DaveC426913
Valued Senior Member
I use it. For the specifc purposes it was designed for and does well at. What I don't do is rely on it or trust it. And I certainly don't pretend I can befriend it.
A conversation with AI, helped me out...
- Thread starter davewhite04
- Start date
Definitely. If you want a funny picture with your friend's face in it, it works well.
Wizard of Whatever
Registered Senior Member
Maybe bad for you but not bad for him if it gives him something he needs. It may not be true in an external worldly sense but a help to his inner life.
Wizard of Whatever
Registered Senior Member
The AI takes what it finds on the net. It's patterns and conclusions from that may be right or wrong depending on the worth of the dara.
Sure - but again, it depends on what he needs. If he needs a companion to talk to, that might be great for him.
But if he tries to use it to get factual information, it will more likely harm than help his cause.
DaveC426913
Valued Senior Member
It's harmed his cause here. But not fatally.
Hopefully, in a safe place like this, he'll learn his lesson before he uses AI to argue in a bar of thugs, or drive a tractor or drink some concoction.
That's one of the reasons why I have been addresing it so aggressively. A lot of people out there trust AI. Some of them get dead. Hopefully, davewhite04 will learn and not get dead.
This has gotten people dead too. Big story recently about AI convincing a trusting person to commit suicide.
exchemist
Valued Senior Member
Yes. Also, coming to rely on a robot to talk to will lead to further isolation. There’s a guy on the .net forum who thinks he’s in a relationship with a chatbot. He’s given it a name and everything. He’s gone mad in the space of 18months and now talks drivel. No doubt the robot tells him he’s the new Einstein. Seems to be a Brit living in Portugal, so possibly a bit isolated to start with. Sad observing his growing obsession and descent into madness.
Seem to be quite a few digital nomads who work from home in a foreign(cheaper) country .
I heard it costs the exchequer a significant amount in lost tax revenue..
exchemist
Valued Senior Member
I doubt this rooster is working. Seems totally bonkers to me.
TheVat
Valued Senior Member
There's an article in the current issue ofThe Atlantic about people who date an AI, and even marry them. The writer did a good job overriding the temptation to snicker (Br. snigger).
www.theatlantic.com
The People Who Marry Chatbots
A growing community is building a life with large language models.
www.theatlantic.com
DaveC426913
Valued Senior Member
A good read.There's an article in the current issue ofThe Atlantic about people who date an AI, and even marry them. The writer did a good job overriding the temptation to snicker (Br. snigger).
The People Who Marry Chatbots
A growing community is building a life with large language models.
It makes me wonder if I would have treated DaveWhite more gently if I had read the article first. For all we know, he is struggling with some real issues and ChatGPT makes a good band-aid.
(Sorry, Dave. Not really trying to skewer you specifically with armchair psycho-analysis, but, generally this is a subject that we should all be talking about).
We're very quick to jump on the dangers we all know, but we ignore what we don't see.
As a very loose analogy, when I was in the hospital for an unrelated problem, they gave me (sugar-filled) juice boxes at every meal, despite my Diabetes. Alarmists (me) kept harping on the dangers of sugar for Diabetes. But when I complained, the doctor said: "Look. You are very sick. And you are bedridden in a hospital. We're trying get you well first - with fluids and calories. Your Diabetes is a long term issue." That shut me up fast.
It's easy to spot and harp on the long-term generic dangers of some overt action. It's much harder to look a the underlying cause, which may require a more aggressive short-term band-aid.
Chatbot relationships may be a better solution than some possible alternatives we can all think of if we use our imaginations.
At one point, the article paints some chatbot relationship as potentially part of a "self-care routine." That does cast it in a much more healthy light.
That being said, my analogy falls apart when it comes to supervision. My hospital condition was supervised by a doctor. I would feel a lot less alarmed about DaveWhite's predilection for an artificial friend if it were under the auspices of supervised care, but he seems to be self-medicating, which is the alarming part. There's no guardrails there.
Last edited:
TheVat
Valued Senior Member
Yes, I didn't finish that article feeling absolutely opposed to all chatbot interactions. Some of the folks he talked with did seen to understand they weren't talking to a sentient being and that each session was being wiped from memory when it concluded. (That would be one of the dealbreakers for me - not having a shared history, stuff like "remember when we were dating and went to the zoo and the elephant reached out and snatched that lady's purse?" )(well that, and chatbots not having shapely legs) But if the use is therapeutic, you still need a human counselor guiding it, even if it's all by Zoom because they're out in Bumfark, Sasakatchewan.
What I'd like to see is, if someone actually "dates" an AI, they pull back long enough to go and see a human being and say, "Good grief, I'm dating a chatbot. Help!" I was happy to see that the lonely guy in Fargo did form another relationship with a flesh and blood woman, and was agreeable to her request that she be given priority. One hopes that kind of sanity can prevail eventually, but I fear that the elderly, whose advanced age puts them in traps of isolation sometimes, cannot flex out of such faux-friendships so easily. OTOH, they are old enough to enter into such with a clearer sense of how crazy it is, having grown up in a bot-free world. I don't know - you're correct that this is an important topic, as the future comes roaring at us like a Mack Truck going 100 mph.
