Republican state Rep. Alex Kolodin said he used ChatGPT to write a subsection of House Bill 2394, which tackles AI-related impersonations of people by allowing Arizona residents to legally assert they are not featured in deepfake videos.
“I used it to write the part of the bill that had to do with defining what a deepfake was,” Kolodin told NBC News. “I was really struggling with the technical aspects of how to define what a deepfake was,” he said. “So I thought to myself, ‘Well, why not ask the subject matter expert, ChatGPT?’”
The bill was signed into law by Democratic Gov. Katie Hobbs on Tuesday. The legislation allows Arizona residents to obtain a court order stating the person identified in the deepfake video is not them.
Kolodin said that the portions ChatGPT created were precise.
“In fact, the portion of the bill that ChatGPT wrote was probably one of the least amended portions,” he said.
Hobbs was not aware of the portion of the legislation being authored by ChatGPT.
“I kind of wanted it to be a surprise once the bill got signed,” Kolodin said, noting that it was part of the plan.
“I didn’t really understand the technical aspects, which is why I turned to a technical service I also didn’t understand”
The fact he kept the use of an LLM, which are known to provide false information, a secret should be grounds for an immediate removal from office.
If the output was wrong and he blindly trusted it, then none of the people reviewing it noticed anything, then they should all have been fired.
Sounds more like he had it generate the text, he checked that it was correct, then other people checked it and also couldn’t find any issues with it.
Autocorrect also ducks up sometimes, that doesn’t mean you should be fired for using it, unless you’re too dumb or lazy to proofread it yourself and correct mistakes.
You assume that someone actively checked that. We know for a fact that legislators routinely don’t actually read the bills they approve, and a disturbing amount of the time issues are found after things are signed into law because they didn’t actually read the damned things in the first place.
He also claims he asked ChatGPT because he needed the definition of deepfake. So instead of going to any number of dictionaries where definitions are located… He asked an LLM AI that has been proven to provide false information, to generate an answer. And then kept that fact a secret.
The fact he admits that he purposefully kept the ChatGPT use a secret until after it was signed into law proves that he knew it was at least a shady idea. It doesn’t matter whether it actually made a difference, that shows a disturbing lack of ethics.
We have apparently evidence that he doesn’t know what a dictionary is, or how to find one, even online. And a willingness to use tools he doesn’t understand to generate laws everyone will be bound by. From his actions here we have shown a pretty clear lack of proper ceitical thinking and ethics.
If you can check that it’s correct, then you can research the topic well enough to know what should be in it, which means you can just write the damn thing yourself.
They aren’t too dumb and lazy to check it, but they are dumb and lazy enough to ask something else to write it for them. That is a highly specific amount of dumb and lazy.
He also apparently could not find a real wctual subject matter expert, so he presumably is not very good at his job.
As long as he fact checked what chatgpt told him, who cares?
Chatgpt is a great starting place for a subject you don’t yet understand, but it should always be fact checked before you apply it to whatever scenario you’re in.
Purposely not disclosing the usage of a tool that is widely reported and known to provide false information until after it is signed into law shows a pretty clear lack of ethics.
As long as he fact checked what chatgpt told him, who cares?
But he can’t fact check it - he was using it because he didn’t know enough about the topic to write that section in the first place.
I was suggesting ChatGPT to a social worker I know who was complaining about the number of reports he has to write - AI could bang out a first draft that he could then work up into the final report because he knows the cases and the relevant laws and procedures. If he didn’t know the subject inside and out it could be dangerous. Even then, he should flag it up for others to look it over in case of hallucinations.
If he didn’t know the subject inside and out it could be dangerous
Could be, but not necessarily. It’s good to use as a starting point for further research. It’s good to introduce you to terminology used in the field you’re researching if you don’t know it already.
Knowing what to search for is very very valuable. This can speed up your understanding of a subject significantly.
The problem is GTP is learning from our existing knowledge base. If legislation is trying to amend a broken system, we don’t want AI to be modeling that system. This case seems fairly harmless, an AI takeover isn’t what we should be worries about.
Something like institutional racism being replicated in a more insidious manner is the concern. Relying on these closed systems potentially gives the type of people who implemented the discrimination being modeled to turn around and say, ‘See, we were right all along!’ If results are held up on a pedestal and AI is integrated into our political and legal systems, it may make changing society for the better much harder.
We shouldn’t universally condemn tools like ChatGTP being used in this way, but we should tread very carefully when it comes to large scale societal changes.
Agreed. Chatgpt is a great starting place. Nobody should take what it says as fact, but use what it says to help you research further.
All joking aside, this is the answer.
Prompt: Imagine that you are a capable legislator…
hate the drivel about “false information” when using an LLM. You can use it to help write things you struggle to describe and rewrite things in a certain style. You don’t have to use it for direct information.
Rewrite the previous text as if you are a legislator.
I strongly oppose the excessive focus on “false information” concerns surrounding the use of large language models (LLMs). These models can be valuable tools for assisting in writing tasks, helping to articulate ideas more clearly, and adapting content to various styles. Their use can enhance the quality and versatility of written communication.
This is the bill for anyone curious. https://legiscan.com/AZ/text/HB2394/2024
ITT: A bunch of people who don’t understand how an LLM works.
He used a grammar checker. OMFG the group think stupidity around here is amazing.
“I was really struggling with the technical aspects of how to define what a deepfake was,” he said.
yeah, just a quick grammar check, nothing else
Semantics is grammar
The legal definition of something is not grammar.
