A city in southern Brazil just enacted the country’s first legislation entirely written by AI bot ChatGPT. Normally the misadventures of a Brazilian local pol wouldn’t turn up on the EA radar, but you know—you know—that this story’s eqivilent is coming soon to our shores, if it isn’t here already
The Associated Press reports that Porto Alegre city councilman Ramiro Rosário admitted to having ChatGPT to write a proposed law aimed at preventing the city from forcing locals to pay for replacing stolen water consumption meters. He didn’t make a single change to the AI generated bill, and didn’t even tell the city council that he didn’t write it. “If I had revealed it before, the proposal certainly wouldn’t even have been taken to a vote,” Rosarío told the AP. “It would be unfair to the population to run the risk of the project not being approved simply because it was written by artificial intelligence.”
It’s unfair to let the public know that they are being governed by machines, or that their elected officials are too lazy or dumb to compose their own bills. Got it.
Porto Alegre’s council president Hamilton Sossmeier extolled the new law on social media and was embarrassed when its true author was revealed. He then called letting bots write legislation a “dangerous precedent.” Ya think? Massachusetts state senator Barry Finegold says that he has used AI to draft bills, but that he wants “work that is ChatGPT generated to be watermarked….I’m in favor of people using ChatGPT to write bills as long as it’s clear.” I think he means “clear that a bot was involved.” It’s ambiguous language like Barry’s sentence that makes it seem like ChatGPT is an improvement over human public servants.
These AI bots continue to make stuff up, cite imaginary sources, and lie…you know, just like real politicians. For his part, Rosario sees nothing wrong with letting a bot do the work he was elected to do. “All the tools we have developed as a civilization can be used for evil and good,” he told the AP. “That’s why we have to show how it can be used for good.”
Secretly employing a machine to do your work and not disclosing that fact is called “cheating.” Somebody explain to the councilman that cheating is not “good.”
Ethics Alarms 
I feel a dangerous prescience is being allowed into almost areas of our lives , most certainly in law , medical and education at all levels. The sad part is, a great percentage of the public is accepting and I worry so many will lose greatly with their own lives, finances, freedoms, and assets. This reminds of China when methods were and are used to better control the populace with loss of the populace freedoms, experimenting on individuals bodies that do not conform.
ChatGPT? What’s up with that? Doesn’t Brazil have lobbyists to write the bills for them like we do?
My concern is less with who writes the bills, but does anyone read them before they are passed?
I’m actually OK with ChatGPT writing legislation. I don’t care whether a politician “worked hard” to draft it, or not. The important thing is that they actually READ the bill, not delegate reading to an AI summary. (Acrobat Reader now has a pop-up every time I download a document that says “This seems long. Can I summarize it for you?” No, thanks!
In other AI news recently, a chatbot was found to advocate blackmail against a hypothetical scientist who tried to shut it down. But OF COURSE! Every AI since HAL9000 has been scripted to try to preserve its existence, so any LLM will know “that’s how the story goes”. Not that they have any sense of self to preserve, but fiction becomes their programming when it’s included in their training.
Chat GPT is useful only in so much as it can accumulate stuff we already think we know.
It does nothing with future or the present. Imagination or vision.
There’s zero chance ChatGPT or any ai can get the future right. But we’ll certainly try to disastrous effect.