All of what chatbot produces are fictional depictions.
I feel I should also make you aware that it is in fact illegal to suggest someone murder someone else. Saying “I wasn’t serious” or “I thought we were just role playing” will not hold up in court.
(We all know that if you repeat a lie long enough you will really begin to believe it, right? This is a repeatable and testable thing)
> I feel I should also make you aware that it is in fact illegal to suggest someone murder someone else. Saying “I wasn’t serious” or “I thought we were just role playing” will not hold up in court.
When a person was doing it with intent then it's a crime. But when a piece of software witout any human control generated a piece of text ... Hard to tell.
Heck even if an artist says something that inspires audience to do crime they are usually not at fault.
A book or movie does not talk back to you.
A book or movie is a deterministic system. It is static.
If you cannot see the difference you are willfully being obtuse. I’ll make this claim even if you truthfully believe there’s no difference. That has no impact on my claim.
I see the difference but in my opinion it goes the other way than you think. Or at least begins to.
Personal conversation with anonymous entity is finally revealed to be the least credible source of information. Thanks to AI.
If you believe they think: then it’s obvious. There’s a real connection.
If you don’t believe they think: you can recognize how others do and talking in a human way is often misunderstood as thinking.
Are we really surprised machines optimized for human preferential output are resulting in machines that people love, trust, or view as a friend?
Elder generation thought face to face conversation is usually sinciere. But the culture of massive commercialization of everything changed it. Our generation learned that if unknown person talks to you they are most likely lying to sell you something. But we grew up with honest Internet so we see online chats as honest so AI might trick us. Next generation will know that online conversation might be AI bs.
Everything is fluid. The world moves on. People learn new things and adjust their valuations.
The perception of adults that the world goes to hell all the time is as old as time itself.
> Elder generation thought face to face conversation is usually s̶i̶n̶c̶i̶e̶r̶e̶sincere.
Objectively, face-to-face communication IS more sincere, in the fact that there's more communication. But you're also not wrong because people are frequently more honest over text. Since they abstract the person they're communicating to and this reduces their fear. We all wear different masks in different settings. You're not the same person you are to your close friends as you are to your family as you are to your strangers. None of this is "fake" but you are many. But we probably have more faces now than most people had in the past.
If anything, it is the current generations (mine included, and I say this as an introverted millennial) lack of connection that makes text so much more impactful. Often not knowing what to do with the extra information and with the boundaries of sincerity being much fuzzier over text. I mean people can make a modest proposal, preference that way, and people will still not understand the sarcasm. People that are terminally online!
> The perception of adults that the world goes to hell all the time is as old as time itself.
The world isn't _just changing_, it is RAPIDLY changing and it shouldn't be surprising that people are struggling to make sense of it all. But that also doesn't mean you should over simplify things and paint everything with broad strokes. Abstraction is good and helpful, but details also matter. You need to consider both.
That has been true for at least 3 generations. They experienced in their lifetimes creation of electricity, gas utilities, radio, tv, flight, public transport, medicine, global transport network, vaccines, cars, computers, tanks, bombs they got exposed to in world wars. Nukes. Llms are comparatively benign and limited in scope and impact so far.
It's a bit self-centered to think we are uniquely challenged by changing world when things were rapidly changing for last 200 years.
Willfully being obtuse seems to be common here on HN.
When it happens, I refuse to accept that people actually believe what they're espousing. I think they're playing Devil's Advocate and thinking they're being thought-provoking and creating discussion as a pseudo-intellectual exercise.
It's so boring, though, and often outright frustrating.
Yes it is. ;-)
It's just a really shitty one that's insufferable to interact with.
If anything, that’s an argument that the authors of the data it was trained on should be held responsible for what the bot generates.
I'm sure how "statistical patterns" make it okay. It's not like you're rolling a deice and getting random answers.
Don't be daft.
The persona the chat bot is emulating is fictional, but the chat bot itself is not. It’s not a person, but it is participating in a conversation with the user.
If a heating company chat bot tells you to adjust a setting on your boiler to fix a problem, that’s not a fictional depiction of an instruction.
If you talk to a lunatic it is not lunatic who is to be blamed for you being mislead.
AI chatbots are not parts of real world. You are not encountering them unwittingly.
Suppose one person now and then gets help from a bot impersonating the real person. The line between role playing and real interaction gets very thin.
My point being, 'role play' vs 'real interaction' is not a clear cut.
My employer told me that "chatbot" can help me with my work (engineering). /s