How much activity on social media is actually real?
How much is bots?
How much is click farms?
And at what point does the distinction stop mattering?
Readers with long memories might remember a slightly unorthodox survey I ran a while back: counting PornBots across platforms. The logic was simple—PornBots are easy to spot, so they’re a useful proxy for non-authentic users.
The conclusion then was blunt: X (Twitter) had more PornBots than the rest put together.
I’ve repeated the exercise this year.
The result? Nothing has improved.
X still has the highest proportion by some distance. Roughly a third of the accounts I encounter are obvious PornBots. Before anyone gets clever—no, this is not a reflection of my browsing habits. The platform is simply flooded with them.
The Problem With Measuring Bots
There is no agreed figure for “how many bots are on social media.”
That’s not because nobody’s looked. It’s because the question itself is slippery.
- Platforms define “bots” differently (spam vs automation vs fake humans)
- Detection is imperfect (the best bots look convincingly human)
- Activity ≠ accounts (a small number of bots can generate a huge amount of content)
So instead of answers, you get ranges.
The most studied platform is X:
Accounts
- ~5% (platform’s own estimate)
- ~9–15% (typical academic range)
- 25–68% (during specific events or topics)
My own experience? Closer to 50–60%.
But accounts are the wrong place to focus.
Accounts Don’t Matter. Output Does.
Even if bots are a minority of users, they can dominate what you actually see.
Content / activity
- Bots may generate 20%+ of discussion globally
- Up to 66% of shared links in one major study
- On specific topics: often 50%+ of posts
So a platform might be:
- 10% bots
- But 40–60% of visible content shaped by them
Which is what people actually experience.
How It Actually Feels: Immigration
Take immigration—one of the most consistently “hot” topics online.
A typical pattern looks like this:
- A real news story breaks (e.g. asylum numbers, a crime story, or a policy change)
- Within minutes, dozens of near-identical posts appear
- Accounts with minimal history push the same framing:
- “Country is full”
- “Government betrayal”
- “Media cover-up”
- These posts heavily retweet each other, creating the illusion of momentum
- Real users then pile in, responding to what looks like a genuine surge of opinion
The key point: the initial wave is artificial.
Bots don’t need to win the argument.
They just need to set the tone.
And because outrage travels faster than nuance, the tone they set is almost always the same. And it creates the impression that most people agree with right wing political positions, when they don’t.
Enter the Click Farms
Bots are only half the story.
Click farms are worse in some ways—because they’re human.
Low-paid workers, often in Asia, running large numbers of accounts:
- Posting at scale
- Using AI-generated images and text
- Targeting specific demographics (older users, political groups)
If you join almost any political Facebook group, you’ll see it:
- Highly emotional posts
- Dubious images
- Content that feels slightly “off” but travels anyway
A lot of it originates far from the countries it claims to care about.
So What’s the Real Number?
If you force an honest estimate:
Across major platforms:
- Accounts: ~5–15% non-human (baseline)
- Content/activity: ~20–40% bot-influenced
- Hot topics: often 50%+
And those “hot topics” are exactly the ones that shape politics.
Are We in the Dead Internet?
The Dead Internet Theory claims the internet effectively died around 2016–17—replaced by bots, AI content, and algorithmic manipulation.
That’s overstated.
But it’s not entirely wrong either.
We’re not in a fully synthetic internet.
But we are in a world where large parts of online debate—especially on issues like immigration—are artificially amplified.
And crucially:
- The amplification is not neutral
- It skews heavily in one direction
- It rewards outrage over accuracy
The Real Problem
The internet isn’t dead.
But honest debate is under serious pressure.
Not because people disagree.
But because the conversation itself is being shaped—subtly, constantly—by actors who aren’t really part of it.
And most users don’t notice.
They just feel that something is off.
They’re right.
Post Script.
For those interested these are the AI PornBot names:
Leila Onyx
Priya Desire
Liana Desire
Emmeline Sin
Abhinav Hands
Aline
Inna Luxe
Kira Adict
Alexa Kiss Trap
Chloe Val
Andrea Luna
Eva Doll
Amelia Hart
Ioanna AfterDArk
Nicol Rivers
Sofia Lee
Sara Hayes
Madie Roslyn
Ta
Mollyyyy Clark
Luluuu
Linda Washington
Kira Control
Alinia
Alinna
Luna Amore
Tessy
Amelia Bloom
Lily Hexwlitz
Elizza Kitten
Kevin Dutton
Ionna Noir
Sara Hayes
Madie Roslyn
Molly Clark
Mollyy Clark
Tessa Banana
Luluuuu
Matta Rife
Linda Washington
Werna Katherine
Gwendolyn Yeates
Joycer Olive
Joy Hume
Not long ago X implemented a feature that showed the country of origin for accounts, but soon disabled it.
If you’re cynical -and I am- you could say the feature was disabled because many right‑wing commenters and thread starters on topics like immigration turned out to be from countries such as India, Africa, and Asia.
The Peak of the pub trade was early 1900’s I think there was around 80,000, but with half the population we have now.
Of course things were different back then, beer was actually safer to drink than water, it was cheap, there was no TV or Radio so people got their entertainment and socialising from pubs.
Since then we’ve had radio and TV, cars to get to different places, bans on drink driving and now the internet and the stuff that provides plus the reasons you mentioned. They are all reasons to stay in.
However it doesn’t all make sense. Why has it reduced so much from the 90’s when I first started turning out? I reject the idea that cheap supermarket booze is to blame, booze from shops has always been cheap. Up until the 2000’s maybe even 2010’s you could buy 8 cans for fiver, but it never stopped people going out. The pint to can price ratio now is arguably worse than it was 90s-10’s
I think the cost of living is destroying local pubs more than the cost of a pint. 1995 I used to work away from home, when I returned I bought my first pint in my hometown of Seaham, it was in the Duke, and it cost me £1.45, or £3.01 in todays money. You can still get a pint £3 in Seaham – happy hours and such, and some places in Sunderland sell pints that low. Granted you can pay up to a 5er for a pint, but you don’t need to.
When people look at their income they are being forced to make choices, a night out is a luxury, and luxuries are the first on the chopping block when finances are tight. Increasing rents, mortgage rates, food and energy are eating the working class alive, combine with things you never used to have to pay for like mobiles, internet and streaming services. You also have more health conscious younger people who maybe still live with parents but are working and have a disposable income, but spend it elsewhere.
Its a real shame because despite alcohols drawbacks, its a stress relief, letting lose on the weekends is what I lived for, memories with mates we still talk about, but also meeting new people. My dad used to be a bus driver we drove through Durham a few weeks ago and for nostalgia sake he went one of his old bus route driving through the likes of Murton, Easington Lane and Shotton, he pointed out the places where there used to be pubs, pretty much every single location is now a house or shop.
You may have guessed but this was meant for https://jon-chadwick.com/2026/02/26/why-pubs-are-closing-uk/
Dont know how that happened but I’ve had a nightmare trying to post it.