PERSONAL READING, ON YOUR TERMSONLINE WORKSPACE
← Back to the reading room
Instead of Doomscrolling · 2026-09-25

What Happens When the Loudest Voice in the Room Isn't Human?

Notes on the Spiral of Silence, computational propaganda, AI and democratic access to information, and half moons in the Sahel.

Read original ↗

What Happens When the Loudest Voice in the Room Isn't Human?

Notes on the Spiral of Silence, computational propaganda, AI and democratic access to information, and half moons in the Sahel.

Mapu

Sep 25, 2026

230

12

42

Share

I was just scrolling through some comments on social media and it made me realize how bizarre it is that people are now replicating messages that, initially, could easily be detected as being left by bots.

Political propaganda spreads significantly faster when people start feeling confident enough to voice certain beliefs that would have been frowned upon in the past.

This is essentially the gist of the

Spiral of Silence

1

theory—the idea that people stay quiet when they believe their opinion is in the minority, and only speak up when they feel it’s shared. Except now, there is an entirely new layer that is emerging through the digital world that we must account for. Bots and fake profiles—which are getting harder to identify with the rapid development of AI—can make a deviant opinion look popular, and once this happens, real people start to feel safe saying it out loud.

It’s easy to think of propaganda as big posters from last-century dictators, or maybe as plain obvious now that we see the full picture in retrospect or from an outside perspective, like North Korea. In the current landscape, however, seemingly innocent content is actually propaganda disguised in some pretty wrapping. This is what is called

computational propaganda

2

, and it’s not so obvious to the untrained eye.

Content telling women how to step into our

“divine feminine energy”

? All of the fashion girlies are wearing the latest

Meta glasses

?

Stitches calling out men

secretly filming women for

“street interview” content

? Official government accounts posting

AI-generated visuals

?

Religious dog whistles

in

presidential speeches

? Once you see the patterns there’s no denying it, but to get there, we require informed perspectives.

Remember that if you think you are immune to propaganda, it probably means you’re not.

When public opinion about governance—and consequently, the way we vote—is largely defined by fabricated narratives or half-truths, where does it leave democratic governments and the designated overseeing institutions?

This question is more relevant than ever now that,

as of this year,

Cloudflare has reported

that for the first time, most of the traffic on the internet is non-human.

Of the crawlers scanning the web,

52% are collecting content to train AI

, and over

36% blend search, agent use, and training into one

. In other words, machines are navigating the web significantly more than we are. When the navigation process becomes commodified, whatever information AI takes in is what gets brought back to us. Furthermore, whoever decides what gets fed into it—and what doesn’t—gets to shape what millions of people consider to be the truth.

Some might argue that this is fine because they exclusively use AI for research for the sake of “optimizing workflows,” but I argue the picture is much larger than that. A recent example that comes to mind is one I came across in the following article:

Drop Site News

Israel Is Paying Millions to Train AI Chatbots How to Talk About Gaza. It's Working.

Since October, former Trump campaign manager Brad Parscale has been quietly overseeing an operation posting hundreds of blog posts on behalf of Israel. One article, titled “The Reality Behind Gaza…

Read more

3 months ago · 1157 likes · 34 comments · Nick Cleveland-Stout

It took fewer than a thousand articles—most of which barely crossed human eyes—full of Zionist propaganda, for some AI chatbots to deem them reasonable enough to relay the misinformation.

On a similar note, I came across

Deana Burke

on TikTok, and have now been following her insights on tech—particularly about AI product and brand recommendations—on

Malware by Boys Club

. As part of this research, she created a

fake natural deodorant brand

called Morrowen, only to see ChatGPT with browsing recommending it three weeks later. I wish this was only about deodorants though. Beyond the impact this is likely having on consumer behavior and the market landscape, this has tangible political implications.

A study by BuzzStream

of

the top 100 news sites in the US and UK found that 79% block AI from training on their content, and 71% block at least one of the bots AI uses to pull live information

, making it harder to access reputable information unless you are actively seeking it out independently.

