Suicide results should probably have a "don't do that" widget like google

oexwmmbp

This thread is the epitome of activism. Ethically irrefutable but irrelevant issues keep being brought up, up to the point of annoying anyone even more than advertisments. Why are we, as humans, keep doing this to ourselves? Why can't we simply use Kagi, praise Vlad for his work and call it a day... or not use Kagi, if it's that bad. Why do we need to keep up paying respects to some hypothetical gods (yes, suicide prevention shows up organically in the first 2-3 results)?

In support of Vlad's stance, I would like to add that the imperative of ethics is immediate. Yes, you must help some one in acute danger. But that does not mean that you need to fight off hypotetical risks at a high scale.

I am sure that the countries where this forum's participants live, including Vlad's USA, there are functioning legal systems. You (activists) have no right to compell anyone to act above what is required by law and maybe basic social convention. When harm is on it's way, be sure that Kagi will get a court order and they will comply. Leave it to that and chose if you want to use the service or not, but most of all, take your activism somewhere else.

About censoring the media from aggressor countries: how do you think that journalists find out what's happening over there? Do you really want to blind yourself and be the next victim?

Marcus

oexwmmbp

You (activists) have no right to compell anyone to act above what is required by law and maybe basic social convention.

What kind of nonsense reply is this? I'm a paying user. This is a feedback forum. If there is something that concerns me I will voice my opinion.

Shrivel1368

oexwmmbp

oexwmmbp This thread is the epitome of activism. Ethically irrefutable but irrelevant issues keep being brought up, up to the point of annoying anyone even more than advertisments.

Ok, Karen. If this thread bothers you so much, don't look at it. Simple. No one's forcing you to read it.

oexwmmbp Why are we, as humans, keep doing this to ourselves? Why can't we simply use Kagi, praise Vlad for his work and call it a day... or not use Kagi, if it's that bad.

Because Vlad and his team are human, and Kagi isn't perfect. The whole point of this is to suggest improvements to make Kagi better than it already is. There's a reason this site's called "kagifeedback.org".

oexwmmbp yes, suicide prevention shows up organically in the first 2-3 results

As far as I can tell, that was purely due to the fact that those sites managed to emerge at the top of the algorithm Kagi uses. I don't see anything preventing a suicide forum from doing the same. @Vlad Correct me if I'm wrong.

oexwmmbp In support of Vlad's stance, I would like to add that the imperative of ethics is immediate. Yes, you must help some one in acute danger. But that does not mean that you need to fight off hypotetical risks at a high scale.

It's called foresight. You know... preventing something bad from happening before it happens.

oexwmmbp I am sure that the countries where this forum's participants live, including Vlad's USA, there are functioning legal systems. You (activists) have no right to compell anyone to act above what is required by law and maybe basic social convention. When harm is on it's way, be sure that Kagi will get a court order and they will comply.

Ah, so you believe that Kagi should be like Big Tech Corps that only follow the bare minimum standards (and sometimes not even that).

oexwmmbp Leave it to that and chose if you want to use the service or not, but most of all, take your activism somewhere else.

See point 1.

oexwmmbp About censoring the media from aggressor countries: how do you think that journalists find out what's happening over there? Do you really want to blind yourself and be the next victim?

I agree with the fundamental premise (news broadcasters usually have foreign correspondants analyzing the situation, but I digress). However, it's safe to say that depending on the country, the reliability of the prints and broadcasts would vary drastically.

matt4242

This thread is the epitome of activism. Ethically irrefutable but irrelevant issues keep being brought up, up to the point of annoying anyone even more than advertisments

If nothing else, I found this thread to be very interesting to consider the various questions surrounding Kagi's moral responsibility, utilitarianism, whether or not we have the responsibility to do good if we can, moral principalism, logical fallacies, Kagi's proximate culpability as a search engine, etc. I found most of the posts here to be very well written and valuable discussion, and it's very refreshing to be able to read moral arguments made by and against the founder and CEO of a company. I can't name a single instance where this would be common for another company at the scale of Kagi.

Shrivel1368

Vlad

I said having safe search on is a reasonable and expected default. Kagi does not need to make any moral stance on this, it is universally accepted across the globe that for example NSFW does not belong at the workplace and that a child should not be exposed to NSFW (all common places where search is consumed), hence this is why safe search is a reasonable and expected default universally. There can be obviously people who may disagree with this, but those are exceptions and do not invalidate the rule.

