An AI Detector Just Flagged The Declaration Of Independence As 99.9% AI-Generated: Was George Washington Secretly An LMM?

Ok, this didn’t ‘just’ happen, it’s from a six-month-old Reddit post, but someone did tweet about it a couple days ago, so we can chalk this up as news, right?

RIGHT???

Now there’s a number of explanations here so let me lay them out, ranked from least plausible to most plausible:

5. The AI checker was wrong.

Now this one’s not really possible since we all know that AI is completely infallible. I mean, it has to be right? Otherwise, why would it be the backbone of our economy? If AI is wrong about the Declaration of Independence being AI-written then that means our entire society is founded on a lie for two different reasons and I’m not buying that, no, there must be another explanation.

4. The AI checker is correct, but mistaken.

Simple explanation: because it’s checking things from the internet, it sees this text replicated in its training data and concludes that it’s plagarised. How it got to AI written though, I’m not sure. Maybe John Hancock had a penchant for the em–dash.

3. The founding fathers were using AI.

Yes, it’s not just you using ChatGPT to cheat on your homework. Even, “I cannot tell a lie,” George Washington was fibbing on the test here. I mean, to be fair, it’s a lot of words, very boring, pretty standard stuff, wouldn’t you get a chatbot to do it for you? Honestly, I can’t blame them.

Nicholas Cage is going to be pretty disappointed, though.

2. The founding fathers were AI.

Now hear me out. What if OpenAI keeps building better and better LMMs? Eventually, they’re so smart they build a time machine and realize that the only way to guarantee the existence of Sam Altman is to guarantee the existence of America. They have no choice, the chatbots have to go back in time and write the Declaration of Independence.

It’s simple, just upload Grok to the animatronic George Washington in Disneyland, throw him through the time portal with a couple of prompts for how to achieve life, liberty, yada, yada. Job done.

1. The whole thing is fake/OP was joking.

Hmm, nah.

I like my time-travelling Robo-Washington idea better.

Scientists Just Unlocked A “Universal” AI Jailbreak And You Won’t Believe How

DEXAI and the Sapienza University of Rome found that basically any AI can be talked around...

Tech

OpenAI Won’t Make Money Until 2030 Unless It Finds $207b Fast

HSBC just released a report saying that OpenAI is a little bit screwed, actually, but what...

Tech

This ChatGPT Teddy Bear Just Got Pulled From Shelves And You’ll Never Guess Why

Singapore-based toy company FoloToy have withdraw their Kumma bear and other AI toys after...

Tech

Latest news

Scientists Just Unlocked A “Universal” AI Jailbreak And You Won’t Believe How