For voting today, I tried using AI. I wonder if this is a good approach. Some candidates in today’s California primary election are quite obscure, and finding good information about them can be tricky. So I used AI for some of my selections.
I always (almost) vote, but in the past when I’ve had difficulty finding good information about some of the more obscure candidates I’d leave that choice empty.
It seemed to work quite well today. Thoughts about this approach?
Done!
To be clear, you used AI to find information about the candidates? That’s a very different thing than using it for voting itself.
The problem with using AI to get any sort of information is that you never know how accurate the information is. This is especially true when, as is often the case in politics, there are people on the Web deliberately spreading false information. You say it “seemed to work quite well”: On what do you base that?
I think it’s a very good use of AI, honestly. Tell it your thoughts about various issue and then ask it who you should vote for.
I’ve wondered if eventually we could get to a system where we each have an AI proxy that votes for us based on what we tell it about ourselves and our priorities. A representative democracy where the representative is an AI avatar of ourselves.
Yes. Thanks for the clarification.
I prefer not to outsource my decision-making to a chatbot that boils lakes and makes stuff up. I wouldn’t do such a thing myself.
I went to Claude.ai and put in this prompt:
“Provide links to pro con newspaper editorials on June 2, 2026 California local initiatives.”
It seemed to do a reasonable job.
If more than one or two of the links are hallucinations, I will be surprised.
Of course I would not ask the chat box what it thinks. I use them as smarter search engines.
Since the candidates are obscure, how do you know it worked well? Did you verify the information it gave you?
My apologies for my not-well-described, lacking OP and any resultant confusion!
To be more clear, I used AI to scrape the web to generate information about some of the candidates for whom not much information is readily available. Or can be difficult to find. I did not use it to do my voting for me.
AI was able to provide some information, but I then had to cross check that information, and also determine the information sources.
This approach was much more efficient than doing the searches myself.
Based on the information produced by AI, I was able to more efficiently gather the information that helped me decide who to vote for.
Yes, you have to! AI is a good start on getting some information but then that information needs to be vetted.
This is what I intended to say in my OP. AI is able to find information that is out there, but then as I explained in this post (and failed to say in my OP), that information needs to be checked and cross checked.
It was able to give me some initial information much more quickly than if I’d try searching for that info myself — which I’d done in many prior elections.
For some candidates and some races, not much was able to be found on them and so I did not vote for anyone running. Even if that person was running unopposed. But whereas in the past my ballot might have 30-40% of the races without a vote from me, this time there were only 3 races where I did not make a vote. As @PhillyGuy put it, the AI was a smarter search engine for me.