A part of my soul dies a little when I read this exact sentence:
“I asked ChatGPT…”
Stop right there. You don’t ask a large language model for the truth. You consult it. There is a massive difference, and failing to understand it will destroy your professional credibility.
We have reached a terrifying inflection point on the internet. We are no longer producing credible sources; we are producing content at scale. Here is the brutal reality of how information works today, and the exact strategy you need to survive it without looking like an amateur.
The Echo Chamber: Who Guards the Guardians?
Companies are cutting costs by replacing human writers with AI because it’s cheaper than paying an expert. And here is the kicker: Google doesn’t penalize AI-generated content. It only penalizes obvious spam. This means anyone with a laptop and basic SEO knowledge can flood the internet with “authoritative” garbage.
Here is the structural flaw that should terrify you: AI models are trained on scraped internet data. What happens when AI starts scraping and training on articles written by other AIs? It’s a snake eating its own tail. If that initial content wasn’t checked for errors, we are in for a world of hurt. We will see fabricated information accepted as historical fact.
Furthermore, algorithms are designed to satisfy your prompt, not to tell you the truth. If your search history implies you believe the earth is flat, the algorithm will spoon-feed you flat-earth validation. It works exactly like your TikTok feed—it gives you what brings you pleasure, not what is objectively true.
So, who will guard the guardians? You have to.
What Actually Makes a Source Credible?
Just because a human wrote something doesn’t automatically make it credible. Humans are biased, lazy, and often wrong. Credibility requires facts, verifiable data, and clear structure.
If you are doing actual research, skip the Google front page entirely. Go to legitimate databases like EBSCO. But if you have to venture out into the wild internet, you need a strict verification protocol.
The Auditor’s Toolkit: 3 Rules for Verification
1. The Domain Interrogation (WHOIS) Anyone can buy a domain today. You need to know exactly how old the source is.
- Head to whois.com/whois.
- Type in the domain (e.g., nova.edu.mk).
- Look at the registration date.
- Warning: A domain might be registered 20 years ago, but if it was bought by a content farm last year, the data is compromised. Use this as a starting point, not a guarantee.
2. The Author Background Check Information doesn’t appear out of thin air. Who wrote it? Google the writer’s name. If they don’t have a digital footprint, a professional profile showing actual expertise, or a history of publishing in that specific field, dismiss them immediately.
3. Machine vs. Machine (The AI Checker) As a teacher, I use tools like Turnitin to scan for AI in student submissions. You need to do the exact same thing for the articles you read. Run suspicious text through a free AI checker. If it pings as heavily AI-generated, you cannot use it as a primary citation. Period.
The Wikipedia Rule
Let’s talk about Wikipedia. Is it a legitimate source? No.
It is a fantastic starting point, but it is entirely community-edited. It is the ultimate example of “design by committee.” Anyone can change a page to fit their narrative. Because of this, it is an index, not a source. Use it to find the actual citations at the bottom of the page, read those original documents, and cite them instead.
Doing real research is messy, time-consuming, and hard. But the alternative is outsourcing your brain to an algorithm. Grab a physical notebook, do the actual work, and protect your credibility.
The Bottom Line: Be the Architect, Not the Tool
Here is the strategic reality: AI is a phenomenal tool for brainstorming and structuring data, but it is a terrible architect for truth. Your value in the future job market won’t be your ability to query an AI—everyone can do that. Your value will be your ability to verify, synthesize, and think critically when the machine inevitably hallucinates.
Doing real research is messy, time-consuming, and hard. But the alternative is outsourcing your brain to an algorithm. Grab a physical notebook, do the actual work, and protect your credibility.
Now, if you’ll excuse me, it’s time for another espresso.
