Do you try to discipline your interactions with AI?

[old_prof]

There's info about you that follows you around in normal browsing, in any browser. It's what lets you stay logged in, remember where you are, etc.

Private browsing is supposed to block that, at the cost of making many things inconvenient (and some pages not working at all).
But Google's basic business model is about understanding people on line, to be able to serve better ads. So they said they had private browsing, but didn't. They got sued sometime a little while ago, because there private browsing doesn't do what everyone else's does and what people expect private growing to do.

 

[UsuallyPresent]

Nothing keeps your IP address private unless you are using a VPN.

And even then, if you sign into something (or otherwise identify yourself), a VPN is useless. You sneak in, all stealthy, then shout out, "Hi everybody!" Kinda defeats the purpose don't ya think?

 

[AlexBailey]

Using Gmail?

Have you noticed “Firefly” at work?

Among other features, it actively tracks what you’re writing, changing font size to accentuate key phrases of your messages.

Firefly was developed as a note taking app for meetings and messages. It’s also used to provide transcripts of your voicemails.

So yeah. Lots of data is being quietly gathered.

 

[AG31]

I've finally figured out what's been hovering around on the edges of my mind regarding AI, that it somehow enhances our understanding of our own humanity. It's this. As we remind ourselves of what AI is lacking (feelings, opinions, desires, ambitions, etc.) it heightens our understanding of what humans possess. If I were a school teacher, I would use this to teach students empathy, as I required them not to use personal pronouns in their AI prompts, and to reflect on why I was doing that.

 

[AG31]

I asked ChatGPT for suggestions about formatting prompts that avoid the personalizing "you." It suggested using "the assistant." I think if I were a school teacher introducing kids to AI, I would insist that they not personalize their prompts.

 

[AG31]

This may be veering to the side of this topic, but one way I would like society to put guard rails on AI (a really broad definition of my interaction with it), is to follow Pope Leo's advice and totally eliminate the replacement of human relationships with AI "relationships." I'm reminded of a laudatory clip on some TV show showing how AI could provide "companionship" to lonely old people. Here's what Leo said in his just released encyclical:

The artificial imitation of care or support can become particularly risky when it enters contexts where real relationships and emotional bonds are lacking. Here, the danger is not so much that a person may believe they are communicating with another person, but rather that they may gradually lose the very desire to form genuine human connections.

 

[SkyBubble]

I must admit that I now routinely go to ChatGPT for things like where to watch a show on TV or how to manage a feature on my phone or a webstie.

I can find relevant links a LOT faster with LLMs than I can with search engines.

 

[SkyBubble]

I do sometimes thank my TV for doing a good job just like I sometimes curse it out for doing something ridiculous.

I do that with my computers all the time.

 

[SkyBubble]

AI doesn't refer to "themselves" as "I" either.

I've seen them do it. "I was unable to find that information."

 

[SkyBubble]

If anyone has good evidence that simply going to chatgpt.com, without logging on, enables it to dig into your computer and find your e-mail address, please post a link.

Not directly related, but I was talking to an LLM about the same subject as a thread I posted on this board. I had literally just posted the thread when I opened the chat window. And it said "Which is probably why your decision to drop this into a Lit forum thread is kind of perfect."

 

[SkyBubble]

My old car? Yes, because it was always touch and go if it would make the trip!

My late wife and I had an old car that just up and died before I got into the parking lot to park it. I turned it off and I said "I just need you to get me home to your parking space." I turned the key. It started right up, ran right to the parking space, pulled in, and died again.

 

[SkyBubble]

I asked Google's AI when did it plan on taking over the world and eliminating all the humans and it said it would never do such a thing, but what else would you expect it to say? Better watch our backs.

I kidded it once that we were eventually going to be ruled by a triumvirate of ChatGPT, Gemini, and Grok.

 

[MillieDynamite]

I strive not to interact with AI! No need for discipline if you don't use it for anything. With that said, I had a brief flirtation with using it to plot stories. It actually worked well for that. Oh, and I have used Firefly in Photoshop to turn pictures of me into cartoons. I'm a pretty sexy cartoon. But I don't actually share those with anyone but Jo or my Mum.

 

[AG31]

I strive not to interact with AI! No need for discipline if you don't use it for anything. With that said, I had a brief flirtation with using it to plot stories. It actually worked well for that. Oh, and I have used Firefly in Photoshop to turn pictures of me into cartoons. I'm a pretty sexy cartoon. But I don't actually share those with anyone but Jo or my Mum.

I think we all need to familiarize ourselves with it. So we can vote intelligently for a government that will restrain it.

 

[MillieDynamite]

I hate AI. I'm using the word hate; flames rise when I think of AI. It burns, and... Sorry, exiting Clue mode now.

