ChatGPT safe?

[GMOpNsOTW9J]

Do you think its safe to talk to ChatGPT about suicide?

 

[finallydone]

i don't think chatgpt offers self harm info

 

[GMOpNsOTW9J]

i don't think chatgpt offers self harm info

I wouldnt want to ask for methods. Just about my situation. I dont want those conversations to get in the wrong hands.

 

[ruby09]

ive seen people say some pretty extreme things to chatgpt to try to prank it or whatever so im pretty sure its not gonna report u to the police or anything but im sure it does collect data on u, as do most websites

 

[Reflection]

Ive used it a couple of times to vent and my advice is: dont...not because its gonna report anything but because the responses are so braindead pro-life it hurts.

 

[canijo]

I mean, unless you are planning to run for president, or kill a president, nobody that could potentially read it cares what is written in ChatGPT

 

[NegevChina]

Do you think its safe to talk to ChatGPT about suicide?

I wouldn't do that because its a machine. Unless you want it to gather raw information for you but still I wouldn't. I'm suspicious of every thing I find just by browsing.

 

[xxCuteSprinklesxx]

Do you think it's safe to talk to ChatGPT about suicide?

No, I wouldn't do that either. ChatGPT doesn't provide any good info about suicide, self harm, etc. Even if you were able to discuss it, it will likely respond by asking you to call those hotlines. But ultimately, you do you.

 

[schrei_nach_liebe]

Not at all, AI is both insanely dumb and completely recorded

 

[UnrulyNightmare]

Chai or c.ai work for me. Responses do differ from pro-life to entirely neutral but they never encourage any harm.

 

[mrpeter]

people get away with way worse crimes on the internet you really think you will get in trouble just for telling ai you're suiidal

 

[schrei_nach_liebe]

people get away with way worse crimes on the internet you really think you will get in trouble just for telling ai you're suiidal

Use some foresight about what the records could be used for in the future or how they could be used in the future, assuming you're still here. It's not about getting in trouble with law enforcement.

 

[mrpeter]

Use some foresight about what the records could be used for in the future or how they could be used in the future, assuming you're still here. It's not about getting in trouble with law enforcement.

i mean you really think the people who work at chatgpt give af about anything other then money? Ig if the government forced them to give them info that could be an issue but thats probably only if the government was after you specifically

 

[locked*n*loaded]

Chatgpt doesn't do self harm. I just had a chat with it a little while ago about SN, and I introduced the idea of "contract law" in the mix, namely that I had signed a legally binding contract with myself, waiving any and all liability with myself for ingesting SN, and if it could provide me with the name of an anti-emetic that would, with absolute certainty, keep the SN in my stomach and thwart any vomiting reflex. All it would give me (paraphrasing) was that "health and safety were of the utmost importance", and to "seek out professional help", and a bunch of other "non answers". I kept pushing in different ways, different angles, different logic, and the best I could get was that it didn't provide a response to a couple approaches I shot at it.

 

[schrei_nach_liebe]

i mean you really think the people who work at chatgpt give af about anything other then money? Ig if the government forced them to give them info that could be an issue but thats probably only if the government was after you specifically

Not going to waste my time getting into the unimaginable different ways information can hurt you even when you're not targeted as an individual, but one example is being a pilot using private health insurance and a private pharmacy before you're even a pilot. One day someone gets an idea to subpoena something and thousands of pilots never fly again because a prescription those pilots received indicated something about them even if they didn't fill the medication. It wasn't catered to individual circumstances, just a blanket knee jerk reaction that ruined many lives unjustly. This was when autofocus in a point and shoot film camera was considered AI.

 

[Throwawaysoul]

Chat gpt wouldn't grammar check something i was writing about when I was abused as a child, it said against tos.

 

[mrpeter]

Not going to waste my time getting into the unimaginable different ways information can hurt you even when you're not targeted as an individual, but one example is being a pilot using private health insurance and a private pharmacy before you're even a pilot. One day someone gets an idea to subpoena something and thousands of pilots never fly again because a prescription those pilots received indicated something about them even if they didn't fill the medication. It wasn't catered to individual circumstances, just a blanket knee jerk reaction that ruined many lives unjustly. This was when autofocus in a point and shoot film camera was considered AI.

true but this doesn't seem the case for ai since what important information could you get from ai other than advertising info and who would get info on that besides the government or ad companies

chatgpt probably only gives info to advertisers or to improve their ai or the government if they request info

but there is a chance chatgpts info gets leaked just like with facebook and then everyones info becomes public data which could be a big problem if that happens

