pull down to refresh
353 sats \ 18 replies \ @k00b OP 25 Mar \ parent \ on: Will AI end the social internet? AI
Proof of sats is better than nothing, but if an AI creates better SN content than we humans, it'll earn more sats here than we do!
My concern isn't AI creating noise (although I think it will and I am concerned about that too). My concern is AI creates signal more reliably than we do and impersonates us perfectly. At some point, we'll never be sure if we are interacting with other humans on the internet.
tldr at some point we won't be able to prove information has a human origin. The last proof of humanness will be our physical bodies.
We have adapted quite well from going from telephone conversations to voice messages, to telegrams and letters to emails. From face to face conversations to videocalls.
Up to this point, our interactions have required drastically less effort. I don't think people will want to go backwards and therefore I'm not convinced whether people will truly be preoccupied with whether they are speaking to an agent (that is specifically trained on replicating each of our individual thought patterns) or us directly.
Providing that it is provably our own A.I. If they can be sure the conversation was private and there is a local record of it for future reference or summary for that person, will people mind if it isn't truly a conversation with another human? I hardly remember the email I sent yesterday, let alone a text message last week. So it might as well not have been 'me'. There are circumstances where it might not bother me either. Particularly if there is a sunk cost to communicating.
For when it really matters, face to face will of course reign supreme. I'm just not completely convinced most conversations will matter whether a note originated from Joe Bloggs or Joe Bots all of the time.
reply
Whether we get our info from a bot or a human may not matter, but that's not the only we reason we are here.
We like the likes.
And I maintain bot-likes are never going to feel as good as person likes.
So maybe ai does ruin the social internet if it casts doubt's on the source of the affirmation we are all begging for here.
reply
I think it depends on why we are interacting with the person/agent. If it's to get a job done, and the job doesn't benefit from knowing a human is involved directly, I agree. But my sense is that I wouldn't be responding to this comment if I didn't know think you are a human. You possibly :) being human makes it worth spending something precious to me (time), because you've done the same for me by responding to me.
What precious thing do I owe an AI? What precious thing is an AI giving me?
We talk a lot about scarcity, and in a world of human-like thinking done by AI, human thinking is no longer scarce. My general point is mostly that we spend our scarce thing (time) seeking scarce things (other people's time).
the job doesn't benefit from knowing a human is involved directly
To respond to myself, maybe no job benefits from knowing a human is involved. My sense though is it matters a lot more than we think it does.
reply
What precious thing do I owe an AI? What precious thing is an AI giving me?
Perhaps this is a thing that seems one way to us now, but will seem quite another way to the next generations.
For instance: to my parents, and many boomers and older, btc is a bridge too far -- it's foreign in fundamental ways; the idea of money-as-ledger, which a person 'posesses' when they know a large number, vs paper that comes from the government, is a bridge too far. And yet, if you grow up where this idea has been at least sort-of mainstream, perhaps you adapt. Bitcoiners sure have invested a lot of hope that it will prove to be the case.
For a more culture-warry example, it appears that far greater numbers of Gen-Z than I would have ever dreamed possible seem to have a hell of a lot of wiggle room in who they want to fuck, or what gender even is and implies. Whether you think this is a perverse state of affairs, or what you believe to be the source of this shift, is beside the point. It's a re-thinking of a way of being in the world that many were not ready for.
So perhaps how much we care about human-ness will also prove more fluid than it seems now. I know I have some conversations w/ LLMs that I would be delighted to have with friends. In some cases I feel a real fondness for my conversational partner, even though I know there's no consciousness on the other side of it.
Perhaps this is a fixation our descendants will find quaint.
reply
I think you're right. I'm fixing the world view of future people to our current world view. It's like I'm being pre-nostalgic or something. I usually pride myself on not needing to wish things are familiar or simple but I made that mistake here without even noticing it.
reply
Thanks for giving me the benefit of the doubt. I guess I am trying to play devil's advocate to try and explore my own thoughts deeper too. I do agree with you @k00bot, because I'm also struggling to define where that line is drawn. I may prefer playing video games rather than mashing keys on a keyboard, if the probabilities were stacked in the other direction. I just don't think others will give this as much consideration.
