Simone and Malcolm Collins expose how viewbotting, clip spamming, and manufactured engagement are completely warping our perception of what's popular online. From Twitch streamers (80% of top creators allegedly botted) to music giants like Beyoncé losing billions of fake views, "woke" games with 200 peak players, Substack subscriber farms, and Kick's massive clip-spamming campaigns — the internet is far faker than most realize.We break down the economics (why botting is a rational business decision), real-world examples (Ben Shapiro, Steven Crowder, Caleb Hammer, Clavicular), how algorithms get gamed, and what this means for discovering authentic content in 2026.Dead Internet Theory just got an upgrade.

Show Notes

* According to some analysts, for the first time in over a decade, bots now generate the majority of internet activity

* At 51-53%

* This is according to multiple reports and sources (see note at the end)

* Note: Breakdowns often separate “good” bots (search engine crawlers, SEO tools) from “bad” ones (malicious scrapers, credential stuffers, ad fraud). Imperva notes bad bots alone rose to ~40% of total traffic in 2025 (up from 37%)

* BTW: Cloudflare’s data (which focuses on HTTP requests they observe) shows a lower but still rising bot share—around 31–32% in Q1 2026 (up month-over-month)—with AI crawlers as the fastest-growing segment. Their CEO has publicly predicted bot traffic will exceed human traffic by 2027, aligning with the broader trend. Some analyses of Cloudflare data cite >50% of HTML page requests as bot-driven in 2025

* There are literal view farms (this is one Brazilian one that was raided two months ago, in March 2026:

* For any platform you can imagine, you can buy viewbots with varying degrees of sophistication, including viewbots that have widely varied IP addresses that have detailed histories, leave comments, mute/unmute while watching streams, etc.

Fame is manufactured

* Major music labels and artists are using botting to look bigger than they are

* An example: Drake accused his own label (UMG) of conspiring with third parties (including Spotify) to bot streams for Kendrick Lamar’s “Not Like Us” to harm him. UMG called it “untrue” and “illogical.” Defamation claims were dismissed; the broader case is ongoing. Drake has also faced separate accusations of using his Stake partnership to fund botting for his own catalog.

* When major companies DON’T use viewbotting, you see embarrassing situations like the pilot episode of Star Trek: Starfleet Academy, which got ~16,000 views in its first 11 hours after release on YouTube. A separate report also said the live premiere peaked at roughly 1,300 concurrent viewers.

* Even major viral figures, like Caleb Hammer and Clavicular, are manufactured to a great extent

Let’s explore just how bad it is

Viewbotting on Twitch

* Around 10% of Twitch streamers with at least 50 average viewers show clear, persistent signs of viewbotting, according to the most comprehensive independent analysis available (Streams Charts / Audiencly 2025 whitepaper, covering Q2 2025 data)

* It’s worse for big creators: Streamer/analyst Devin Nash (and his agency) analyzed the top 500 Twitch streamers and estimated 400–430 (roughly 80%) show signs of viewbotting or being botted (30–40% of viewers as blatant bots + another 5–15% via embeds).

* This is based on chat activity monitoring, user-list sampling, logged-in/out ratios, and known botnet cross-referencing

* Creators argue Twitch is a platform where viewbotting ia necessary for survival; if you’re not doing it, you’re not competitive

* Doesn’t help that discoverability is very low

Devon Nash on the Unit Economics

In a recent video, Devon Nash, a professional on the brand marketing side of the equation (he’s Chief Marketing Officer at Novo), explained how viewbotting is a no-brainer smart decision for streamers and agencies based on the unit economics:

* Viewbots cost approximately $0.01 to $0.02 per viewer hour, which translates to about $135 to $185 per week to add 500-750 viewers to a stream. This weekly cost includes features like chatting and custom chat messages to make the viewers appear authentic. For a full month of viewbotting, agencies spend less than $800 to artificially inflate viewer counts.