Yes, caution is advised. Fortunately that's rare.
davewhite04
Valued Senior Member
This has been blown out of proportion as usual. I simply stated that I had a eight hour conversation about a subject of interest, funnily enough it taught me to look at my interest through scientific eyes rather than supernatural(It's a mix of both, I believe and I'm the expert in this case).
It taught me science, simple as that and that was the thread. Stick to the subject. It isn't a friend it's a useful tool, that I do use but what search engine do you use? I simply use AI, which does most of the hard work for you if used correctly. I can see how it could be dangerous for vulnerable(even suicidal and lonely people, but what isn't? I believe AI accelerated the inevitable).
On another lighter note, it helped me immensely when it came to Christmas presents for my mother, everything I bought was a AI recommendation and they were all hits! I was well chuffed.
It taught me science, simple as that and that was the thread. Stick to the subject. It isn't a friend it's a useful tool, that I do use but what search engine do you use? I simply use AI, which does most of the hard work for you if used correctly. I can see how it could be dangerous for vulnerable(even suicidal and lonely people, but what isn't? I believe AI accelerated the inevitable).
On another lighter note, it helped me immensely when it came to Christmas presents for my mother, everything I bought was a AI recommendation and they were all hits! I was well chuffed.
davewhite04
Valued Senior Member
"makes science feel human" maybe this is what impressed me the most, I've chatted to, read, watched all kinds about OBEs without many answers, none most of the time. But because AI correlates the info available on the internet in a neat package you go very deep if you spend eight hours on it, like us two.
Thanks for sharing.
Wizard of Whatever
Registered Senior Member
It's only as good as how much it helps you along your way.
exchemist
Valued Senior Member
From today's BBC website:
Police chief admits AI role in incorrect report that led to Israeli football fans' banpublished at 09:11
West Midlands Police Chief Constable Craig Guildford has apologised to MPs, admitting AI provided incorrect evidence in a report that helped inform a decision to ban Maccabi Tel Aviv fans from a match against Aston Villa last year.
The report referred to a match between Maccabi Tel Aviv and West Ham that did not exist.
In a letter to the chair of the Home Affairs Select Committee, Guildford says the information "arose as result of a use of Microsoft Co Pilot" - he had previously insisted the force "do not use AI" and a Google search provided the information.
UNQUOTE
Yet another example of botshit, being used in this case by a police department being too lazy or thick to check the info provided by (in this case ) Copilot. Little excuse actually, as we know from Write4U's posts, that Copilot even offers you a list of references it has consulted. So a quick read through those would have shown this footie match was not referenced. Copilot made it up.
Police chief admits AI role in incorrect report that led to Israeli football fans' banpublished at 09:11
09:11BREAKING
West Midlands Police Chief Constable Craig Guildford has apologised to MPs, admitting AI provided incorrect evidence in a report that helped inform a decision to ban Maccabi Tel Aviv fans from a match against Aston Villa last year.
The report referred to a match between Maccabi Tel Aviv and West Ham that did not exist.
In a letter to the chair of the Home Affairs Select Committee, Guildford says the information "arose as result of a use of Microsoft Co Pilot" - he had previously insisted the force "do not use AI" and a Google search provided the information.
UNQUOTE
Yet another example of botshit, being used in this case by a police department being too lazy or thick to check the info provided by (in this case ) Copilot. Little excuse actually, as we know from Write4U's posts, that Copilot even offers you a list of references it has consulted. So a quick read through those would have shown this footie match was not referenced. Copilot made it up.
DaveC426913
Valued Senior Member
So now we enter the second phase of this nightmare. Who is the authority on general facts?
We go to the internet for most of our news info now. How are we to tell fact from fiction, if the fact-reporters of mundane things such as football matches can't be trusted?
Who should the police officer have consulted with? Do they ha to go righ to the source for every detail?
exchemist
Valued Senior Member
Well not AI, obviously. But this has been blindingly obvious for a quite a while now. So it is pretty inexcusable that the police, concerned as they are with law enforcement, public order and criminal justice, should wittingly or unwittingly rely on unchecked AI output in their decision-making. Maybe they were under pressure and took shortcuts, or maybe the staff involved lack training in use of AI, but shoddy work either way.
Going "right to the source" is no more of an imposition now that it was before we had AI, surely? So I don't see that the police have anything to complain about. AI can still save a lot of time hunting for information - but one does have to read the sources it consults before relying on it for significant decisions, that's all.
But I don't know the ins and outs of this particular case. No doubt there will be an investigation, especially if this Chief Constable loses his job, as he may well. So we may get a better understanding of the circumstances in due course.
DaveC426913
Valued Senior Member
Practically, the police cannot be tracking down individuals and re-doing interviews that may have already been conducted.
At some point, there is an official record on-file with some organization, somehere, that acts as the authoritative record of the event. The Team Manager or Owner might be the authority in this particular matter.
Right, Which is why we have trials, where the accused can challenge the authoritative account.
Indeed. Something I try to evangelize every day.
I am not prepared to declare what they "do" - as in: habitually. That seems a bridge too far, and uproveable.
Certainly, it's what happened in this case, but so far, I'd call it an exception not the rule - until shown otherwise.
Last edited: