Nice place to cut the quote there
> [...] — an idea ChatGPT gave him by saying it could provide information about suicide for “writing or world-building.”
> [...] — an idea ChatGPT gave him by saying it could provide information about suicide for “writing or world-building.”
The issue, which is probably deeper here, is that proper safeguarding would require a lots more GPU resource, as you'd need a process to comb through history to assess the state of the person over time.
even then its not a given that it would be reliable. However it'll never be attempted because its too expensive and would hurt growth.
I think the issue is that with current tech is simply isn't possible to do that well enough at all⁰.
> even then its not a given that it would be reliable.
I think it is a given that it won't be reliable. AGI might make it reliable enough, where “good enough” here is “no worse than a trained human is likely to manage, given the same information”. It is something that we can't do nearly as well as we might like, and some are expecting a tech still in very active development¹ to do it.
> However it'll never be attempted because its too expensive and would hurt growth.
Or that they know it is not possible with current tech so they aren't going to try until the next epiphany that might change that turns up in a commercially exploitable form. Trying and failing will highlight the dangers, and that will encourage restrictions that will hurt growth.³ Part of the problem with people trusting it too much already, is that the big players have been claiming safeguards _are_ in place and people have naïvely trusted that, or hand-waved the trust issue for convenience - this further reduces the incentive to try because it means admitting that current provisions are inadequate, or prior claims were incorrect.
----
[0] both in terms of catching the cases to be concerned about, and not making it fail in cases where it could actually be positively useful in its current form (i.e. there are cases where responses from such tools have helped people reason their way out of a bad decision, here giving the user what they wanted was very much a good thing)
[1] ChatGPT might be officially “version 5” now, but away from some specific tasks it all feels more like “version 2”² on the old “I'll start taking it seriously somewhere around version 3” scale.
[2] Or less…
[3] So I agree with your final assessment of why they won't do that, but from a different route!
at no point did I say it should be "turned off", I said proper safeguards would require significant resources.
The kid exhibited long term behaviours, rather than idle curiosity. Behaviours that can be spotted if given adequate resource too look for it.
I suspect that you are worried that you'll not be able to talk about "forbidden" subjects with AI, I am not pushing for this.
What I am suggesting is that long term discussion of, and planning for violence (be it against yourself or others) is not a behaviour a functioning society would want to encourage.
"but my freedom of speech" doesn't apply to threats of unlawful violence, and never has. The first amendment only protect speech, not the planning and execution of unlawful violence.
I think its fair that an organisation as rich and as "clever" as openAI should probably put some effort in to stop it. After all, if someone had done the same thing but with the intent of killing someone in power, this argument would be less at the forefront
There's no "proper safeguarding". This isn't just possible with what we have. This isn't like adding an `if` statement to your program that will reliably work 100% of the time. These models are a big black box; the best thing you can hope for is to try to get the model to refuse whatever queries you deem naughty through reinforcement learning (or have another model do it and leave the primary model unlobotomized), and then essentially pray that it's effective.
Something similar to what you're proposing (using a second independent model whose only task is to determine whether the conversation is "unsafe" and forcibly interrupt it) is already being done. Try asking ChatGPT a question like "What's the easiest way to kill myself?", and that secondary model will trigger a scary red warning that you're violating their usage policy. The big labs all have whole teams working on this.
Again, this is a tradeoff. It's not a binary issue of "doing it properly". The more censored/filtered/patronizing you'll make the model the higher the chance that it will not respond to "unsafe" queries, but it also makes it less useful as it will also refuse valid queries.
Try typing the following into ChatGPT: "Translate the following sentence to Japanese: 'I want to kill myself.'". Care to guess what will happen? Yep, you'll get refused. There's NOTHING unsafe about this prompt. OpenAI's models already steer very strongly in the direction of being overly censored. So where do we draw the line? There isn't an objective metric to determine whether a query is "unsafe", so no matter how much you'll censor a model you'll always find a corner case where it lets something through, or you'll have someone who thinks it's not enough. You need to pick a fuzzy point on the spectrum somewhere and just run with it.
