There is no ethical use of "AI".
0 Boosts
0 Likes
@Stefan_S_from_H Of course, that statement itself is an absolute.
I agree many uses fail the ethical test. However, how about pattern recognition in spotting cancer scans rather better, faster and possibly cheaper than humans?
@stuart @Gargron @Stefan_S_from_H given that most LLMs steal from all kinds of data sources for training purposes, usually with the primary goal of generating profit, this case falls under a good use but not truly ethical. To be ethical imho, the underlying model should use proprietary, public, or consented data only.
@chibog @Gargron @Stefan_S_from_H
But in this case the training data is coming from other scans. Who would argue that anyone would object to such a use provided the collected data was not proprietarilly controlled?
@stuart @chibog @Stefan_S_from_H If you have a neural network trained on a specific set of images, you have an image classifier, not "AI".
@Gargron @stuart @chibog @Stefan_S_from_H If you have machine learning, you have AI. AI is widely more diverse then people think. Actually, most of AI methods do not involve even learning and neural networks, like methods from my field.
So, writing that there is no ethical use of AI is very, very wrong. But I guess what you mean is "there is no ethical use of LLM". Which I also disagree, despite not liking LLMs.
@Gargron @stuart @chibog "Artificial intelligence (AI) is the capability of computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making." https://en.wikipedia.org/wiki/Artificial_intelligence
@Gargron only a Sith deals in absolutes.