Quote:
Originally Posted by john70t
First there were students copying literature or buying homework or term papers off the internet.
Basic text searches sometimes caught them cheating.
But AI can take that same info, scramble and rewrite it a little, and that breaks the search.
1% same? 50% same? What overlap constitutes plagiarism?
And not just in simple education papers...a large number of artificial fake medical studies have discovered.
So much information. The review boards just can't keep up.
The data can used to justify release of new drugs on the market. This is a huge amount of data obtained from numerous sources. And the 'test subjects' can be located in foreign countries and off any known US record databases, or even exiting in real life aka fake little brown villagers. Without witnessing and verifying in-person, anything worth billions can be made up in a couple nanoseconds.
Faked data can have so many levels of fake inputs complicating investigations into whether it can be verified. And without the tedious task of actual people in charge and being held personally responsible for the outcomes -cough- international studies without a vertical human chain of command should be rejected.
Pandora's Box has been opened.
|
Drug companies have been faking and altering medical trial data for decades. This is nothing new.
__________________
‘07 Mazda RX8
Past: 911T, 911SC, Carrera, 951s, 955, 996s, 987s, 986s, 997s, BMW 5x, C36, C63, XJR, S8, Maserati Coupe, GT500, etc
|