The Firehose and the Feed | Weeknotes #427

The Firehose and the Feed

I’m one of those people who live their life with disappearing messages turned on wherever they are available. Partly because I’m a millennial, who grew up with just enough space for 15 text messages on their first mobile phone. But I also have it turned on because I don’t think of chat, MSN, SMS, Signal or whatever to be “writing” and worth keeping around. It’s a kind of speech. But sometimes I do wish I still had transcripts!

It’s been over a decade since the Brexit vote happened. At the time the group chat was full of people talking about the “bad vibes” on social media and the feeling of experiencing an info war for the first time. I wish I could show you just how sensitive people were to currents on the social seas. We have of course since; found out that 1% of accounts generated 1/3rd of all messages during the referendum campaign period, and most of them were bots.

From the distance of a decade this is all immediately recognisable as Bannon’s “flooding the zone with shit”. But even before the gamergate era , 4chan figured out that platforms could be “played” by making the information environment ungovernable. Earlier coordinated raids got strange or offensive phrases trending and were proof-of-concept experiments. A Demonstration, in the early days of social media, that if you applied enough volume, fast enough, across enough accounts, you could take control of what everyone else was looking at. A real insight into the mechanical physics of Code Space in the early 2010’s for me.

After Brexit and Gamergate, but before Bannon’s “Flood the zone” strategy came to light in an interview in 2018, RAND released a super interesting report called “The Firehose of Falsehood”.

It described in strategic form, what my friends and I had been experienced “being inside of” during the Brexit campaign. That propaganda that behaves less like a single persuasive argument and more “like weather”. Instead of trying to win by proving one true singular narrative, you win by throwing out many stories quickly though multiple channels, repeating them, and keeping the whole bit going even when claims are contradicted or challenged.

The effect being, that the information environment becomes noisier and more exhausting to navigate. When volume stays persistently, attention gets pulled towards whatever is loudest, most emotionally charged, or most easily repeated. Slower and more careful analysis struggles to stay visible.

This is all par-the-course in 2026. I think everyone reading this knows this is just how the internet works now AND that it’s all massively exacerbated by algorithmic feeds. But I think there are two problems with assuming everyone knows this:

1. “All social media is manipulated” as an online background condition isn’t an interesting story to the mainstream media, unless a specific example suits them and their own narrative/manipulation of the weather.
2. Lots of people don’t know! All the millions of boomers out there sharing ring footage of a baby saving a cat from a bear, or liking and posting abhorrent things about immigration. Has nothing to do with slop at all, it’s all info war.

A couple of years ago I read a super interesting paper from Karthik Srinivasan at the University of Chicago Booth School of Business called “Paying Attention”. Which explains why the firehose fits the physics of code space so well.

The paper treats attention as a reward: likes, upvotes, views, and comments are all modelled in the paper as acting something like a wage. Because receiving these things tends to make people produce more content afterwards. What the paper also finds is that ‘the wage of attention’ has diminishing returns. The first units of attention given to someone changes their behaviour online the most. Which means that spreading small bursts of engagement across many accounts, likes, RT’s, replies etc can be more productive than using bots to boot and single accounts voice.

When I read the RAND paper and Paying Attention side by side, the physics of the incentive landscape were made quite a bit clearer to me.

It is worth noting, that similar logics spring up amongst real communities too. The TPOT community on Twitter (the loose, intellectually inclined cluster of writers, thinkers, and other chronically online people) built a strong culture around linking and boosting one another’s posts. It was an unspoken part of participating in the community: people surfacing ideas they found interesting, helping good writing find an audience it might not otherwise reach. But the underlying mechanic is the same as the one describe in the Srinivasan paper. Mutual amplification distributes attention in small, motivating doses across a network rather than concentrating it on a handful of accounts. Almost all fandom communities work this way too.

The attention economy rewards the same behaviours regardless of whether the content being spread is a coordinated disinformation campaign or a genuinely interesting essay. The platform does not distinguish between the two. It just measures the engagement.

