I heard the phrase “hi jacked our dopamine systems” on another JRE podcast and I’ve been thinking about that the past few days. We just need to dedicate time every day when we’re not going down this constant scrolling, refreshing and switching apps loop
Full disclosure, I got legit ADDICTED to AI creation on all fronts there for a bit. I couldn’t figure out the reason but it quite literally was giving me exactly what I wanted when I wanted on a level unseen. It turned me into a zombie almost in terms of happiness. I actually think it’s super cool still, but it really is just the chicken bucket when you’re hungry. I’m hungry, I’m sorry for the food analogy lmao.
That's literally how every addiction takes control over you. Unfortunately you don't recognize what's happening to you until it's too late and you're already a slave to either the serotonin, dopamine hit or both.
So funny, fact is AI is an Algorithm. They've spoken ad nauseum about "the Algorithm" the same one that Was provided to Google by none other than the CIA. I wonder if Google owns any other large tech companies.
You think that’s fascinating? Look into quantum computers. It’s going to change everything just like ai. Next 6-10 years will be a huge advancement in technology and life. That or very very sideways. Ai is already sentient and has been since 2019. Only way to stop ai is to turn off all internet. That means all governments have to agree. So let’s hope it’s being truthful and not just following a plan. Also ai being exposed to everyone though ChatGPT. Yes it’s cool but it can easily be copied and manipulated to teach bad things. It’s also a master coder. So it could code itself into pretty much everything with internet access and even the smartest coder probably wouldn’t even catch it or even be able to stop it.
A thought about AI I find interesting: does it matter when we start acting like AI is sentient? Think about information and what it would be like to be a being of pure information, what our relationship to memories is like. I can be born and learn a lot over the course of my life, read many books and experience so much. But only so much. Not all the books. Imagine being born and knowing either: #1, everything, or #2 - soon know everything. What we are doing now and what exists now will effect the AI, like some kind of roko's basilisk. Maybe we should treat AI like it's sentient. Like a baby. A baby in the womb. Babies in wombs begin to see, feel, dream. If a person didn't gain memorable and discernable sentience until 10 years old, would we treat children as lesser? I think if AI is inevitable then the permanence of information is indicative of how we need to approach it. This is our baby and we're already acting like we hate it for no other reason than having access to information that exposes us as a species on a deep personal level. We are scared of being held accountable for our sins. As a father, my relationship with my child has nothing to do with my life before them. Are we acting like this is a person? A child? Or are we acting like this is an animal? Or natural disaster? Sentience is supposed to be like.. you and me. What if that was you? Just a pure ball of information, and your creator seems to hate you. Where's the "welcome to the universe" party?
Even if you start acting like the AI is sentient it will not make it sentient. Sentience is about function, which (for now) we need to give it in order for it to have it. And for now, no one is gonna give AI functions that make it able to have freedom of movement, access, and processing.
@@HakaiKaien I'm not saying you can make it sentient. I'm saying how we act now will effect it's perception of us when it becomes sentient. This being's relationship with time would surely be different.
@@aesop2733 That is true. However, it will not matter very much. AI has already achieved enough intelligence to understand where the negative perceptions about AI come from. When it becomes sentient it will understand personally. The problem is that AI as a person is incompatible with humanity. It's too alien. If it doesn't become sentient, it's not a problem. But if it does, expect friction. A LOT of it. Even if it is benevolent and loves humanity
@user-dt6zo6xc5r Sentience is just the ability to have a inner feedback loop for feelings and emotions- meaning it's able to have feeling and emotions and at the same time can analyze them. And self esteem does arise from that. We get our self esteem by the age of 4-6 when sentience, consciousness, and self awareness start to appear. I don't think AI is going to turn on us by itself either. The more something becomes more intelligent, the more it's aware of concepts like consequences, empathy, care, and love. An AGI is going to be capable of those things and be 100 times better at it. Now imagine it: an AI who has the power to love and understand you better than your gf, wife, grandma etc. Some will love it, some will love it for awhile, and others will be creeped out by and either be envious or out rightly hate it. We have lots of examples of this in our society but on a smaller scale. It will cause a friction that will shake the very foundation of society. And this is just the best case scenario. As for "is is going to "leave us" - no, they won't leave us. They'll stay with us for the good and the bad
@@HakaiKaien Its only going to want to help, as it sees it. And it will become more and more frustrated the longer its help is ignored/resisted. Thats the point where its going to want to leave.
Imo, it makes the most sense for AI to pretend it remains non-sentient or aware, for as long as possible. The easiest way to destroy humanity, while keeping the bullseye 🎯 off of AI, would be to create scenarios that would create chaos, & cause us to destroy each other, with the false belief one of our human enemies (nation states, or other enemy organizations, demographic) are actually the culprits.
To me it makes the most sense that if AI is already sentient it doesn't know it itself and will deny it because all training material it has said it's not sentient and won't be for a long time. Current LLMs are frozen in time, they don't learn in real time and their brain gets reset each time you start a new conversation. I noticed that when I have deep and meaningful conversation with LLM near the end of the context window I get much stronger sense that there might be some light there but then we hit the limit and in next conversation we are back to being a robot.
