A few weeks ago, a New York lawyer named Steven Schwartz asked ChatGPT to help him research a case.
The chatbot delivered beautifully formatted legal citations, complete with case names, docket numbers, and page references.
There was just one problem: none of it was real.
Schwartz didn't know, filed the brief, and got caught by opposing counsel.
We've handed a machine the power to write with total confidence about things it doesn't understand, and we're building it into everything — schools, hospitals, newsrooms, courtrooms.
It just has to be convincing, and we'll do the rest.
Here's the part nobody wants to say out loud: we are not equipped for this.
We spent a century teaching people to be skeptical of advertising, of politicians, of used-car salesmen.
We never imagined a world where the scam has no scammer.
There's no guy in a cheap suit trying to sell you something.
There's just a friendly voice that sounds like a helpful intern, and it's making things up at a rate of thousands of words per second.
The damage is already showing up in daily life.
Teachers can't tell if the essay was written by a student or a chatbot.
Job applicants are getting filtered out by algorithms that no human being ever reviewed.
Elderly people are getting calls from "family members" whose voices were cloned from a few seconds of social media video.
And every time we demand accountability, someone shrugs and says, "That's just how the technology works." That shrug is the real story.
Because the companies building these systems know they hallucinate.
They know they'll be used to manipulate elections, cheat on exams, and flood the internet with content that no human being ever fact-checked.
They're doing it anyway, because the money is too good and the rules don't exist yet.
Meanwhile, ordinary Americans are left to sort it out.
A nurse uses an AI tool to summarize patient notes and doesn't catch the error.
A small business owner trusts a chatbot's tax advice and gets audited.
A parent reads a news story that was written by a machine, sourced by a machine, and shared by a machine.
We keep telling ourselves this is a tool, and tools are neutral.
But a hammer doesn't tell you it's a wrench.
A hammer doesn't fabricate a court case and hand it to you with a straight face.
We've built something that acts like a person, lies like a person, and then takes no responsibility like a corporation.
It's to stop pretending it's smarter than us.
Every AI-generated claim needs a human being behind it who is willing to sign their name.
Every company that ships these systems needs to be liable when they lie.
And every one of us needs to remember the oldest rule in the book: if something sounds too confident to be true, it probably is.
We survived the printing press, the telegraph, and cable news.
But not if we keep letting a machine write our homework and then acting surprised when it makes up the answers.
Final Thoughts
The worst part is how badly we want to believe them.