There is a practical implication buried somewhere in the above though I think.

That if the first units of attention are the ones that change behaviour most, then being liberal with the like button actually matters. Leaving a comment, sharing a post, replying to something you found worthwhile are small acts that actually work. In the face of “the firehose” the modest counter-move is to be deliberate about where your attention goes, and cultivating the things that you want to see more of.

This is of course, is not an argument that’s new around here, or anywhere else. Your attention is sovereign, and you should engage in the conflict of disinterest.

But I woke up this morning trying to think though the terrain, and what the rules of engagement might be given the nature of the weather.

On The Blog

World Literacy

This is one of those posts that arrives because an idea has come up three times in real life in the last few weeks and when that happens it’s best to write it down.

The idea is World Literacy1.

Audiences now arrive at a media franchise already knowing how worlds work. Not that world, specially, but worlds in general. Audiences have a new fluency, with expectations and a set of standards they didn’t consciously acquire, and they can’t necessarily articulate if asked. They got them from everywhere else. This is quite new. And if you are running a world, it’s one of the most important things to understand about the room you are walking into.

Start Select Reset 📑

Subscribing to SSRZ supports my online work and creative projects.

Get the Zine!
£5/month 💌

As a thank you, I send you my zine four times a year, just like it’s 1994.

No spam. No email. Cancel at any time.

Photo 365

The Ministry Of My Own Labour

I’m on the look out for work! Short term engagements or a more permanent situation.
Do you like my work and know someone who might find it useful?

• Finished tweaks to final 302 script
• Wrote and recorded pre episode drop. ‘Liner notes’ episode.
• Recorded 302!!!
• Edited 302.
• Did first pass layout of text essay for Start Select Reset
• Sorted out woocommerce so i can sell the zine. (Haven’t flipped the switch yet)
• Long call about agents, and problems that might arise in the near future.

Terminal Access

This piece on the Dutch military jailbreaking the F-35s vendor locked software just like an iPhone is extremely interesting, and seems like Europe really is taking software independence seriously.

Altogether, any kind of jailbreaking of the F-35’s systems would come with a serious risk of legal action by Lockheed Martin and additional friction with the U.S. government. What would have to happen for a country like the Netherlands to pursue that course of action would also likely be just one symptom of a much more serious breakdown in relations with Washington. Doing this could easily prompt a cutoff in spare parts and other support, if that had not already occurred, which would leave jailbroken jets quickly bricked on the ground. To be clear, cracking the software would do nothing to mitigate the downstream impacts of being shut out from critical sustainment pipelines.

Reading

Still reading A Million Years of Music by Gary Tomlinson. I’m about 3/4 of the way though now. We have just got to early homosapien language, symbol abstractions, speach and its impact on “musicking”. This book is a amazing. Just every chapter blowing my mind.

I’m still listening to Empire of Imagination: Gary Gygax and the Birth of Dungeons & Dragons by Michael Witwer. Unless i have unexpected long train journey in the next few weeks i might park it and move on to something else. I just cant get past the style of the thing.

I started Impromptu Man: J.L. Moreno and the Origins of Psychodrama, Encounter Culture, and the Social Network this week. The the father of psychodrama and *all* group therapy. He was an early critic of Sigmund Freud. Currently reading about origins and impact of Esalen Institute had on wider culture.

Songs for Sleeping Dogs – Ben Hackett

I came across this 2025 debut album by Ben Hackett this week. A solo album thats a mix of indie, folk and new age. Living up to its title, it has a dreamy cadence throughout, but tips in to psychedelia in places. I’ve mostly had this album on as I’ve been working and its very hypnotic and immersive. Really enjoyed it.

My favourite track on the album is Electric Pond Ripple. Here’s a video of him playing it live. On the album it uses piano not synth as the main keys instrument.

Remember Kids: