What "Good Enough" Costs the Internet
Session 0.5 · ~5 min read
"Good enough" is the phrase people use when they have decided that quality is someone else's problem. It sounds pragmatic. It sounds efficient. In practice, it is the mechanism by which entire information ecosystems degrade.
When one publisher decides that AI-generated content with no fact-checking is "good enough," nothing happens. When ten thousand publishers make the same decision simultaneously, the floor drops out of the information economy.
The Trust Erosion Cycle
Trust in online content does not collapse overnight. It erodes through a predictable cycle that feeds on itself.
'Good enough' becomes the norm"] --> B["Readers encounter more
low-quality content"] B --> C["Readers develop skepticism
toward ALL online content"] C --> D["Even quality content
is trusted less"] D --> E["Publishers see declining
engagement on good content"] E --> F["Publishers reduce investment
in quality"] F --> A
Each loop makes the problem worse. As readers become more skeptical, they spend less time evaluating content and more time dismissing it. This penalizes careful, well-researched work as much as it penalizes slop, because the reader has stopped distinguishing between them.
When "good enough" becomes the standard, the floor drops. Readers stop trusting written content. Genuine expertise gets buried under confident-sounding nonsense.
What Gets Lost
The costs of the "good enough" standard are concrete, not abstract. They show up in measurable ways across the information ecosystem.
| What's Lost | How It Manifests | Who Pays |
|---|---|---|
| Specificity | Generic advice replaces detailed, experience-based guidance | Readers who need real answers |
| Accountability | No author, no editor, no one standing behind the claims | Everyone who acts on bad information |
| Institutional memory | Hard-won domain knowledge gets buried under AI-generated rewrites | Entire professions and industries |
| Economic viability of writing | Commodity pricing eliminates the middle tier of professional writing | Writers, editors, publishers |
| Reader attention | Content fatigue drives readers away from text entirely | Anyone who communicates through writing |
The Institutional Memory Problem
Some of the most valuable content on the internet was written by practitioners who documented their hard-won knowledge in blog posts, forum threads, and niche publications. A plumber who wrote about the specific challenges of replumbing Victorian-era houses. An engineer who documented the edge cases in a particular database configuration. A teacher who shared lesson plans that actually worked with difficult students.
This content was never optimized for SEO. It was written because someone knew something useful and decided to share it. It ranked because nothing else addressed the same questions with the same specificity.
AI content farms target exactly these queries. They generate surface-level articles on the same topics, optimized for the same keywords. The original practitioner content gets pushed down in search results or drowned out entirely. The institutional memory of the internet, built up over two decades of people sharing what they know, is being overwritten by text that knows nothing.
The Reader's Response
Readers are not passive victims of this process. They adapt. The adaptations are not good for anyone who publishes content.
AI slop repeatedly"] --> B["Develops pattern recognition
for AI content"] B --> C["Starts dismissing content
that 'looks AI'"] C --> D["Dismissal extends to
legitimate content"] D --> E["Reader shifts to
trusted sources only"] E --> F["New voices and
small publishers get
locked out"]
The most concerning adaptation is the shift to "trusted sources only." When readers lose faith in open-web content, they retreat to known brands, paywalled publications, and personal recommendations. This is rational behavior, but it concentrates attention on a smaller number of publishers and makes it harder for new voices, no matter how expert, to build an audience.
The Difference Between Good Enough and Actually Good
"Good enough" and "actually good" are not points on a spectrum. They are different standards with different outcomes. Good enough asks: will this pass? Actually good asks: will this serve the reader?
| Standard | Good Enough | Actually Good |
|---|---|---|
| Question | "Will this rank?" | "Will this help someone?" |
| Process | Generate, skim, publish | Research, draft, review, revise, publish |
| Evidence | "Studies show..." (no citation) | Specific source, year, finding, and context |
| Voice | Generic "helpful assistant" | Identifiable author with a perspective |
| Shelf life | Until the next algorithm update | As long as the information remains accurate |
This course is not about avoiding AI. It is about refusing to accept "good enough" as a standard. The sessions that follow will build the systems, workflows, and quality controls that separate production-grade content from slop. Starting with Module 1, we stop diagnosing the problem and start dissecting the mechanics of what makes AI content bad, so you can learn to make it good.
Further Reading
- The Internet Is About to Get a Lot Worse (The Atlantic)
- The Age of AI Slop (The New Yorker)
- The Dead Internet Theory Is Becoming Real (New York Magazine)
- Dead Internet Theory (Wikipedia)
Assignment
- Write a 500-word essay, entirely by hand (no AI), on what "good enough" means in your field. What is the difference between good enough and actually good? What gets lost when the standard drops?
- Save this essay. It is your baseline writing sample. You will compare it to your work at the end of this course to measure your own growth.
- Be honest. If your own content has been "good enough" rather than "actually good," say so. This is a diagnostic, not a performance.