* Twitch sponsorship rates typically range from $1 to $3 per concurrent viewer (CCV), with $1.50 to $2 being the standard rate for a 2-hour gaming sponsorship. For a streamer with 1,000 viewers at $2 per CCV, a single 2-hour sponsorship generates $4,000 in total revenue. The agency typically takes 20% commission, earning $800 per deal, while the streamer receives $3,200.

* Nash demonstrates how agencies can achieve massive returns by combining viewbotting with multiple sponsorship deals.

* Starting with a 300-viewer stream and adding 700 botted viewers creates an apparent 1,000-viewer stream for approximately $150-180 per week. If the agency secures just two 2-hour sponsorships for that inflated audience, they earn $1,600 in commission while spending less than $400 on viewbots.

* This creates what Nash calls “a money printing machine” where agencies multiply their investment several times over.

Viewbotting on Substack

There are websites that sell Substack subscribers (as low as ~$0.02 each), sometimes claiming they use “real people” added manually rather than pure bots. Whether these are organic or farmed/incentivized accounts, they still represent artificial inflation

* here’s one: you can buy low, medium, and high quality subscribers).

* You can also buy comments, likes, views, shares, plays, restacks, searches, comment likes, comment restack, comment shares, aves, messages, comment replies, and save as image

In April, the Observer covered how Andrew Tate’s Substack saw its total follower count drop from 1.1 million to 980,000 after analysis of a sample of 1,000 paying subscribers found that 75% had no biography, publications, or visible activity—and half were created in a 16-day window. Investigators concluded he had imported a pre-existing (likely harvested) email list. Substack’s standards and enforcement team reviews bulk email imports and acts when they appear illegitimate.

Earlier, creator Rebekah Jones lost hundreds of thousands of subscribers in apparent purges (documented on X in 2025), with charts showing dramatic drops after bulk fake additions.

Viewbotting on YouTube

Fake views have existed since at least 2009, with media attention by 2011 and a major 2012 purge in which YouTube removed billions of fraudulent views, including over 1 billion from Universal Music Group artists (e.g., Beyoncé, Rihanna, Justin Bieber, Nicki Minaj).

Physical view farms continue operating globally in 2026. In March 2026, Brazilian police raided a large-scale YouTube view farm with dozens (or hundreds) of smartphones rigged to ceilings, running 24/7 to loop videos and simulate views/interactions. (IT LOOKS CRAZY) Similar operations have been documented in Vietnam and elsewhere, often targeting music videos or algorithm gaming

A 2024 academic study analyzing nearly 100,000 YouTube videos from over 1,000 French channels over 1.5 years found fake view removals (“corrections”) on ~90% of channels and 78.5% of videos. These corrections occur in daily batches (often around 5 p.m.) and frequently happen late in a video’s lifecycle—after most organic views have accumulated—rather than in real time. Notably, videos corrected later tended to be more popular overall, suggesting fake views can temporarily boost algorithmic recommendations and perceived popularity before being stripped.

Clip Spamming

Devon Nash also changed how I view YouTube discovery with his breakdown on how people—including Clavicular and Caleb Hammer—are manufacturing virality by spamming clips of their work on platforms like YouTube (see: Exposing the New Manufactured Viral Content Industry)

The video explains how a paid “clipping economy” is artificially hijacking short‑form algorithms on YouTube, TikTok, Instagram, and X to manufacture viral influencers and promote the streaming platform Kick,

How the clipping system works

* Campaigns run inside large Discord servers (20–30k+ people) or invite‑only groups where each campaign corresponds to one streamer, podcaster, or brand.

* Clippers pull 30–120 second segments from long‑form streams and upload them as shorts on TikTok, Instagram Reels, YouTube Shorts, and sometimes X, under their own accounts.

* They are paid on a flat CPM basis, typically around 0.10–0.40 dollars per 1,000 views but sometimes up to 2–3 dollars or specific bounties like 3,000 dollars per million views for particular clips.

* Payment usually happens in USDT and often only once a minimum aggregate view threshold (for example 100,000 total views across all the clippers’ uploads) is reached, incentivizing people to spam 50–100 clips across multiple accounts.

Because the CPM is on top of the platform’s own ad revenue...