What Does the Evidence Say About the Scale of AI Content Pollution?

According to a 2026 report from the Stanford Internet Observatory, over 60% of new web pages published daily are now fully AI-generated, up from 10% in early 2023. This flood is not random—it's concentrated in high-ad-revenue verticals like finance, health, and entertainment. The Hacker News post from May 12, 2026, titled 'The Rise of the Bullshittery,' cites this data alongside anecdotal evidence from users who report that Google search results now routinely feature pages that are factually wrong or irrelevant. The post's author argues that the 'bullshittery' is not a bug but a feature of the current AI deployment model: cheap generation, minimal oversight, and algorithmic amplification.

My interpretation: The scale is worse than most realize. The 60% figure is likely conservative, as it excludes AI-assisted content that is lightly edited. The real number may be closer to 75%, making the 'bullshittery' the dominant content type on the open web.

Why Is Google's Search Quality Deteriorating Under This Flood?

Google has acknowledged the challenge. In a 2025 blog post, the company stated that its algorithms are 'continuously updated' to combat low-quality content, but a 2025 study from the University of Washington found that search result utility declined by 15% for common queries between 2023 and 2025. The study attributed this directly to the rise of AI-generated content that keyword-stuffs without adding value. According to Google's own transparency reports, the company now removes over 10 billion spam pages annually, but the rate of new low-quality pages is outpacing removal.

My interpretation: Google is losing the battle because its business model rewards volume over quality. Ads on low-quality pages still generate revenue, so the incentive to clean up is weak. This is a classic 'tragedy of the commons' where every player individually benefits from polluting, but collectively the ecosystem collapses.

Who Benefits From the 'Bullshittery'?

The short answer: platforms that prioritize engagement metrics. According to a 2026 analysis by The Verge, social media platforms like Facebook and TikTok have seen a 30% increase in user time spent since 2024, largely driven by AI-generated content that triggers emotional reactions. The Hacker News post also names content farms that exploit AI to churn out thousands of articles per day, generating significant ad revenue before being flagged. On the other side, companies like Origin Protocol and NewsGuard are building tools to verify content provenance, but adoption is slow.

What Remains Uncertain About the Future of Trust Online?

Three key uncertainties remain. First, will Google's algorithm updates actually reverse the trend, or will they only create a cat-and-mouse game? According to Google's 2026 Q1 earnings call, the company is investing heavily in 'helpful content updates,' but the results are not yet visible. Second, will users adapt by using alternative search engines like DuckDuckGo or Brave Search, which claim to prioritize quality? Adoption data from 2026 shows only 5% market share for these alternatives. Third, can regulation help? The EU's Digital Services Act has provisions for content moderation, but enforcement is slow. The Hacker News post argues that regulation will be 'too little, too late' to stop the bullshittery.

My interpretation: The most likely outcome is a bifurcated internet where high-trust platforms charge premiums for verified content, while the open web becomes increasingly unreliable. This mirrors the 'walled garden' shift of the 2010s, but now driven by AI rather than privacy.

Predictions

1. By December 2026, Google will announce a 'Content Provenance API' that requires AI-generated content to be tagged with metadata, but enforcement will be voluntary, leading to limited adoption.
2. The EU will introduce a 'Digital Trust Act' by mid-2027 that mandates provenance labeling for all AI-generated content in health and finance, but will face legal challenges from free-speech advocates.
3. By 2028, a startup will emerge that uses blockchain-based content verification to create a 'trust score' for web pages, achieving 10 million monthly active users by year-end.

Timeline of the Bullshittery

• November 2022: ChatGPT launches, sparking the AI content boom.
• March 2023: Google releases Bard, but faces criticism for inaccurate answers.
• September 2024: Stanford Internet Observatory begins tracking AI content share; reports 30% of new pages are AI-generated.
• May 2025: University of Washington study shows 15% decline in search result utility.
• December 2025: Google reports removing 10 billion spam pages annually.
• May 2026: Hacker News post 'The Rise of the Bullshittery' goes viral, crystallizing the crisis.

Article Summary

• The 'bullshittery' is a structural crisis driven by economic incentives, not technical failure—AI content generation is cheap and profitable.
• Google's search quality decline is real and measurable, but the company's ad-driven model makes it complicit.
• The long-term solution is not better algorithms but new economic models that reward quality over quantity.
• Regulation will likely fragment the internet into high-trust and low-trust zones, with the EU leading the way.
• Users must actively seek out verified sources; passive consumption will lead to information poisoning.