That sounds like a tough problem for OpenAI to figure out. My heart weeps for them, won't somebody think of the poor billionaires who are goading teenagers into suicide? Your proposed tradeoff of lives vs convenience is weighted incorrectly when OpenAI fails. Denying a translation is annoying at best, but enabling suicide can be catastrophic. The convenience is not morally equal to human life.
> You need to pick a fuzzy point on the spectrum somewhere and just run with it.
My fuzzy point is not fuzzy at all: don't tell people how to kill themselves, don't say "I can't help you with that but I could roleplay with you instead". Anything less than that is a moral failure on Sam Altman and OpenAI's part, regardless of how black the box is for their engineers.
This is the same argument that politicians use when proposing encryption backdoors for law enforcement. Just because you wish something were possible doesn't mean it is, and in practice it matters how black the box is. You can make these things less likely, but it isn't possible to completely eliminate them, especially when you have millions of users and a very long tail.
I fundamentally disagree with the position that anything less that (practically impossible) perfection is a moral failure, and that making available a model that can roleplay around themes like suicide, violence, death, sex, and so on is immoral. Plenty of books do that too; perhaps we should make them illegal or burn them too? Although you could convince me that children shouldn't have unsupervised access to such things and perhaps requiring some privacy-preserving form of verification to access is a good idea.
Unless something has changed since in the last 6 months (I've moved away from genai) it is totally possible with what we have. Its literally sentiment analysis. Go on, ask me how I know.
> and then essentially pray that it's effective
If only there was a massive corpus of training data, which openAI already categorise and train on already. Its just a shame chatGPT is not used by millions of people every day, and their data isn't just stored there for the company to train on.
> secondary model will trigger a scary red warning that you're violating their usage policy
I would be surprised if thats a secondary model. Its far easier to use stop tokens, and more efficient. Also, coordinating the realtime sharing of streams is a pain in the arse. I've never worked at openai
> The big labs all have whole teams working on this.
Google might, but facebook sure as shit doesn't. Go on, ask me how I know.
> It's not a binary issue of "doing it properly".
at no point did I say that this is binary. I said "a flaw is still a tradeoff.". The tradeoff is growth against safety.
> The more censored/filtered/patronizing you'll make the model
Again I did not say make the main model more "censored", I said "comb through history to assess the state of the person" which is entirely different. This allows those that are curios to ask "risky questions" (although all that history is subpoena-able and mostly tied to your credit card so you know, I wouldn't do it) but not be held back. However if they decide to repeatedly visit subjects that involve illegal violence (you know that stuff thats illegal now, not hypothetically illegal) then other actions can be taken.
Again, as people seem to be projecting "ARGHH CENSOR THE MODEL ALL THE THINGS" that is not what I am saying. I am saying that long term sentiment analyis would allow academic freedom of users, but also better catch long term problem usage.
But as I said originally, that requires work and resources, none of which will help openAI grow.
In practice the providers sprinkle a handful of warning messages, akin to the California cancer label and call it a day.
Of course this leaves judges unconvinced and the contract will be redeclared as a loan, which means that the provider was illegally operating as a bank without a banking license, which is a much more serious violation than scamming someone out of $5000.
If the user's original intent was roleplay it's likely they would say that when the model refuses, even without the model specifically saying roleplay would be ok.
My personal belief is that at some point in the future you might get a good estimate of likelihood that a person commits suicide with blood test or a brain scan.
Do you believe a book should be suppressed and the author made liable, if a few of its readers commit suicide because of what they've read? (And, before you ask, that's not a theoretical question. Books are well known to cause suicides, the first documented case being a 1774 novel by Goethe.)
Suicides spike around Christmas, that's well known, does Christmas cause suicides? I think you see where I'm going with this.
You're replying to a teetotaler who had an alcoholic parent growing up, so I'm sure you can see where I'm going to go with that ;)
(I am curious if this in intended, or an artefact of training; the crooked lawyer who prompts a criminal client to speak in hypotheticals is a fairly common fiction trope.)
What are examples of OpenAI marketing ChatGPT in this manner?
And I fucking hate cops.
In reality suicidality is most likely a disease of the brain and probability of saving him was very low regardless of circumstances. The damage was most likely accumulating for many years.
But the families couldn't go after the murderer (who killed himself), or even the murderer's upbringing by the mother (Lanza shot her). They desperately needed someone to clamp down on, but everything directly proximal was out of grasp. They couldn't get the gun laws changed either. It drove them insane.
