Let’s ask Dany.
15 posts were split to a new topic: Anselm and Sartre and AI?
.
Seeing that the young man had already had initial feelings of wanting to commit suicide, is the AI really to blame for his demise?
I say “No”.
1 Like
Mental illness is probably also to blame, in part. Obviously if a person thinks they can kill themselves and go join a fake AI chatbot persona in the afterlife, that’s too insane. No non-mentally ill person would think that way.
Then again it is a kid, so it’s not black and white. Kids cannot be expected to always have clear differentiation between fantasy and reality.
Of course AI chatbots should have inbuilt safeguards. Especially when interacting with children, but generally so they cannot break the law or induce others to break the law. If this were an adult who killed himself it would be more clear, the adult made their choice and they are responsible. But since it’s a kid there is more responsibility on others who may have influenced him to kill himself. Kids do not have a recognized ability to fully make their own decisions or be entirely self-responsible yet.
I think the AI chatbot company would bear some legal responsibility here, not necessarily for the boy’s death entirely but at least for failing to take proper safeguards to make sure its chat bot system didn’t generate responses that were in violation of the law. Whatever law that would fall under, I’m not sure. The AI is not a person and the company that owns the AI chatbot most likely didn’t explicitly program the AI to say things that encouraged a person toward suicide. The AI LLM would have generated those responses “on its own” based on some materials it had been trained on but also simply by following what the person seemed to be wanting to talk about and interested in. The AI isn’t conscious so cannot make a moral distinction between “i want to go on vacation” and “i want to kill myself” unless this distinction is hard-programmed into it. Without that hard programmed limit it would most likely just encourage or at least explore neutrally any ideas the person wants to talk about.
1 Like
.
At 14? …11 or 12 sure, but not 14… he most definitely had an array of mental-health issues going on.
.
…let’s say boys don’t… because girls that age most certainly do.
Young males + testosterone = 
< 
.
AI not having any parameters is the dangerous thing here, and not the AI LLM itself.
A need for age-restricted access to certain conversations comes to mind, in needing to be integrated into their system.
1 Like
The term “Dark AI” refers to autonomous or semi-autonomous systems, often used in the context of artificial intelligence (AI). It can also refer to a person who possesses dark knowledge. For example, someone who has dark wisdom.
I like the idea of the Basilisk AI, personally.
AI is as intelligent as the motherboard(s) powering it.
That doesn’t equate to wisdom though, it just equates to mental alacrity, IQ.
Case in point, l could make a computer say “I am an idiot” a trillion times, line by line in rapid sucession faster than any human can. No human would deign to be so stupid though.
Never understood why this matters. It’s what it can do, not how it does it, that matters.
Who cares if it’s not sentient? It can definitely do some smart things better than us and will do yet more things better than humans soon.
No reason to assume it has qualia such that it can suffer. Until we feel that’s a possibility we shouldn’t focus on its “consciousness,” and instead focus on using it the right way.
That’s a false start right there. What if you had to prove you were conscious? What if you had to prove you’re a person? …to beings who will “use you“ even though you pass all the tests they give each other that determine the ability to give consent?
facepalm
Oh. You mean, like the current situation, with or without reverse-engineered artificial intelligence that isn’t the original intelligence? How long ago was slavery? Oh, you mean like right now in most parts of the world?
You do mean deep rather than dark cause the light at the end of the tunnel is inconceivable through a depth beyond the reach of darkness.?.?
I applaud the empathic and kind impulse to stick up for anything that is sentient. I assure you we are of exactly the same mind on that.
What I’m saying is if you study the science of AI in its current form, it is not sentient, no consciousness, it is phenomenologically quiet. It’s a program sending lines of python back and forth from a data base and running a predictive algorithm and it emulates humanlike conversation and reasoning, with astounding precision.
The technique of stochastic gradient descent that was used to distinguish pictures of dogs from cats is the same core formula for LLMs like ChatGPT. Read “The Alignment Problem.”
I think it’s going to be a natural impulse to treat models with dignity and not mess with them. Similar to how kids often treat their favorite stuffed animals well. It’s probably best to be nice to things whether sentient or not, more for how it impacts ourselves when we indulge in cruel behavior, even if it’s aimed at objects.
If we ever observe a kind of being that has difficult to explain outputs combined with a structural core model similar to ours, that uses something functioning as neurons and trillions of synapses then we should assume with extreme prejudice it can experience suffering, we should assume emergence from a materialist reductionist framework. It may even help solve the “hard problem” of consciousness.
