People have been talking about "Deepfake" pictures and video for years now - artificially created media based on publicly available source media that shows something entirely made up. ChatGPT adds the ability for AI to write documents on arbitrary subjects.
The problem for people passing these off is that the fonts used to print them don't look handwritten. They're too perfect, to regular. Real people have subtle variations in their handwriting so you can catch this sort of thing out.
Like other machines, this starts with asking ChatGPT to write an essay based on the assignment prompt. That generates a chunk of text, which would normally be stylized with a script-style font and then output as g-code for a pen plotter. But instead, Devadeth created custom software that records examples of the user's own handwriting. The software then uses that as a font, with small random variations, to create a document image that looks like it was actually handwritten.
It uses a pen plotter to write the document, so it could be written in a spiral notebook. The future is stupid, and entirely untrustworthy.
How would the world be different if the idiot that made the fake Dan Rather George Bush TANG memo had just used the correct font????
The future is stupid and will be lead by liars and cheaters......
Wow... wouldn't it just be easier to actually DO the work?
OldNFO, I had a college professor who told us that if we cheated, we would only be cheating ourselves. He made a point to detail what the current cost per credit hour we were paying was.
Just one more reason not to believe anything you didn't witness with your own eyes.
Post a Comment