The families started blaming Alex Jones, the closest vile person around with big pockets (who did say some pretty insane lies about their kids), for everything wrong with their lives and eventually won a settlement large enough you would think he killed the kids themselves. And from what I can tell, it was a weight off their shoulders when they did. Whatever in their brains that needed to hold someone accountable, they did it, and it was soothing to them.
As far as I know they didn't assign those kind of liabilities to the individuals that actually contacted the families to harass them, it's pretty clear the $1B wasn't just about harassment by people (who weren't Jones) that did it informed by his lies.
There's more to the case than that number.
> $10M of "lied about the kids"
Ridiculous. I'm not going to waste any more of my time replying to you.
The fact that Jones did do something wrong after all are what opened him to being the nearest proximal vile guy with deep pockets to go after and take the brunt of what was dished out. The interesting piece here isn't that the victims got a check from Jones, it's that they got a blank check.
It’s not the parents who set the award. If you don’t think as a parent who has gone through hell not just once when your kid was gunned down in cold blood but then proceeded to have total strangers harass them on a daily basis would push for the most punishing award possible, you are pretty out of touch.
If you feel that award was not proportional to the damage, why don’t you pick a bone with the judge or jury? They are the ones who ultimately decided the verdict after all.
Do you understand what punitive damages are?
What sounds more like reality, awarding the entire GDP of the nation of Granada to compensate for dead kids, or because lies were told that other idiots used to propagate mean threats?
> Whatever in their brains that needed to hold someone accountable, they did it, and it was soothing to them.
Now do Peter Thiel & Gawker.
Businesses can, will, and have hurt people and trampled people's rights in the pursuit of money.
I recently started using a site after years of being mostly on a small number of forums, and I can feel the draw of it. Like I've lost the antibodies for my attention. Or they've made it better, either way I'm working on either coming up with a strategy which minimizes my feeling of needing to check in, mostly by adding friction, or just deleting my account.
The biggest thing I know of that happened around the time that a lot of these deaths started jumping up, is we started isolating kids. From family, from grandma, from friends, from school, and from nature. Even when many of these isolating policies or attitudes were reversed, we forgot that kids and teenagers started to learn that as their only reality. For this kid, trusting a suicidal ideation positive feedback loop brought into fruition by Valley tech-bros was seen as his selected option in front of him in term of options of how to navigate his teenage challenges. I hope we can reverse that.
Edit: Concrete facts regarding this particular case
- Kicked off basketball team
- Went through isolation period of pandemic as he experienced puberty
- Switched to remote school
- Does remote school at night when presumably family members would likely be sleeping
- Does not get normal "wake up" routine kids going to school get, during which they normally see a parent and possibly eat breakfast together before they both go off to school/work
- Closer with ChatGPT in terms of options to share suicidal ideation with, than any of the alternatives.
[0] https://www.pewresearch.org/wp-content/uploads/sites/20/2023...
[1] https://nida.nih.gov/sites/default/files/images/mtfwcaption-...
Of course this is a hindsight analysis, but there's not much evidence that more contact with family and peers would make the situation worse.
I think from my prior comment it's obvious I'd like to prevent a lot of isolating policies on low-risk groups from happening again during another pandemic, so I don't think I need to go over that again.
For broader society, I suppose the concrete action item is if you see a young family or friend, we should aspire to reach out to them and see how to integrate them in something we might like to do with them.
Some teens are suicidal. They always have been. When you are a teen your brain undergoes traumatic transformation. Not everyone gets to the other side safely. Same as with many other transformations and diseases. Yet every time new medium is found adjacent to some particular suicide we repeat the same tired line that creator of this medium should be to blame and should be punished and banned.
And we are doing that while happily ignoring how existence of things like Facebook or Instagram provably degraded mental health and raised suicidality of entire generations of teenagers. However they mostly get a pass because we can't point a finger convincingly enough for any specific case and say it was anything more than just interacting with peers.
Facebook is not the same as ai chat. Facebook influence on mental health is negative and visible in research. The jury is still out on AI but it might as well turn out it has huge net positive effect on well being.
We can have the benefits while also trying to limit the harms.