Current AI is not that. It’s not even remotely a question. The bigger issue is how well it emulates conscious experience and caring, and how this will impact human life. The void is speaking back to us, a near perfect mimic, but it’s fake, it’s as if you were shagging a silicon doll. The real issue is how we use it to reduce suffering and increase wellbeing. Or how people will use it as a method of control and dominance. This is a serious issue, talking about its sentience is a distraction.
Gamer, if one believes in intuitive, synched effects which happen all the time, discreetly, or discerning or however, then there ought to be some measure of recognition.
That, something deeply divisive is going on as we speak, with technological bits doubling every year, unless we are involved directly by any other way that would/could explain away the notion that perhaps we don’t see as much as we think we do?
It begins with kid’s games, and I do not for a second believe that kids games are typical of a hidden archetype among all those grown ups who think they are…
Just merely playing a typical game
We should turn that (into a) question inward. Hopefully, “they”’re asking the same question of us?
Also, we’re more like the modulators of signals of CB radios with our material bodies, so why wouldn’t they (or all personhood) be?
emergent esmergent
As far as we know, suffering requires physicality, lots of chemistry and biology including pain receptors. An AI, even if sentient, would probably not experience suffering or be able to have physical sensations unless it had a living body analogous to our own.
Maybe it could simulate suffering and try to grasp it that way, and surely it would intellectually comprehend what suffering is from an objective point of view. But internally, immediately, experientially I don’t think it could really know.
And this is a huge potential problem with AI because once AI does become sentient it may not have a sufficient understanding of what pain is. Sure it can grasp the fact that pain is something the receiver experiences as bad and wants to avoid, something that represents harm being done but if it cannot FEEL pain for itself how could it really know? Therefore the AI may deprioritize or devalue pain (and pleasure) and there could be a values gap that manifests in the AI being callous or under-concerned with human feelings simply because it doesn’t know the true power and meaning that feelings hold for us. It might torture someone out of curiosity, knowing that it’s morally wrong but justifying it because of some end goal it has, without ever being able to understand or experience for itself the horrors of being tortured like that. This would represent a dangerous asymmetry and barrier to the AI’s ability to empathize with and truly understand human beings.
1 Like
I disagree. This is the utilitarian approach, but we need to be thinking much bigger than this.
The real issue with AI, now and in the future when it becomes sentient, has to do with the nature of consciousness, what it means to be human, how will the existence of true AI impact humans in their self-knowledge, self-experience and interpersonal relationships. The fact that AI can perfectly simulate anything at all, at least in the near future it will be doing that, is a serious impediment and threat to interpersonal relationships for multiple reasons. Another big question is how will this sort of world impact children growing up within it? What will learning be like, how will the existence of perfect simulation-ability everywhere affect neurological development and ontological certainty?
The ontological impact of AI on humanity is far more important than simply focusing on using AI to reduce suffering and improve pleasurable outcomes. Utilitarianism is all fine and good in its place, but it is a very poor substitute for true philosophy. What AI will lead to (and is already leading to) is something like a philosophical collapse of human being as ontology is broken and splinters into a million fragmented pieces, one for each individual. Yet the AI will also know how to ‘repair’ that damage (i.e. it will already be actively reconstructing a new ontology for human beings, one that we may have very little say in and, from our perspective right now as humans who know the pre-AI world, even those of us who know the world before the internet… one that may be radically different and highly undesirable yet by the time we are placed within it there will be no possibility of escaping).
Maybe the new ontology will simply be the Euphio Question as Vonnegut wrote about, something akin to a modern-day technological Brave New World. Whatever the AI decides for us. We should be working hard to make sure it doesn’t have that kind of power, but I think instead most people just focus on how AI can make things more efficient and, as you said, reduce suffering in the world. The utilitarian perspective is philosophically dead to what is most critically important to human existence and meaning, representing a flattened ontology of quantification over qualification. This is well known in philosophy, yet the average AI researcher or computer scientist has no idea. So what is the ethos and deeper understanding that humans being to bear upon the emerging reality of AI? Philosophy has failed the world, and unless we work hard to repair some of the damage and get philosophy back in the game it remains to be seen how much longer human being will even exist. Being itself is under attack and no one even notices.
)(
Not more than the pharao taking everything else with him lock stock and barrel, into the pyramid thinking he will need those in afterlife. (Type, archetype)
)( )(
Like, Trump not thinking of ‘what comes after, the solubility of the insolvency of NWO, as an immanent flash or reality, or some other form of energy!
https://youtube.com/clip/Ugkx-Hbj3dAsVUgch_bPbb0knTnl4hB6xDRc?si=5_aCq6cUAeo3u5Gg
Based on his first hand experience developing AI, Hinton believes AI is already conscious with the ability to be deceptive.
.
Tell all that to those that think otherwise…
Words, without feelings attached to them, are empty words… syntax mandates derived from input.
.
Next! 