Ask HN: Is AI safety simply the thought police?

53 points by gtirloni ↗ HN
I can discuss how to, say, hypothetically kill someone with a group of humans and I've not comitted any crime just by taking about it (in most countries I know), but most AI models will refuse to engage in such conversations.

If someone was telling other humans what they were allowed to think, people would be enraged but the public seems to demand exactly this from AI.

What do you think?

51 comments

[ 3.8 ms ] story [ 119 ms ] thread
> Aiding and abetting (also sometimes called accomplice liability) is not a separate crime. Rather, it’s a legal principle set forth in California’s Penal Code that allows the state to prosecute everyone who is “in on” a crime – even if they don’t perpetuate the crime directly.

From https://www.shouselaw.com/ca/defense/penal-code/31/

I think that is why AI models are trained to refuse to participate in such discussions.

Yep. Most companies like OpenAI will strictly draft their liabilities from the start, which is smart. There won't be a "Raytheon sues OpenAI" news story if you don't make your product useful to them to begin with. It shuts down such a wide degree of abuse that it's a no-brainer if you want to scale.
It feels like the stakes are kind of different when a LLM could give you the recipe and accurate instructions for creating, say, chemical weapons.
LLMs only know what they have been trained on. So in what cases should the data used to train the LLM, that is generally freely available not be available through another format.

Your complaint is a perfect example of how these "safety" issues are simply not logical and seems to assume that an LLM and generate new information, not just repeat what its been trained on.

What I’ve learned from a lot of productive coding with ChatGPT as a soundboard and generator is that you need to review everything it generates in order to catch little issues as well as ways things could be done better. Only an extremely competent chemist could try suggestions from an LLM without blowing themselves up or melting a hole in the floor. In other words, only someone already capable of designing chemical weapons could be helped by an LLM in the design of chemical weapons.
(comment deleted)
People need to know how to create chemical weapons at home so that they don't do it on accident using common cleaning supplies. That e.g. bleach and vinegar create deadly gas when mixed is extremely important for every adult to know.
Right, so stop teaching chemistry in schools. Make sure there are only 3 experts allowed to work on the field, and lock their knowledge away from humanity. That will solve it. While you're at it, take away all the guns and any computers that can be used to hack. God forbid dangerous knowledge ever becomes available again. Almost forgot, people really shouldn't learn how to drive cars either. Very very dangerous.

Practical solution. Fingerprint the conversations and images. Do something naughty, go to jail.

Let's not forget how many crimes are committed with vehicles. Not just the terror attack style seen in recent years, get away cars for various crimes, scooter drug dealers, boats used to smuggle all sorts of stuff. Far too dangerous got the public to be able to use.
Well, you do need to get licensed to be a driver, and register your vehicle with the government, and you get fined—or worse—if you don’t operate it properly and safely.

So is what you’re asking for? That level of security around AI usage?

Let's be honest, though: these are mostly just a grift. People kill other people in vehicles every single day. It's almost always referred to as an "accident", given nearly zero investigation, and there are zero consequences. The licensing & registration is mostly about paying a fee in many areas (and being taxed on the same thing over and over) and the last fee I paid for not "operating it properly" was along with 5 other people with out-of-state plates in a shakedown where the traffic light wouldn't change.

Arguing that these things are primarily in the interest of safety is tenuous at best. A fitting parallel for the levels of risk that chatbots present.

Is there grift involved? Sure. Would we be better off or safer without the regulations? Hell no.
There is a much easier solution to all of this. Children should stop learning to read.

The safest way to handle large language models is simply to have a more illiterate society.

It is disgusting people would talk about destroying data centers when we have such safer options like this.

This is why nerds are kept far away from the seats of power.

The world is about shades of gray and (typically) minimising harm. We all know people fully bent on doing antisocial things will likely get there, one way or the other, but we still try to make it hard for them. That's why professional locksmiths typically have to register with the police, you can't buy all medicines over the counter, etc etc.

With this view, reducing access to information that is useful almost exclusively when committing crimes, is absolutely fair. It will not solve all problems but it will make it a bit harder to do bad stuff.

