I wonder how difficult it would be to bias a model so that it subtly corrupts election results when performing OCR.
Surely not hard but why?
How OpenElections uses LLMs
11–20 of 60 posts
Re: How OpenElections uses LLMs
#12Very interesting! Is this the state of the art for accurate OCR of tabular PDFs, or is there other work in the space to compare against?
There's lots of posts on HN for developments and companies doing OCR and Document Extraction. It's a classic CV problem but still has come a long way in the past couple years
Yeah, this is a very well-traveled road, but LLMs have made some big improvements. If you asked me (the guy who wrote the original piece linked above) what I'd use if accuracy alone was the goal, probably would be AWS Textract. But accuracy and structure? Gemini.
Re: How OpenElections uses LLMs
#13I wonder how difficult it would be to bias a model so that it subtly corrupts election results when performing OCR.
Surely not hard but why?
Easier to steal elections?
Don't have to bother with gerrymandering, or slick legal ways to arrest people for voting with the wrong documents. Or just good old fashioned intimidation, like making the polling place the police station or the ICE detention facility.
It's just a lot smoother process when you can simply write some software to manipulate the count.
Who's gonna check?
(No, seriously, Who's gonna check? Because you also need to layoff everyone in that department once you're in power.)
Re: How OpenElections uses LLMs
#14Why is the original source data not available anywhere digitally?
Since it's printed it is clearly already in a database somewhere. Why can't that just be made public too.
Seems bizarre to OCR printed documents (although I am aware of many companies doing this to parse invoices, etc.)
Re: How OpenElections uses LLMs
#15Why is the original source data not available anywhere digitally? Since it's printed it is clearly already in a database somewhere . Why can't that just be made public too. Seems bizarre to OCR printed documents (although I am aware of many companies doing this to parse invoices, etc.)
Welcome to government data.
One key problem is that the US has tens of thousands of local governments, and each of them get to solve problems in their own way.
Digital literacy of the kind that understands why releasing a CSV file is more valuable than a PDF is rare enough that most of them won't have someone with that level of thinking in a decision making role.
Re: How OpenElections uses LLMs
#16Earlier quoted context omitted.
Surely not hard but why?
Easier to steal elections? Don't have to bother with gerrymandering, or slick legal ways to arrest people for voting with the wrong documents. Or just good old fashioned intimidation, like making the polling place the police station or the ICE detention facility. It's just a lot smoother process when you can simply write some software to manipulate the count. Who's gonna check? (No, seriously, Who's gonna check? Beca…
Corrupted OCR won't help you steal elections. The result counting is a different process, with well designed checks and safeguards.
The problem is that once the counts are done and have been reported a lot of places then print those results out on paper and then scan those papers into a PDF for anyone who asks for a copy!
Re: How OpenElections uses LLMs
#17Earlier quoted context omitted.
Surely not hard but why?
Easier to steal elections? Don't have to bother with gerrymandering, or slick legal ways to arrest people for voting with the wrong documents. Or just good old fashioned intimidation, like making the polling place the police station or the ICE detention facility. It's just a lot smoother process when you can simply write some software to manipulate the count. Who's gonna check? (No, seriously, Who's gonna check? Beca…
Many jurisdictions do risk-limiting audits using the original ballots, so futzing with the results wouldn't necessarily make that easier. Also, cast vote records are public in many states - those are records of each ballot cast. So people can check.
Re: How OpenElections uses LLMs
#18Earlier quoted context omitted.
Surely not hard but why?
Easier to steal elections? Don't have to bother with gerrymandering, or slick legal ways to arrest people for voting with the wrong documents. Or just good old fashioned intimidation, like making the polling place the police station or the ICE detention facility. It's just a lot smoother process when you can simply write some software to manipulate the count. Who's gonna check? (No, seriously, Who's gonna check? Beca…
You may consider reading about risk limiting audits. https://www.voting.works/audits
Re: How OpenElections uses LLMs
#19Earlier quoted context omitted.
Easier to steal elections? Don't have to bother with gerrymandering, or slick legal ways to arrest people for voting with the wrong documents. Or just good old fashioned intimidation, like making the polling place the police station or the ICE detention facility. It's just a lot smoother process when you can simply write some software to manipulate the count. Who's gonna check? (No, seriously, Who's gonna check? Beca…
Many jurisdictions do risk-limiting audits using the original ballots, so futzing with the results wouldn't necessarily make that easier. Also, cast vote records are public in many states - those are records of each ballot cast. So people can check.
I think you mean risk limiting, right?
Re: How OpenElections uses LLMs
#20This is such an excellent example of a responsible and thorough application of vision LLMs to a gnarly data entry problem.
It’s also an excellent example on how lack of forced machine-readable format for gov publishing is a PITA.
If I was in power and wanted to continue said rule, I’d definitely discourage the adoption of any standardized formatting for election results.
Not, you know, for any nefarious purpose…but because what we’ve used forever was good enough for grandpappy, so it’s obviously good enough for us.
/cough