The number one thing influencing what I write about nowadays is what I’ll call the unreliable information hourglass. It’s my unacademic and exaggerated idea that while information used to be less reliable as it aged, the incentivization of fake news and AI-generated listicles means that in many cases, information (online) is now less reliable the more recent it is. 

That creates a bimodal, lopsided hourglass distribution, with the highest concentration of untrustworthy sources occupying the two extremes of the timeline. And that means the pinched section in the middle is, I believe, the most worth writing and reading about.

From there to here, from here to there

Churches and states maintained such a stranglehold on information throughout early history that, as recently as 2013, historians were making shaky claims about what year we actually live in. 

The Phantom Time conspiracy theory postulates that a Roman Emperor and a Pope sat down and agreed to jump the Anno Domini calendar ahead 297 years to bolster their claims to power. A subset of the population was more willing to believe that we live in the early 1700s than they were to trust historical sources many hundreds of years old. It's not hard to see why!Information from the time was overcentralized. It was authoritative but impetuously self-interested. 

Today, sources are overly decentralized. They are novice and even more self-interested than those who came before them. “It’s easier than ever to share information and perspectives from different parts of the world,” Ethan Zuckerman wrote in Rewire. And yet, “We may now often encounter a narrower picture of the world than in less connected days.”

Everything from claims that Sam Altman was ousted for secretly achieving and unboxing AGI to supposedly academic support for LK-99 as a superconductor were simultaneously front-page news and widely distrusted. The more those claims aged, however, the easier it was to see them for what they were.

There are mountains of sand at the two extremes of the unreliable information hourglass; rough and jagged nuggets, difficult to separate from the whole. The grains of information that sit between these dunes, the Goldilocks era of reliable information, was accessible at unprecedented levels yet still relatively expensive to get attention (pre-Twitter et al., at least). That’s where you’ll find the topics I’m writing about.

At the hourglass’s isthmus, somewhere between the founding of the Associated Press in 1846 and Facebook firing its Trending list human curators in 2016, information gatekeepers were able to justify their existence with expert commentary and fact-checking. 

These are not academic claims! Someone could dismantle my armchair assertions with a single Wikipedia page (Woodrow Wilson’s tirade against fake news in his America First speech, from 1915? The AP self-censoring reports on Nazis during WWII?). My point is not that we should all be creating and consuming academic, peer-reviewed history papers. We have enough arrogant, amateur historians and conspiracy theories to last us decades.

Instead, I’m arguing that if you want to write about something like, say, building software that breaks the norm, I believe filtering it through the lens of a well-cited article on Eudora’s popular-yet-unfriendly interface is not only more interesting than a stream of consciousness “Thought Leadership” post–it’s also a much stronger signal that there was time and effort behind the writing.

History and quote-focused articles make it far easier (and more enjoyable!) to build an audience than writing single-voice opinion pieces or tutorial articles that only tend to work at scale.

Media archaeology and exhibits

My focus on writing, versus broader “content” generation, is intentional. Producing videos like those from Veritasium and CGP Grey or podcasts like Hardcore History and Acquired require more than a little algorithmic luck and multimedia expertise. And yet, audio and video aren’t conducive to fact-checking mid-consumption (for me, at least). Few people are pausing YouTube or Spotify halfway through to check a citation. Here, blogging and newslettering have the advantage.

I get unreasonably excited about obscure finds from the pre-PPM-advertising internet: An old USENET post from someone who later became famous, a Geocities site with the earliest known use of popular slang, an April Fools RFC from the early 90s. They are quaint and quirky, and I can pepper them throughout my writing as a sort of search query for others like me. Digital paper trails are catnip for me and the people I want to interact with.

I don’t want an idea or explanation or story to be summarized or dumbed down for me. I want to go on a journey, starting at the beginning, adding context and nuance, and arriving at a (but not the only) possible conclusion. Single-source information cannot give this to me, and so I avoid turning around and giving it to others. Good writing is a melting pot of primary sources that have stewed for at least two or three decades.

“With copies of the Internet over time and cross correlation of data from multiple sources, new services might help users understand what they are reading, when it was created, and what other people thought of it,” Brewster Kahle, founder of Internet Archive, said in a 2015 interview. “People might be able to give a context to the information they are seeing and therefore know if they can trust it.” 

If you want to earn people’s consistent trust and attention, Kahle provides a reliable formula. Responsibly aggregate stories and quotes that educate and entertain. You’ll likely be surprised at how little time it takes to dredge up citations and references to surprising moments in tech history.

How and where I dig 

I don’t think the list of places I frequent when doing research matters as much as the small idiosyncrasies I attribute to each of them. I’m happy to share any tangible tips I can think of here, but the joy of wading through the Goldilocks era of reliable information comes from developing your own grasp of each wellspring. It’s better to immerse yourself in these digital dig sites and take your own lay of the land.

Research almost always ends up being over half the work of producing an article, but not because it is difficult or slow-going. On the contrary, there’s usually too much good stuff to sift through! I find that it pays off later to marinate in whatever I stumble across, to marvel that it happened and that there’s a record of it, to pull on threads even when I’m reasonably certain I won’t include them in my writing. Doing these things is the reason the writing part flies by so quickly.

“Web 1.0, where we had the Wikipedias, the Internet Archives, you had blogs and things like that, by participating in the internet, you left something that people could build on.” Brewster Kahle reminisced during a panel with the President of the New York Library. “And Web 2.0, if you define it as kind of these platforms, it just scrolls away. It's not called pages anymore, it's called a feed. I mean, a feed. Right? Isn't that what you do to horses?” We can all do better than that.

I go to the Internet Archive for old software manuals but prefer Google Books’ search function for basically everything else. Hacker News has a pretty great search function, but it’s better to use a `site:` search via Kagi if I’m on the hunt for insightful or insider comments. Matt Guay, my partner in tech history, has some great tips for finding old USENET posts.

Then, I put what you find in chronological order. Connect dots between people and platforms, events and outcomes. Distinguish between theories and certainties. And if you want people to trust you, put your name on the (by)line. Kagi and Google both let you search within custom date ranges, and I find that anything before ~2012 is far more likely to have an author byline than you see today. Reach out to them! Most are more available to answer questions than you might think.

If you’re not scared to publish a bit of history, you might be doing it wrong. For example, Matt wrote The Very First Web Apps for Zapier in 2017 and is constantly finding older, better “Firsts” since then. That’s fine! The people who have reached out to him have done so more out of mutual curiosity than one-upmanship.

Finally, don't be afraid to ask for quotes and interviews. One of my favorite Pith & Pip articles is on the History of Sticky Keys, which we stumbled into writing after reading old copies of PC World and seeing an attribution to the inventor (or, rather, one of them. It was a wild calculus-was-discovered-twice bit of tech history). We had a harder time getting the inventor to stop talking than we did securing the interview! 

At best, it’s front page. At worst, it's still better than AI slop

Maybe the unreliable information hourglass and the tech history in the middle of it is just my megrim. An unhealthy need for cognition. But as long as I publish with links, caveats, and a diminished ego, my writing has a decent chance of getting attention (from people I want to read my work) and being referenced down the road.

AI is improving all the time. I’ve set a reminder to reread this article in a year to see if I have a machine competition in the media archaeology game. Not that it would change much. I enjoy the work of research and investigation, peer review with Matt and others, and sharing the stories I’ve unearthed with others who appreciate them.

Don't write about tech history (or hire me to write about tech history for your company) because it pays off. Do it because it creates community, empathy, and trust. And because it's fun.

Header photo by Birmingham Museums Trust on Unsplash