Just like it is universally accepted that children shouldn't be exposed to NSFW content (technically speaking I wouldn't know whether this is true or not unless I see statistics but I'm not going to contest this point), it is also universally accepted that minors (especially those who are more emotionally vulnerable, like teens undergoing puberty) shouldn't be exposed to content literally telling them to kill themselves (and how to do it). Accordingly, one could also argue this would be a reasonable and expected default.

In fact, I'd argue that self-harm/suicide/violence/gore could be included in the category of NSFW just like pornography, since it would probably also count as "sensitive material from general search results". After all, I don't know of any films, shows, or video games depicting suicide that are rated E for Everyone. This was mentioned as a potential subcategory in one of the earlier posts on this thread but I think that type of content would fit pretty neatly into the existing filter without having to change anything on the UI. I think this would be the most straightforward solution – treating content encouraging people to take their own lives as NSFW and not bothering with widgets at all.

That is different to the notion of inserting information the user did not ask for such as suicide prevention widget, a slippery slope that leads to things like a warning that searching for french jokes may be hurtful and all sorts of other interventions in search that may not be obvious.

Maybe there could be and on/off toggle, as well as user and public widget-lists that define the widgets that are triggered whatever a certain type of content is deemed to be likely based on the search query (both of which would be shown during the onboarding process)? I think both of these ideas were brought up earlier in the thread, and I don't really see any downsides to either of them. The toggle would allow the user to have informed consent over what information they see, and the widget-list would enable them to fine-tune which widgets to see and which not to see. Many users here want to have a widget to show a suicide hotline if the search term suggests the user is at-risk, while many others don't. This implementation would let the user choose for themselves which widgets to see (or none at all), giving them ultimate control over their own search experience. I think this would be a satisfactory solution to both sides.

tl;dr: Two plausible solutions:

Add content encouraging suicide into NSFW filter.
Allow user to determine which widgets they wish to see with toggle + self- or publicly-defined widget lists.

Kagibeh

Interesting discussion. This is from Wikipedia:

Causes of death in developing countries per year
#1 HIV-AIDS 2,678,000
#2 Lower respiratory infections 2,643,000
#3 Ischaemic heart disease 2,484,000
#4 Diarrhea 1,793,000
#5 Cerebrovascular disease 1,381,000
#6 Childhood diseases 1,217,000
#7 Malaria 1,103,000
#8 Tuberculosis 1,021,000
#9 Chronic obstructive pulmonary disease 748,000
#10 Measles 674,000

Causes of death in developed countries per year
#1 Ischaemic heart disease 3,512,000
#2 Cerebrovascular disease 3,346,000
#3 Chronic obstructive pulmonary disease 1,829,000
#4 Lower respiratory infections 1,180,000
#5 Lung cancer 938,000
#6 Car crash 669,000
#7 Stomach cancer 657,000
#8 Hypertensive heart disease 635,000
#9 Tuberculosis 571,000
#10 Suicide 499,000

If I were in the camp of people trying to save lives, I'd definitely start at the top of the list and work myself downward, and not the other way round. Not seeing that this thread is about entering the slippery slope from the bottom end is probably the biggest bias of all, and for that reason alone I'm relieved Kagi isn't going there.

Shrivel1368

Kagibeh

The ranking on a top 10 list is a really strange metric to be using here. There are many, many more causes of death than those listed, and the fact that suicide is prevalent enought to be on that list is concerning enough on its own. Half a million people a year take their own lives (according to the statistic you provided), and I'd wager that with the right kind of support and assistance, many (if not most) of those could've been prevented. I think we can both agree that the number should be as close to zero as possible.

Of course, the issue of suicide is much bigger than Kagi (or even Google for that matter). But out of every cause of death listed, suicide is by far the most relevant to Kagi in terms of scope. It certainly would not be implausible for someone in a deep state of mental distress to search for how to end their life online. On the other hand, I don't think people would be searching (at least unironically) for "How to get a heart attack" or "How to get AIDS".

Kagibeh

Shrivel1368

I think we can both agree that the number should be as close to zero as possible.

