Will it leak or be announced that a campaign speech, advertisement, or political speech in an EU country or the United States was written by ChatGPT or other LLMs, in 2023?
Related questions
🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ2,457 | |
| 2 | Ṁ1,992 | |
| 3 | Ṁ1,171 | |
| 4 | Ṁ1,029 | |
| 5 | Ṁ559 |
@jonsimon Yes that's why I agreed that it should be resolved in the affirmative. The "making a point about AI" thing was mentioned below, I think my mistake in predicting no was forgetting about House floor speeches.
Actually, speaking of which, does that count as "major"? "Major" is left undefined and this speech was made to like a couple dozen people.
It was already used by a judge to make a ruling
https://www.theguardian.com/technology/2023/feb/03/colombia-judge-chatgpt-ruling
@MetroWind Well at least this is what Bard has to say:
It is difficult to say for sure whether or not a politician would publicly admit that their speech was written by AI. There are a number of factors that could influence a politician's decision, including the political climate, the audience, and the politician's own personal beliefs.
In some cases, a politician may be more likely to admit that their speech was written by AI if they believe that it will help them to connect with voters or to make their message more persuasive. For example, a politician who is running for office on a platform of technological innovation may be more likely to admit that their speech was written by AI, as this could be seen as a sign that they are embracing new technologies.
On the other hand, a politician may be less likely to admit that their speech was written by AI if they believe that it will damage their credibility or make them appear less intelligent. For example, a politician who is running for office on a platform of traditional values may be less likely to admit that their speech was written by AI, as this could be seen as a sign that they are out of touch with the people they are representing.
Ultimately, the decision of whether or not to publicly admit that a speech was written by AI is a personal one that each politician will need to make for themselves. There is no right or wrong answer, and the decision will likely depend on a variety of factors.
There are minor speeches being made constantly. If a speech-writer publically says they've used ChatGPT to substantially write or draft one their speeches, does this count, even if it's not something huge like a State of the Union?
Politicians speak,
Words calculated to seek,
Support from the masses,
But written by machines, not asses.
I think that the likelihood of this happening is a lot higher than the likelihood of this happening and us finding definitive proof of it soon. It'd be harmful politically and therefore kept secret and publicly denied.
@ConnorDolan Speaking of which, what is your standard of proof? How is a potential leak verified to be indeed true? Must the person who used the AI admit it?
If it's "somebody accuses someone else of using AI to write a speech" then I'd like to change my position, because that will almost certainly be a smear regardless of if it's true.
@ConnorDolan Hmm, it seems like just an accusation isn't enough. I think some supporting evidence will be required-for instance, if Scott Aaronson's GPT detector was able to detect these things.
Do you require finished version to be written by AI? Or is it still "YES" if it turns out LLM made the first draft or helped with editing?
@XavierDunikowski I think "substantially written". So, if it was written by AI then tweaked, that counts. If it was a couple bullet points from ChatGPT, that doesn't count.
If that electron decides to run for president, the speech is already in the bag…
Related questions