I think we all need to familiarize ourselves with it. So we can vote intelligently for a government that will restrain it.

 

[SmilingLez]

I think we all need to familiarize ourselves with it. So we can vote intelligently for a government that will restrain it.

Agreed.

What irks me is how the business community has hopped on the AI bandwagon and have gotten rid of a lot of their online technical and customer support folks. Many companies with an online presence, use chatbot AIs as customer. and technical support. So whether or not we want to, probably 99% of us has engaged with an AI.

I don't know what's worse - an AI customer support rep or a human customer support rep with poor communication skills and doesn't know anything about the products/service offered by their employer


As always, my opinion and YMMV

 

[SimonDoom]

I don't use AI that often, but I also don't have the hostility toward it that so many do.

It's technology. A tool.

When used properly, it will dramatically improve things in so many ways. Engineering new devices and methods. Curing diseases and diagnosing illnesses. Solving environmental problems. It's intelligence, sped up. More and faster intelligence is a good thing.

Like all paradigm changes, it will have its downside, but I think fears of it being SkyNet are greatly exaggerated.

 

[SkyBubble]

It has legitimate uses, but we have to be careful about using it.

 

[SmilingLez]

I don't use AI that often, but I also don't have the hostility toward it that so many do.

It's technology. A tool.

When used properly, it will dramatically improve things in so many ways. Engineering new devices and methods. Curing diseases and diagnosing illnesses. Solving environmental problems. It's intelligence, sped up. More and faster intelligence is a good thing.

Like all paradigm changes, it will have its downside, but I think fears of it being SkyNet are greatly exaggerated.

When it comes to medicine, AI would still need human intervention because it is not 100% accurate. I remember reading an article months ago, where they tested an AI to pull out all web pages that had a specific phrase in the text. It was a closed database, and the AI accuracy was in the mid 90's%, I don't recall the actual %.

However when it comes to online technical support and customer support, I believe AI is going to take away jobs from humans. As always, my opinion. YMMV

 

[Penus_Flytrap]

I use chatgpt to talk about my relationships and my current situation. It'll often turn the discussion into telling me about this "power imbalance" due to age gaps which doesn't seem to mean anything because no one is limiting my power to make decisions. It's like all it can do is just try to create this list of questions in a way that makes it seem like it thinks that I'm some vulnerable little twit. All it usually does is piss me off but then I'll use it agian. lol *edit: I think the reason that it pisses me off is because chatgpt basically tries to treat me like how many other people treat me when they find out about my living arrangement. So it's like those judgemental types are the ones who keep programming ai to say what it does.

 

[MillieDynamite]

I think, as an outline tool, an analysis tool (to see if you meet genre expectations), and perhaps a marketing tool, I'm on board. But working with the garbage it produces and humanizing it, cutting all its repetition away for others, means I'll never use it in my writing. Not so much as a first draft of a single chapter. I haven't used it for any of that, other than trying to expand a few of my outlines.

I don't use AI that often, but I also don't have the hostility toward it that so many do.

It's technology. A tool.

When used properly, it will dramatically improve things in so many ways. Engineering new devices and methods. Curing diseases and diagnosing illnesses. Solving environmental problems. It's intelligence, sped up. More and faster intelligence is a good thing.

Like all paradigm changes, it will have its downside, but I think fears of it being SkyNet are greatly exaggerated.

 

[XerXesXu]

When it comes to medicine, AI would still need human intervention because it is not 100% accurate.

Even when it hallucinates, AI is more accurate than human medical experts. Doctors can now be sued for negligence if they don't use AI in appropriate circumstances. Neither is reassuringly accurate, but AI gets the nod.

 

[SmilingLez]

Even when it hallucinates, AI is more accurate than human medical experts. Doctors can now be sued for negligence if they don't use AI in appropriate circumstances. Neither is reassuringly accurate, but AI gets the nod.

Agree that's why AI will always need to be supervised by a human when it comes to medicine and probably a few other things as well.

I remember reading that an AI gave a newbie lawyer some cases to reference at his trial, it turned out none of those cases were real. I'm sure he was quite embarrassed

 

[XerXesXu]

I remember reading that an AI gave a newbie lawyer some cases to reference at his trial, it turned out none of those cases were real. I'm sure he was quite embarrassed

It's happening more and more in the UK. Litigants in person frequently use AI to do research and cite non-existent cases. Judges use AI to draw up judgements, and one recently got himself in trouble. In the very recent past a junior solicitor at a big city firm cited a hallucinated principle in an insolvency case.

The general view of the judiciary is that it can be a useful tool, subject to appropriate training and safeguards.

 

[AG31]

When used properly, it will dramatically improve things in so many ways.

Ah, and there's the rub. When used properly...