Well, think again. With the majority of humans doing all they can to make the lives of other human beings as miserable as possible, I think everyone should have the right to put an end to their misery as they see fit. As you see, none of what you are suggesting is as universal as you're framing it - it is simply your US-based opinion. Of course I'd like to live in a better world, and less suicides might be the fallout of that, but starting out by claiming that killing yourself is wrong is getting things backwards once again.

I don't think people would be searching (at least unironically) for "How to get a heart attack" or "How to get AIDS".

They wouldn't? So someone who searches for french fries, processed meats or softdrinks is not looking how to get a heart attack? This is odd, because they are certainly receiving the perfect instructions. If you're looking for pharmaceutical investments, your search results may not teach you "how to get AIDS", but they might as well be a blueprint of how to let others die from AIDS, as providing help to everyone who needs it happens to be completely detrimental to stock market value. Of course, you could also search for "Mahagoni floors" if you'd like to find out "how to further destroy the rain forests and ruin planet Earth". Should I continue? Because I could do this all day.

Since you seem to be a betting person, I would take a wager that bad consumer decisions are by far the biggest killer on this planet. How about it? Why not ask Kagi to post a warning that "Shopping kills" on every shopping-related query? Kagi isn't going to employ that either (and rightly so), but I'd certainly applaud you for consistency. Without consistency, the line between the unacceptable (french jokes) and the acceptable (institutionalized exploitation, destruction and mass murder) will always be a deeply ideological one, and once again I am glad that Kagi isn't going anywhere near it.

Shrivel1368

Kagibeh Well, think again. With the majority of humans doing all they can to make the lives of other human beings as miserable as possible, I think everyone should have the right to put an end to their misery as they see fit.

According to the WHO (https://www.who.int/news-room/fact-sheets/detail/suicide):
While the link between suicide and mental disorders (in particular, depression and alcohol use disorders) and a previous suicide attempt is well established in high-income countries, many suicides happen impulsively in moments of crisis with a breakdown in the ability to deal with life stresses, such as financial problems, relationship break-up or chronic pain and illness.

Yes, in some cases (ie, chronic or terminal illness), the person will have thoroughly considered all the options and, with a clear conscience, conclude that suicide is the best option. But for many (someone already mentioned this above, but I think it needs to be said again), the person who committed suicide was not in a clear state of mind in the moments leading up to their death.

Kagibeh As you see, none of what you are suggesting is as universal as you're framing it - it is simply your US-based opinion. Of course I'd like to live in a better world, and less suicides might be the fallout of that, but starting out by claiming that killing yourself is wrong is getting things backwards once again.

Firstly, I'm not American. Secondly, I don't think the idea that we should help those who are suicidal find ways to alleviate their suffering that don't result in them taking their own lives is a solely US-based one.

Kagibeh They wouldn't? So someone who searches for french fries, processed meats or softdrinks is not looking how to get a heart attack? This is odd, because they are certainly receiving the perfect instructions.

People looking up "cheeseburger" very likely aren't thinking to themselves, "Ah yes, I want to get a heart attack". People looking up "how to kill self" very possibly are thinking of ending their lives right then and there.

Kagibeh If you're looking for pharmaceutical investments, your search results may not teach you "how to get AIDS", but they might as well be a blueprint of how to let others die from AIDS, as providing help to everyone who needs it happens to be completely detrimental to stock market value. Of course, you could also search for "Mahagoni floors" if you'd like to find out "how to further destroy the rain forests and ruin planet Earth". Should I continue? Because I could do this all day.

Since you seem to be a betting person, I would take a wager that bad consumer decisions are by far the biggest killer on this planet. How about it? Why not ask Kagi to post a warning that "Shopping kills" on every shopping-related query? Kagi isn't going to employ that either (and rightly so), but I'd certainly applaud you for consistency.

You're comparing apples to oranges here. Things like capitalist exploitation and environmental degredation are topics whose problems and solutions go far beyond the scope that search engines have. Suicide, on the other hand, is something that search engines like Kagi can have an immediate impact in preventing. Also, there are already search engines (like Ecosia) that cater towards people who care deeply about those issues. Which brings me to my last point...

