[wongarsu]

A decade ago, when chat bots were a lot less useful, a common piece of etiquette was that it's fine for a bot to pretend to be a human or God or whatever, but if you directly ask it if it's a bot it has to confirm that. Basically of the bot version of that myth about undercover cops.

I don't see a downside in requiring public-facing bots to do that

Not sure if that's what the proposal is about though, it's currently down

[tenpies]

It bothers me that they didn't consider that this should be bilateral: a bot must confirm that it is a bot, and a human must confirm it is a human.

I wouldn't want humans pretending to be bots, for a variety of reasons.

[rapind]

> I wouldn't want humans pretending to be bots, for a variety of reasons.

It would be so embarrassing if your AI Girlfriend / Boyfriend turned out to be real.

[arkis22]

your AI girlfriend/boyfriend is totally real, its just that its a south east asian who works at a pig butchering scam farm.

https://www.economist.com/leaders/2025/02/06/the-vast-sophis... https://www.economist.com/briefing/2025/02/06/online-scams-m...

[comex]

A law like that would probably be unconstitutional if it applied broadly to speech in general. Compare United States v. Alvarez, where the Supreme Court held that the First Amendment gives you the right to lie about having received military medals.

It might work in more limited contexts, like commercial speech.

[yjftsjthsd-h]

> I wouldn't want humans pretending to be bots, for a variety of reasons.

I don't have an opinion yet, but I can't think of a specific reason to object to that (other than a default preference for honesty). Could you give an example or two?

[MobiusHorizons]

Probably the main risk would be people trusting what they think to be an automated system and trusting it to act for a specific purpose. That or people saying things they think should be private. I’m not saying it’s actually safe to interact this way with bots, but the trust expectations are different enough.

[scarab92]

What you'll find is that most people form a knee jerk opinion first, most often in opposition to change, then retrospectively seek reasons to justify their opinion after the fact.

In other words, people, generally, cherrypick evidence for their opinions, rather than picking opinions for their evidence.

A good sign this is occuring is when the reasons provided are vague, the prevalence of that negative outcome is rare, or are hypothetical scenarios which rely on companies or people behaving in unlikely and unnatural ways (like ignoring broader incentives).

The result is luddite-ism, and proposals exactly like this one, whereby regulations are proposed even in the absence meaningful and demonstrable harm.

[hsbauauvhabzb]

Privacy

[hunter2_]

Your sentiment reminds me of this: https://www.goingconcern.com/when-ai-stands-for-actually-ind...

[banku_brougham]

So compel speech from a person? "Congress shall make no law..." Really the most basic civics education would benefit us all, I think there are some youtube videos about this.

[EFreethought]

To paraphrase a line in the Bible: Are the bots here for the benefit of man, or is man here for the benefit of the bots?

[BobbyTables2]

I’ve had a number of encounters with ISP tech support where the humans seemed a lot like bots…

[pishpash]

All support is basically like bots nowadays, that's why they can be replaced by bots so easily.

[scarab92]

It's all neural nets.

The differences are just implementation details (i.e. whether the transistors are carbon or silicon based).

[rightbyte]

I think they used to have a lot of canned responses which made them feel like bots.

[Galatians4_16]

Freeze Peach says I can be a bot if I want to.

[chrisco255]

Some humans pull it off very well.

[mjbale116]

> I don't see a downside in requiring public-facing bots to do that

Your statement attempts to give an impression of a middle ground, but what it actually does is delegating the action to the human - who has limited energy and has to make hundreds of other decisions.

Your statement sounds like what a lobbyist might whisper to a regulator in an attempt to effectively neuter the bill.

People not versed in technology do not - and do not have to - know what an LLM is or what it can do.

These matters need be resolved at the source and we must not allow hopeful libertarian technologist DDoS the whole society.

[nico]

Archive links on another comment: https://news.ycombinator.com/item?id=42968477

[mixxit]

ah the good ol '/finger [Nick|Address]

[geor9e]

I remember the opposite. The chat bots popular in 2015 were trying to pass the Turing Test and would deny being a bot. The 2025 chat bots popular now are pretty good about explaining they are a bot. Of course, that's just generalizing the popular ones - anyone can make their own do as they please.