Post #4203468

RE: https://mastodon.social/@gvwilson/116997366488785600 If you accept something is unethical, its usefulness is irrelevant, isn't it? Unrestrained medical testing on humans would be useful* -- but we oppose that because it's wildly unethical. We may have learned more from the Tuskegee Syphilis Study than we could have any other way; you can argue that it was "useful" but it was not ethical. Nobody should have to weigh its usefulness in saying, "We should not conduct medical studies on unsuspecting humans." That is, of course, an extreme example. But it's ridiculous to complain that people who find something unethical won't consider the usefulness of the thing. This is just trying to pave the way for sweeping the ethical considerations under the rug because - if we stop to actually consider the ethical implications - then the conclusion might be, "We should not use this." And that's why society is the way it is today: if something serves capitalism, the ethics have to be swept aside. I've said this before, but if the roles were reversed and AI usage benefitted workers -- if the driver of LLM usage was bottom-up and people were using it to do the same amount of work in half the time and then futzing off the rest of the day to play with their kids or pursue hobbies while collecting the same pay -- suddenly business would care *a great deal* about the ethics of using these tools to "steal" time from their employers. But because all of the negative impact is on workers, communities, and even other companies not involved in the whole AI push, we dare not consider the ethics - only the utility of these things - lest we decide, "Let's stuff that genie back in the bottle and throw it out to sea." * In the long run, though perhaps not for those particular humans.