No. Especially for local elections I believe that both the search results and any AI summarization run a risk of mistaken identity. There often aren't enough results for a person in a local race and can pick up people with a similar name.
Yes I have. Typically the websites and news media ask questions I do not care about. So I ask it MY questions. Like what is their profession, how long in politics, stance on freedom of speech, environment, economy, taxes, defense. This yielded a much better result than scanning sites. Precise answers. I did NOT ask it to rate the races. For judges retention, normal information is limited, I asked why or why not from both democratic and republican perspective. Likewise special measures. Reduced my research from a day to just an hour or so.
Even that use of "AI" is sus: when I have Claude write some code, it has goofy problems a lot of the time. Bit-fields off by 2, missing field structs, public enums and variables that aren't and shouldn't be used publicly. Why is use of "AI" in other contexts not as error prone? I stand by my answer. I'm not even going to offload my mistake-making to a program that gives wrong information every so often.
When researching candidates, is it more or less accurate than the media and ads? Is a mistake or two in a larger blurb about a candidate going to move the needle? Do your human written resources have more or less errors?
Many people are using Ai for deep research and finding it one of the most useful use-cases. You check your code, why do you assume all people don't look at the citations that come back with their report?
I agree eyes-free usage in coding or otherwise is a bad trend. But humans are going to human and I'm not sure the Ai returning information is actually worse than the major media outlets. The opposite really
No. Especially for local elections I believe that both the search results and any AI summarization run a risk of mistaken identity. There often aren't enough results for a person in a local race and can pick up people with a similar name.
Yes I have. Typically the websites and news media ask questions I do not care about. So I ask it MY questions. Like what is their profession, how long in politics, stance on freedom of speech, environment, economy, taxes, defense. This yielded a much better result than scanning sites. Precise answers. I did NOT ask it to rate the races. For judges retention, normal information is limited, I asked why or why not from both democratic and republican perspective. Likewise special measures. Reduced my research from a day to just an hour or so.
No. I think it would be unethical of me to offload decisions that important to programs that create output on a statistical basis.
you don't have to use them to give an answer, most people doing this are using them to get informed so they can make better decisions themselves
Even that use of "AI" is sus: when I have Claude write some code, it has goofy problems a lot of the time. Bit-fields off by 2, missing field structs, public enums and variables that aren't and shouldn't be used publicly. Why is use of "AI" in other contexts not as error prone? I stand by my answer. I'm not even going to offload my mistake-making to a program that gives wrong information every so often.
When researching candidates, is it more or less accurate than the media and ads? Is a mistake or two in a larger blurb about a candidate going to move the needle? Do your human written resources have more or less errors?
Many people are using Ai for deep research and finding it one of the most useful use-cases. You check your code, why do you assume all people don't look at the citations that come back with their report?
I agree eyes-free usage in coding or otherwise is a bad trend. But humans are going to human and I'm not sure the Ai returning information is actually worse than the major media outlets. The opposite really