Raw LLM Responses

I don't think you guys did a very good job covering this. One elephant in the room here is the fact that these tools have a 90% success rate. That means 9/10 times, it is accurate. But not 100%. And that means that, much like lie detector tests, it should not be relied on as evidence. But that doesn't mean it shouldn't be used. It is still a useful tool for the police to know where to look. 9/10 times it points them in the right direction. As for how companies use it for housing, employment, etc., which you might know if you had bothered to look into it instead of screeching about race, it's used to IDENTIFY people. It's used in situations where, for example, you are doing everything over the internet instead of meeting in person. In person, you'd hand someone your state ID, they would look at it, look at you, and know that it's you. Instead, when you do it online, the system looks at your ID pic, looks at a live picture of you, and compares the two. ID verified. No need to go down to the office with your ID right then and there. Which is good if your employer or housing company exists in another state, which makes it more difficult for you to go down there. Now, of course a human could do this too with a live feed or something like that, but automating it makes it faster and reduces workload, and in 9/10 cases it gets it right. The 10% it doesn't can be reviewed by a human. That's a 90% reduction in the tedious wotkload of ID verifications. Quit turning everything into a racism. Yeah, the tech isn't as good at identifying black people. It's good at identifying most people. Should we just not use a useful tech because one group doesn't get as much benefit from it? That's stupid. How about instead we work on improving the technology so it does better at identifying black folks?