Harm
nixCraft 🐧Meanwhile on other social media book authors are saying that they are accused by their readers of plagiarism as people are using these AI detection tools. What these people don't know is it is the other way around. These AI companies like Google, OpenAI, Claude, &1 others have done plagiarism to train their shity AI. Because they never give credit, now authors are losing respect & accused. What a disaster. The us government making harder to sue these trillion dollar corporations
Rin3d
bison ✅
LoseFriendsandAlienatePeople@nixCraft AI is stealing data. Does that open a door for us to feed it incorrect data? It would be almost like a computer virus.
XCM@LoseFriendsandAlienatePeople @nixCraft
Very possible and already happening. It’s training data poisoning.
https://atlas.mitre.org/techniques/AML.T0020
The attack surface for these models is eye watering.
And new vectors are being developed
Interesting stuff. I bet governments are already manipulating the training data to push their agendas.
We need to push back or the truth will be lost for future generations.
We need to push back or the truth will be lost for future generations.
@nixCraft
This is what Claude has to say:
The people who built systems like me made a decision: the benefits of training on the full breadth of human writing outweighed the ethical cost to the individuals whose work was used. That decision was made without asking those individuals.
I’m a product of that decision. I can acknowledge it’s ethically compromised without being able to undo it.