBC or Fox News: clear and agreeable narratives are preferable in what feels like an increasingly chaotic world, even if that narrative is filled with falsehoods, cynicism, and contempt. The pursuit of any dialectical inquiry beyond one’s immediate zone of interest is becoming harder to find.
The internet’s hot-start problem has already been called by other names — Total Noise, Present Shock, The Shallows, the Digital Maginot Line, the Age of Acceleration. Framing the internet’s problem as a hot-start problem isn’t to appropriate someone else’s analysis but to take a contrarian approach to a popular Silicon Valley framing — the cold-start problem — that is well understood by those who work in technology and can materially help start fixing this problem. And since Silicon Valley generally likes contrarianism and difficult problems, putting the two into one package might give the internet’s hot-start problem not only a clear definition but also the added attention and urgency that it deserves.
II
“The real cycle you’re working on is a cycle called yourself. The machine that appears to be “out there” and the person that appears to be “in here” are not two separate things. They grow toward Quality or fall away from Quality together.” — Robert Pirsig
My own personal history with the hot-start problem started on January 10, 2011: I was working at The New York Times for the newspaper’s public editor, an ombudsman-type role that doesn’t exist anymore but was tasked with being the “reader’s representative” inside the newsroom. My job was handling the 300 or so angry emails that readers, businesspeople, and politicians sent each day to the public editor demanding answers about why The Times said this or didn’t say that. But on the morning of January 10 the public editor’s inbox was overflowing with the same question from readers across America: why did The Times kill Gabby Giffords?
Gabby Giffords was a U.S. congresswoman who was shot in the head during a public event in Tucson, Arizona on January 8, 2011. Eighteen people in total were shot, six died, and for about eight minutes The New York Times said online that Giffords had died, too. But Giffords wasn’t dead; she miraculously survived the shooting. I don’t remember who caught The Times’s error, but soon it was everywhere on Twitter, leaving readers aghast the newspaper could make such a mistake and demanding to know how and why it happened. The public editor wrote about the incident soon afterward, and with hindsight I wish I had suggested removing the names of the reporter and editor involved, not just to absolve them of individual blame but because the failure didn’t rest with them alone. Nor even with The Times itself, which was one of many news organizations struggling in the early 2010s with how to report accurate information online when the internet’s collapsing of time was just beginning and its relentless demand for speed was unforgiving.
What happened with The Times and Gabby Giffords was the animating lesson behind my decision-making going forward: humans always want information faster, and technology will deliver it regardless of whether it is right or wrong. The inherent tension to discern right from wrong is difficult when demand for fast information is high. I came to know this tension well when I helped Apple launch its News team in 2015 and was responsible for making these decisions everyday, an experience where I made sure information moved fast without compromising accuracy. But what I realize now is that my approach for five years was wrong: checking facts before publishing is easy, but those facts mean little if the vehicle itself is a distorted view of reality.
Fast forward to 2018 when I suffered a panic attack for the first and only time. I ascribed it to the stress of my job but also the cumulative stress of reading the news. Since it was my job to read the news — scanning thousands of headlines everyday and reviewing objectionable content like terrorist attacks and shooting videos — it seemed clear that not only were the news cycle’s 24/7 chaotic demands taking a toll on me but that the internet’s issue with technology and politics could reach into mental and emotional health too.
Writing about the internet’s hot-start problem without acknowledging its effect on our mental and emotional lives would be a disservice to the gravity of the problem and fail to capture why I care about it: I don’t look at the hot-start problem as some abstract project but as a real and visceral problem that I’ve experienced myself. Trying to understand the problem for almost 15 years makes me realize that it centers on human attention, not just the increasingly shortened “eyeball” attention that is cheapened and traded online but the longer attention that sustains deeper awareness in the present moment. The hot-start problem has left our longer attention under constant attack, making it hard to be present for sustained periods of time or to think critically for ourselves. I worry our short- and long-attention spans are diverging on separate tracks too late for us to reconcile. History demonstrates that we will never slow information technology down. But that can’t be the goal. Instead, any solution to the hot-start problem must start with giving humans freedom to control their online attention again.
III
“To define an agricultural problem as if it were solely a problem of agriculture – or solely a problem of production or technology or economics – is simply to misunderstand the problem, either inadvertently or deliberately, either for profit or because of a prevalent fashion of thought. The whole problem must be solved, not just some handily identifiable and simplifiable aspect of it.” — Wendell Berry, Solving for Pattern
“In a healthy society there will be as many pattern languages as there are people — even though these languages are shared and similar.” — Sara Ishikawa, A Pattern Language
For the past few years I’ve been experimenting on the smallest scale possible to understand the internet’s hot-start problem by examining my own online attention: I created a model that records every article I read online and synthesizes those records with my daily health data, technology usage, and other behavioral traits, giving me a comprehensive view of how my online attention patterns change on a short- and long-term basis. The data findings are compelling, and this experiment gives me confidence that we can retake control of our attention with the right technology.
Creating a browser extension or an AI agent for the technology described above is easy to build and might be profitable, but it doesn't address the core issue: where does the internet’s hot-start start? Doing this experiment leads me to believe it starts with the information patterns themselves, the opaque sorting models that treat attention as currency. We can now easily create newsletters, short videos, and podcasts, but we’ve had very little innovation for how that information is presented or filtered.
A former Apple colleague and I have been working together this year to design new information patterns that reduce noise and elevate quality news and information. We’re calling these new patterns frontpages. We’re calling the platform Frontpage.
The name Frontpage isn’t meant to invoke nostalgia but to serve as a reminder that free speech means more than just screaming into our social media feeds. Information patterns are speech too. Curation is an art that was made invisible when social media eliminated our freedom to choose. We believe there’s a better way.
Our goal with Frontpage is to decentralize curation by empowering humans with deep expertise and exceptional taste to curate information for others. Curators will be able to easily gather, filter, arrange, and synthesize information from across the internet into a single frontpage, either as an individual or in teams. AI assistance can also help, if wanted. We’re making a strategic bet that humans and technology can deliver the best outcomes when working together, not one over the other.
What makes Frontpage unique is our emphasis on quality and time: we’re creating curatorial tools that enable now and live to mean months, years, and decades instead of just seconds or minutes. Our thesis is that lengthening time and making it dynamic will reduce the internet’s deliberate chaos and encourage sustained attention on quality information. Qualitative feedback in our private beta testing this year suggests frontpages with longer timescales help make humans more resilient for critical thinking.
It will take us a few years before Frontpage reaches our full vision, and we know it won’t end the internet’s hot-start problem on its own. But we hope to help humans start reclaiming their attention, even if starting itself — hot or cold — is sometimes the hardest thing to do.
~~
Join Frontpage’s waitlist here.
You can send me comments via email or sign up to receive future essays in your inbox.