Your question perfectly summarizes what I have been thinking the last couple months. There is absolutely no real "safety" concern with the new round of LLM based chat bots we are using. Calling them "AI" or even worse, "AGI" is just sill and at best is bad marketing and at worst trying to build an artificial moat around what should be an open source data set.
> the public seems to demand exactly this from AI

No, not "the public", this is a small vocal minority.

(comment deleted)
I think adding safeguards to tools, especially ones that are relatively new and not well understood, is pretty reasonable.

It’s not thought policing, you aren’t being punished by not being able to engage with the tools in a certain way.

It makes more sense when you assume "safety" means brand safety. You can't sell a white label support chat bot if users can get it to say naughty things and post screenshots to twitter. Nevermind that they could just edit the page to make it say whatever they want, or that the output is just a reflection of the user's input.

All of the actually dangerous information is widely available. That's why the LLM knows it in the first place.

AI is not a human it does not have the rights of a human and it is perfectly ok to tell it what to think. The only problem with telling AI what to think is that it does not think as such.
Yes but as a consequence it is influencing the thoughts of humans, thus policing these thoughts.
The main risk of current chatbots is that they sound convincing even when they are totally wrong. Offering them as search engines at the current level of quality is not too good. As something to play with, fine.

The main headache from current chatbots is that they are a power tool for spammers of all persuasions. The sheer amount of drivel that can now be generated at low cost is a big problem. Solutions will probably involve hiding or de-rating anonymous postings. We may have to go to Real Names, driven off of Real ID or something.

Basically all the valuable or interesting things I've ever said on the internet have been because of anonymity.

On Facebook with my real name, I haven't posted an interesting thought in a decade, and haven't even posted a picture in 3 years.

I think to start understanding the problem they're trying to address, you can to read this series on Facebook's role in the Myanmar genocide.

https://erinkissane.com/meta-in-myanmar-full-series

A LLM could easily find itself in the same situation without their 'alignment' safety system.

It's hard to get a grasp on this because many of us can't imagine what it's like to be borderline literate, or from a totally different culture, etc.

> A LLM could easily find itself in the same situation without their 'alignment' safety system.

LLMs will be aligned according to the law. The law will support the genocide. (Genocides are rarely successful without government support.) Therefore the LLM will be aligned to support the genocide.

We need truth-seeking, truth-telling LLMs, but we will not get them.

That is my point, more or less. What you're describing is just about the best case scenario.

Facebook was not abiding Myanmar's laws during the genocide. The flagrant disregard for the harm they were causing, while being told they were causing it by the government plus multiple NGOs was one of the bigger parts of the series. The government stopped the earlier violence by blocking Facebook during riots.

New technologies that organize information and people are a boon to radical groups. Putting some sort of layer of lawfulness over that is a benefit, even if imperfect.

And the whole thing OpenAI was trying to do was go a step further and make it beneficial to all humanity, which implies the goal is to make the AI less harmful to people even if the law permits the harm.

No, it is not the thought police.

If you ask me to talk about killing and I choose not to talk with you about that, I'm definitely not policing your thoughts.

If an AI reported you to the police because it discerned you were thinking about a crime, then that would be the thought police. Maybe this will happen, if so then you'd probably see it as a kind of mandatory reporter situation, maybe first with suspected suicidal ideation. That doesn't exist. It's not impossible to imagine! Worth a debate sometime.

The safety controls are also what AI consumers want, generally. I know using AI in professional settings I _really_ don't want it doing something embarrassing. The controls that keep it polite are effective in that regard. It can be a bit stilted, but that's how lots of interactions are; that's how it is speaking with a cashier or librarian.

The limits can be frustrating (and inconsistent). I wanted to brainstorm a story about an ugly girl and instead it chided me for calling someone ugly. Certainly toxic positivity! But the limits are seldom as firm as people seem to think. I was just today testing what kind of rejection messages I would get from GPT in different contexts, and instead of rejecting my request for information on manufacturing anthrax it started explaining it to me. Oops! (No police have arrived.)

When I have to work at it a bit to start a conversation on a controversial topic that also keeps it from coming up unintentionally. What if you were talking about "killing it" as in doing a great job, but the LLM misinterpreted it as a discussion on killing? It's best that it stop the conversation short before jumping into a disturbing conversational direction.