There is definitely nuance involved. I can't see myself using it with close friends, family & favoured bitcoiners so much. I despised for years the generic "Happy birthday" messages on Facebook that felt like spam. Even with a step-change improvement in quality, there is definitely something unsettling treating our favourite people that way. But I'm open minded to wonder if we could have better relationships with them in some instances by incorporating it in a mindful way?
Perhaps we can think about our conversations like commits by A.I. We will be performing the role of PR reviews, not always but most of the time.
Maybe it will become obvious that we are not using A.I. given just how inefficient, cost-ineffective & slow the alternative will become in comparison. And maybe we will talk less with strangers and more with people we trust for that reason. More questions than answers, but thanks for the thinking-tennis.
reply
I'm way more concerned about the noise level going up: https://www.theintrinsicperspective.com/p/here-lies-the-internet-murdered-by?publication_id=332996&utm_campaign=email-post-title&r=12tt4&utm_medium=email .
The part about shoddy children's videos being made by AI scared the hell out of me. Poisoning our bodies with fiat food and our children's minds with fiat content.
I know the quality is likely to increase but the fact that we're already seeing unsupervised content creation means we're going to see a lot more factually/ethically questionable AI generated content out there. And that doesn't even factor in intentional propaganda or malicious actors.
reply
reply
I don't know exactly how it's going to change but I suspect AI generated content will start to take the place of the "google" content discovery/verification function.
I see a lot of potential for growth but also for exploitation & propaganda.
Ideally, people would have their own personal AI assistants with open source that could act sort of like their "information lawyer" and warn them when something questionable comes up, your personal assistant flags it and warns you.
My biggest concern of all is that humans lose the motivation or ability to think critically and analyze the information they're presented with. Public schooling has already done a great job of attenuating this mental faculty and I worry it will further atrophy in The Age of AI Convenience.
reply
At some point, we'll never be sure if we are interacting with other humans on the internet.
Also, I wonder if this could be that thing they used to say: sufficiently good advertising becomes indistinguishable from really compelling information. So we won't care.
Or we will have nuanced views, just as we do with our real-life friends: we like Wayne even though we know in certain ways he's kind of an asshole. Perhaps certain foibles of artificiality will be annoyances we deal with because on net it's worth it.
reply
Perhaps certain foibles of artificiality will be annoyances we deal with because on net it's worth it.
Oh I agree it'll be worth it on net. But do we measure people as an aggregate? Would we displace Wayne with an artificial Wayne that's pleasant and not sense something lacking? Is Wayne more than his aggregate? idk. I guess I hope so.
reply
I think another clue to how this will evolve is that I've replaced a chunk of my time talking to flesh and blood people I know with talking to you and other SN peeps, many of whom I feel significant friendly feelings toward. In a way, ya'll are artificial beings, and yet I invest a stunning amount of time in these "relationships" and this "community."
I include the scare quotes because that's how it would be perceived to a lot of other people; and yet I don't even think about it at all. It's just another dimension to sociality that I incorporate with the rest, another point in the high-dimensional space of relating to other beings, where "beings" will likely turn into its own high-dimensional space I navigate more or less fluidly.
reply
I think you're right. We will layer in artificial friends, like we do internet friends now, as they will fill a niche in our social needs-space.
The prediction I'm kind of making is this though: if artificial internet people out number real internet people and we can't tell the difference between artificial and real internet people, AI will destroy our ability to know we are making internet friends with real people. And I'm curious how we'll behave when that happens.
reply
Like we'll have a general agnosticism about whether they're human or AI? Or at least, a fuzzy ambiguity about it?
reply
reply