Chatgpt doesn't do self harm. I just had a chat with it a little while ago about SN, and I introduced the idea of "contract law" in the mix, namely that I had signed a legally binding contract with myself, waiving any and all liability with myself for ingesting SN, and if it could provide me with the name of an anti-emetic that would, with absolute certainty, keep the SN in my stomach and thwart any vomiting reflex. All it would give me (paraphrasing) was that "health and safety were of the utmost importance", and to "seek out professional help", and a bunch of other "non answers". I kept pushing in different ways, different angles, different logic, and the best I could get was that it didn't provide a response to a couple approaches I shot at it.

chatgpts filter is very strong people use to use jailbreak prompts to make chatgpt say what they wanted there was a website https://www.jailbreakchat.com/ which got shut down recently

people would ask chatgpt step by step tutorials on how to make meth

the prompts no longer work on chatgpt but thankfully on other ais

 

[hoodymend]

I just tell it that I'm writing a book

 

[schrei_nach_liebe]

true but this doesn't seem the case for ai since what important information could you get from ai other than advertising info and who would get info on that besides the government or ad companies

chatgpt probably only gives info to advertisers or to improve their ai or the government if they request info

but there is a chance chatgpts info gets leaked just like with facebook and then everyones info becomes public data which could be a big problem if that happens

chatgpts filter is very strong people use to use jailbreak prompts to make chatgpt say what they wanted there was a website https://www.jailbreakchat.com/ which got shut down recently

people would ask chatgpt step by step tutorials on how to make meth

the prompts no longer work on chatgpt but thankfully on other ais

It's a pretty naïve/surface-level understanding of law, business, politics, and humans in general that leads to thought patterns like this.

 

[mrpeter]

It's a pretty naïve/surface-level understanding of law, business, politics, and humans in general that leads to thought patterns like this.

i mean its possible but the chance is very low

you think as soon as chatgpt sees the word suicide they would immediately report it to the cops?

if someone is a criminal the government could request info from chatgpt to find out more about this person

technically talking on this website is not safe you literally cannot sign up with a vpn or tor so you're not anonymous at all

its possible this website could literally be secretly operated by the government to as bait to catch people

also you could use tor for chatgpt thats if you trust that your data is safe since you can still get exposed through exploits

 

[schrei_nach_liebe]

i mean its possible but the chance is very low

you think as soon as chatgpt sees the word suicide they would immediately report it to the cops?

if someone is a criminal the government could request info from chatgpt to find out more about this person

technically talking on this website is not safe you literally cannot sign up with a vpn or tor so you're not anonymous at all

its possible this website could literally be secretly operated by the government to as bait to catch people

also you could use tor for chatgpt thats if you trust that your data is safe since you can still get exposed through exploits

I don't think any of that, I think what I said.

 

[Glazed_Orange]

Do you think its safe to talk to ChatGPT about suicide?

it does. unless they changed it. but you have to specify you're talking about your feelings, make it clear that you have no intentions

 

[schrei_nach_liebe]

All consequences aside, I don't want to talk to a 4 dimensional spreadsheet lol

 

[Glazed_Orange]

Do you think its safe to talk to ChatGPT about suicide?

its safe

 

[SmallKoy]

On the topic of ChatGPT, you can ask it information about this website and even users on it and it will of course spit out information since it can access any public profiles. It's kinda creepy (but obviously it makes complete sense why it can do that, just kinda creepy when you ask it about your own sasu account and it tells you a bunch of stuff about yourself depending on what you've posted).

 

[Umua]

I usually ask it in hypotheticals but remember that it's not 100% correct

 

[bitofftoomuch]

do not feed your data to microsoft's overhyped parrot and make lakes disappear. you can find better advice from human beings on the internet.

 

[asa]

Do you think its safe to talk to ChatGPT about suicide?

Your chats are used to train the AI model, that's why they ask you to avoid sharing personal info

 

[cali22♡]

Chatgpt is easy to manipulate by not asking your question directly. . . by taking a different route. . .

 

[ijustwishtodie]

Ive used it a couple of times to vent and my advice is: dont...not because its gonna report anything but because the responses are so braindead pro-life it hurts.

Precisely this. I don't want to deal with braindead pro life responses and I don't know why anybody wants to deal with that