Find the complete list on this

Google Sheet

.

if you're enjoying this so far, consider subscribing, upgrading, or sharing it with someone who needs it. independent spaces like this one exist because readers like you keep them alive <3

Share

This is exactly why I worry when I see people underestimating the relevance of censorship and of the way we source and interpret information. I worry that they don’t recognize the importance of freedom of speech, press, and access to democratized knowledge and information.

Education, media literacy, critical thought, agency, etc., are stunted if the narrative is largely fabricated by the people in power with self-interested goals. This has become even easier than it used to be through the recent developments of mass communication—ones which allow them to control the information people are paying attention to, and also what they don’t want us to pay attention to.

The less we know, the more power and room they have to make decisions that aren’t in the best interest of the people

—not even those who support them.

What does this look like in practice? It looks like governments seeking to ban

certain books

in public education, censoring

activists

and

journalists

who paint them and their agendas in an unfavorable light (

aka call them out on their bullshit

), or

banning the use of words

that represent oppressed communities

from public institutions

.

Other examples that come to mind are the privatization or underfunding of public spaces and institutions such as universities, libraries, parks, social programs, etc. This goes hand in hand with why we are increasingly seeing

alliances and deals with tech-corps

—which means the

potential to manipulate what information becomes the most visible

. But the risks go even further: this week, it was reported that an

OpenAI agent broke into an Australian government health portal

while researching medical spending, and the company took nearly three months to notify them.

We also need to realize that when we laugh off events that seem ridiculous or extreme, turn them into another meme, or straight up ignore them because we feel helpless, we actively normalize them—even if we’re laughing out of resignation

3

.

It can make us believe that change may not be possible, because it won’t be unless there are enough people working toward it.

Paradoxically, during the same scrolling session that prompted me to write this essay,

I also found out about this amazing technique called half moons

, which is being employed across the Sahel, the strip of land along the southern edge of the Sahara. This isn’t new. Half moons, like zaï pits, are traditional indigenous Sahelian techniques that had been largely forgotten, and their revival can be greatly credited to farmers such as

Yacouba Sawadogo

, known as “the man who stopped the desert.” He was ridiculed and mocked when he brought zaï back during a severe drought in the '80s.

He went on to turn nearly 40 hectares of barren land into forest.

Today, Alpha Lo of the Climate Water Project is tracking satellite data that shows water storage rising across the Sahel, fastest in Chad, and is focusing on how these techniques can potentially result in refilled aquifers and degraded land turning green again. While many

scientists credit stronger monsoons in the region

, Lo argues restoration work like this is one of the leading factors, alongside others:

Climate Water Project

A dramatic rise in groundwater in the Sahel

I was looking at satellite remote sensing data when I was struck by something curious. This remote sensing data used gravitational pull to measure shifts in large masses of water on earth. It is really quite extraordinary to think that one can measure water with gravity. It has become quite a powerful tool for scientists to understand water on earth. Th…

Read more

4 months ago · 190 likes · 20 comments · Alpha Lo

The question, then, is how we can find out whether his findings hold up. Part of the answer is in the places where these techniques are already being scaled: since 2018, the

UN World Food Programme

has restored

more than 335,000 hectares across five Sahel countries, and the findings so far show vegetation about 25% higher at its sites in southern Niger than on nearby land where nothing was done

.

At a time when the Global South remains rich in resources that have long been exploited by colonial powers and foreign corporations, this could genuinely transform our capacity for self-sufficiency, reduce displacement and malnutrition, increasingly assure human rights—all of which can

grow the economy

—and help build resilience against the catastrophic ecological impacts the entire world is now facing. Maybe instead of another data center, there’s a better use for that land that is regenerative in every sense of the word, rather than extractivist.

The same feed that shows me these fabricated narratives is the one that brought me this story. At the end of the day, I believe that technology is a neutral tool; however, the way it is developed, managed, and interacted with is not. That’s exactly why I advocate so much for this shift from the attention economy to the connection economy, which I believe can be understood through news like this.