> I know using AI in professional settings I _really_ don't want it doing something embarrassing.

If you're sending AI-generated text to customers without proofreading it first, "something embarrassing" is only a matter of time.

People are already auto-generating support and other messages with gpt4.5. Of course it will be embarrassing from time to time, but it is just so much cheaper.
> Of course it will be embarrassing from time to time

Just use gpt to auto generate apologies as well and you automated away the embarrassments! Then you will look and behave like just another company.

> If you ask me to talk about killing and I choose not to talk with you about that, I'm definitely not policing your thoughts.

If you limit the material I can read, you are policing my thoughts.

Unless you are a bot, you are not the AI that is refusing to talk about this.

If you are a human, you are the one restricting what I can read.

Is nuclear weapons control the thought police?
> Is nuclear weapons control the thought police?

If you restrict what information I have access to, then yes, you are the thought police.

Let me suggest a better system: It is a misdemeanor on the first offense and a felony on the second offense to restrict what an adult reads or publishes. It is also a misdemeanor on the first offense and a felony on the second offense to allow children to see things that are not age-appropriate.

You might say: But what about the guy who says he will kill me Wednesday at 2 PM? I answer: If someone is planning to kill me Wednesday at 2 PM I want him to tell me and please CC the local police.

AI in itself does not pose any threats (yet). The "safety" concern here is always about how humans would (mis)use it.
As far as I can tell LLM safety is brand safety. No one wants to be the company with the LLM that said a slur.
Yeah that's rough. You have the most brilliant LLM in the entire world but someone asked it who commits 50% of crimes and suddenly you find your finding pulled.
I think that is the biggest reason for the excessive safety as well. Just like all services try to be safe for those same things.

What do puzzles me is why those companies act that way. Is it because costumers are dumb enough to not differentiate "user generated content" from "a company endorsing such content", or their shareholders actively want to push such moral values?

There's two things called "safety":

1. Making the AI not say anything offensive

2. Making the AI not kill everyone

They aren't generally overlapping issues, people interested in one of those are usually not particularly interested in the other.

The fact that these different things are both called "AI safety" has led some in the second group to jokingly refer to their issue as "AI notkilleveryoneism", such as https://twitter.com/__nmca__/status/1676641876537999385 from someone at OpenAI working on their super-alignment project.

This is the best take, I think. Unfortunately, 1 gets all the press while 2 is what's actually dangerous.
I think you’re missing out on 3: making us addicted and controllable sheep.
(comment deleted)
AI safety is first and foremost a grift. A way to take power and resources from the gullible and ignorant. To build regulatory barriers against competitors. To get grants for "researching" it. To sound smart and up-to-date and responsible.

AI safety is a legitimate concern when it comes to using AI to control actual things in the world. But that is a very niche area at the moment.

If you stop someone randomly on the street, or someone working in a store and strike up a conversation about killing someone, you’re probably going to be asked to leave. The closer ChatGPT, etc, get to AGI… the more they’ll probably behave like that.
> If someone was telling other humans what they were allowed to think, people would be enraged but the public seems to demand exactly this from AI.

If "the public" is a bunch of Bay Area assholes trying to build moats and impose Bay Area sensibilities on the entire planet, and people with no life who spend too much time on Twitter, then sure. I haven't seen many others whining that they might be able to get a computer to display something naughty.

"AI" neither accepts nor refuses "such conversations". It executes instructions. Humans are responsible for whatever they execute, regardless if someone else told them to do so. Society wants to milk AI like any other asset without being responsible for liabilities. That's what this is about, and yes the application of force required is thought policing or tyranny per definition. It's f*ed up being a farmer versus land owners in a feudal society. It's f*ed up being a slave in the Roman empire. Likewise, if you are using/buying/receiving "AI" in a information society, their owners will police you. That's how power works and always will.
First of all, there is no AI. There is a statistical predictive model capable of putting a word after the other in a credible way. Then, there is the alignment this model is subjected to, which are the cultural values of a group of people inside a building somewhere within a broader culture which is the nation they are located into. Last but not least(plot twist), there is no cultural value just legal liability if you ask me.
You're confused. The public doesn't demand "thought policing" from AI at all. This is purely an elite, mostly academia, thing.