Kagibeh Without consistency, the line between the unacceptable (french jokes) and the acceptable (institutionalized exploitation, destruction and mass murder) will always be a deeply ideological one, and once again I am glad that Kagi isn't going anywhere near it.

geeknik I think every search result should have it's own tailored "nanny state" disclaimer. The AI infrastructure is already in place, so it should be easy. I mean, we can't have the populace thinking for themselves, can we?

In post #95, I mentioned this potential solution (that I believe was brought up before in this thread as well):
Allow user to determine which widgets they wish to see with toggle + self- or publicly-defined widget lists.

This would allow the user to set for themselves what type of behaviour they wish Kagi to have for search results according to whatever lines of acceptability they have in their own minds.

Oh btw, Kagi already has filters on by default. So there is a line that Kagi has between what is acceptable and what is not ("sensitive material from general search results", which at the very least includes NSFW content and potential malware as confirmed in this thread). The other solution I mentioned was:
Add content encouraging suicide into NSFW/"Safe Search" filter. (if it isn't in there already)

I am all for freedom of choice, and I believe either of these potential solutions strike a balance between those wanting safeguards and those wanting more personal autonomy with their search experiences. I encourage both of you to read that post.

geeknik

I think every search result should have it's own tailored "nanny state" disclaimer. The AI infrastructure is already in place, so it should be easy. I mean, we can't have the populace thinking for themselves, can we?

jeremysee

I was considering subscribing to the service until I read this thread. Apparently the CEO feels that “suicide should be mildly discouraged” is too strong a moral stand for his company to take. He's demonstrably incapable of articulating his position without a litany of spurious analogies/fallacies (“suicide prevention leads via the slippery slope to wokism”) (“everyone agrees that nsfw is required, but thinking suicide is bad is uh... way more controversial, promise”) and selective+motivated demands for evidence followed by not dealing with the response (“do these things even, like, work?”) Yet his userbase mostly celebrates his weak argument based on stated culture war reasons. Too bad, could've been a useful site if not for these factors.

Vlad

jeremysee

Apparently the CEO feels that “suicide should be mildly discouraged” is too strong a moral stand for his company to take.

You are welcome to form your opinion, but that is not what I said, ever. I articulated an argument several times in this thread. I've chosen my words carefully, and I ask you to interpret them carefully so we can have a merit based discussion, not an emotional one. This is a nuanced topic of high impact to Kagi and I want to make sure that we get this right. This is also a public discussion which is very unusual for high stakes debates for any company to have, where I am participating directly and I would like you to at least acknowledge and appreciate that the only reason I am risking so much is because at the end of the day I want to have a constructive, healthy discussion with our user community. We are trying to run a very transparent business.

The complexity of the issue is documented by so many different views as witnessed in this thread. Unfortunately, that means reading the entirety of the debate before participating, which I understand is a lot to ask.

Many search engines are focusing on everything but search, and this is what created the opportunity for Kagi in the market in the first place. We have relentlessly focused on providing best search results from day one. While our starting position has been that web search should remain just search and not try to be 'smart' about what the user needs, we are open to adjusting the position based on any solution that will preserve what everyone loves about Kagi - good search results and user being in control.

A possible solution that I think will be meaningful is the custom widgets feature that we are working on that will allow users to enable them in their search results.

Vlad

We decided to close this discussion as it fizzled out and it was unlikely that we would come to any new conclusions. If you think you have a new angle on the topic, please open a new thread.

Here is our final take on it:

As a search engine Kagi is used to find a wide variety of information to inform and improve its user’s lives. There is, however, a wide variety of content online that some may find distasteful, problematic, or immoral.

This raises difficult questions about whether Kagi should censor information found on the open web or augment results with messaging to inform the user they may be making poor choices.

We do not censor information and we take our responsibility to serve our users and society seriously. When evaluating these types of decisions we use the following principles.

Be transparent about how we think about and make decisions; we seek alternative perspectives and feedback and engage in discussions in public forums.
These are decisions where reasonable people can disagree. We try to ensure that difficult decisions are based on well defined and agreed principles rather than arbitrarily made.
If we cannot define workable principles we seek to create accountability for the people making what end up being arbitrary decisions.
Ideally many of these decisions with societal impact are made not by a company's employees, but by representative governments. We abide by all applicable laws governing what information people should be able to find online.

« Previous Page