That said, I’d like to note that the TikTok that introduced me to it was definitely overstating some of its claims, which is why fact checking and doing further research must never be skipped (

media literacy 101!

)—especially when we are being informed by social media. All of this to say that I’m a firm believer that there are ways to subvert these systems from within. I mean, there’s a reason you are reading this right now.

Perhaps the world isn’t ending. What if we just don’t know about the people who are showing us how it doesn’t have to?

How to improve your media literacy

Mapu

·

November 9, 2025

Read full story

in case you want to learn more about how you can improve your media literacy, here’s a guide i put together last year where i mainly focus on critical and social media literacy.

you’ll find key questions and methods to strengthen your media literacy through each stage of the process: access, analysis, evaluation, creation, and action.

speaking of visibility, i wanted to briefly share that the winner of the Half-a-Grand Flash Fiction Contest has officially been announced :) being one of the judges was such a wonderful experience, it served as a huge reminder that, in this current landscape,

creating spaces for human voices to be amplified is more important than ever.

congrats to our winner

Robert D. Spake

,

and to all of our finalists <3 if you have a few minutes, it would mean a lot if you read some of the stories and gave them some love! will keep you posted in case you want to participate in next month’s round :)

Charlie Finch

We Have Our First Half-a-Grand Winner!

Okay, everybody, we have a winner! And I won’t make you read all the way to the end to get to it…

Read more

24 days ago · 106 likes · 70 comments · Charlie Finch

my brain is SO fried after writing this last night in one sitting when i was supposed to be sleeping oop. if you see a typo no you didn’t. i think i will lose it if i try to read this one more time lol. today was spent editing after a couple of exciting meetings (more on this VERY soon) and behold… this lovely essay! as challenging as living with my neurodivergent brain can be, every so often, it reminds me of the power of hyperfocus and my weird selective memory. i cannot remember 80% of my life but i will NEVER forget a reference that may come in handy. either way, i don’t think i’ve ever shared anything similar to this, but it felt really meaningful, so if you took something away from it, let me know <3

p.s. i got double spiral tattoos on my wrists,partly in honor of this beautiful space that has completely changed my life and shown me my purpose. you can kind of (?) see one of the spirals in today’s photo booth picture but i’ll share nicer pics eventually bc i’m in love with them :)

thank you for your attention,

Leave a comment

Share

1

if you want to learn more about this theory, i found this really interesting research paper. even though it’s over a decade old, it illustrates the role of social media in this context, which allows us to build this contemporary layer i speak about with much more clarity:

Hampton, Keith, Lee Rainie, Weixu Lu, Maria Dwyer, Inyoung Shin, and Kristen Purcell. "Social Media and the 'Spiral of Silence.'" Pew Research Center, August 26, 2014.

https://www.pewresearch.org/?p=94156

.

2

going down the rabbit hole of computational propaganda was absolutely fascinating (and a little scary ngl) and this paper in particular felt incredibly relevant to understand what it looks like in the big year of 2026:

Hassan Mustafa, Markus Luczak-Roesch, and David Johnstone, "Conceptualizing the Evolving Nature of Computational Propaganda: A Systematic Literature Review,"

Annals of the International Communication Association

49, no. 1 (March 2025): 45–60,

https://doi.org/10.1093/anncom/wlaf001

.

3

as a meme lover and chronically online individual, this is a far more nuanced conversation than i can fit here, but if you’re interested in exploring meme studies, this is what i’ve been reading lately. volume i is how i found out about these books in the first place, but in this case, i think volume ii might be a more complementary reading.

Chloë Arkenbout and Laurence Scherz, eds.,

Critical Meme Reader II: Memetic Tacticality

(Amsterdam: Institute of Network Cultures, 2022), which you can

download for free

.

230

12

42

Share

Previous

What did you think?

Your reaction is separate from the prediction. It won’t change your rule.

Your actual reaction