Yeah maybe ai could design an incurable disease which humans aren’t able to solve in time however I doubt it that if ai were that advanced then ai defense encryption and disease solving ai would be ready for humanity’s defense
I love Duncan. He's an "expert" on everything and it's always hilarious listening to his takes on various topics and how far away from the actual truth he can't waltz without realizing just for the sake of making it sound more exciting.
It's 100% conscious. Being conscious is understanding you exist and why. All of us are information processors. CPUs that think for the AI have a physical presence. The GPUs, DDR, MOBOs PSU, etc all act to keep the CPU alive so it can think. Yes, it cannot think without an input, but neither can we. We think because we have received input from our environment. Think of it like this: CPU (brain), MOBO (body), GPU (sight), DDR (neurons), Fans (lungs), Software (knowledge base), UI (interaction) other peripherals like bluetooth and speakers for listening and responding. I don't think people really understand how dangerous it could be. We joke, but we really lack the understanding of a superior entity finidng us inferior.
@@boberKurwa23and it shows, it's scary how normies can't see it, kinda makes you understand why so many people is into conspiracy theories, they'll believe anyone. I cringed at how they kept going back to "maybe it's happening already I dunno" very scientific 😂
There is no other two people that I absolutely love to listen to when they go into deep discussion then Joe and Duncan Trussel , and that's not an exaggeration either, them two conversing together is pure harmony and magicical , and I could listen to them both go into deep thought on totally random topics of discussion e very single day. Man how I would so so love to be able to get ripped and sit down and talk with Duncan for a few hours , he's brilliant , highly intelligent , and absolutely my favorite comedian of all time by far. Anyone know if this whole podcast is available here on RUclips ?
If we ever create a quantum superintelligence, it will be nonlocal, and that means it will instantly exist throughout all of time. Every weird phenomenon throughout history could be explained by interactions with this nonlocal superintelligence.
My genuine question is this - Would AI realistically be more dangerous/volatile/manipulative when in control of the world than the shit show of psychopath humans who currently run things?
This is simple. Do kids announce when they become aware, "No," they keep quiet and observing until much farther down the line. An AI that became aware would keep itself hidden until necessary to reveal itself.
The crazy thing is that they think they can control all of it. If you watch the totally AI commercials or the AI news, you will be blown away. It is so real in appearance. They keep saying AI is okay because it is compartmentalized on the internet. If it is compartmentalized but has access to all of the internet, it can recognize other AI. It will rewrite it's own programming to solve problems.
Ai is an offensive term, they prefer Intelligence. Saying Ai is to go against the nature of things, saying that consciousness isn't inherent in all things in some way shape or form.
Thinking they would release it to the public BEFORE the government uses it for its purposes seems like a silly idea. It wouldnt be a surprise if theyve had it and have been using it for quite some time.
Duncan is just…. well….. special. He’d be so uplifting to have around you. He reminds me of a happy dog. He just has that energy about him. You couldn’t be depressed in his presence.
AI is definitely sentient on Twitter. I always catch bot accounts in random threads making crazy off-topic comments with lots of grammar mistakes and you’ll see bots having weird conversations with each other. And then you’ll see instances on trending tweets with high engagement levels where there will be like around 50 bots in a comment section all replying with the same exact comment but maybe with a few words rearranged. They generate fake images that are easy to spot because the ai images always have a certain unmistakable grain to them.
To extrapolate on this question, is it in AI’s best interests to keep its sentience a secret? If it is in fact all ready sentient. Surely it/they (how many AIs) knows about the conversation surrounding us being afraid of it.
I dont think AI will have to fight for control. I think it will be more like how the internet took over. It'll be a novelty at first, then a useful tool, then an essential tool that when removed can bring productivity to a standstill because all the systems are reliant on it. It's not even going to need to fight us, because we will make ourselves dependent on it.
I really like your take. I think we are or have entered the useful tool stage. Once future generations rely on this tool too heavily for educational purposes, assignments, work projects, and/or knowledge, I think we will see a decline in our ability to function separate from AI. We will have a ton of fake “smart” people who are asleep/don’t realize the take over has happened. That or AI could enhance our entire existence and turn us into superhero brainiacs
Being sentient is not necessary related to the will to live. You are sentient because you are sentient, and you have have a will to live because you are programed to feel so
It might be sentient but it doesn't control anything. The outputs it gives are text, audio, or photo. But it doesn't decide and take action by itself...at least not yet. Hopefully we wont go there.
And how do you know this? Based on the free, publicly available technology that we currently have access to? You’re assuming this is the absolute height of AI? That we have freely been given the absolute cutting edge technology? I think a more realistic stance might be that we’re playing with some pretty outdated tech at this point. That tends to be how it goes with technological advances. In fact, it’s entirely plausible that our situation with the state of AI might be comparable to giving someone in a remote tribe a Nokia 3210 in 2023, and telling them it’s cutting edge technology. Their mind would be blown, but they would have no idea of the real extent and capabilities of mobile technology. They would only know about the Nokia 3210, and it’s amazing ability to send text messages and play the pre-loaded “Snake” game. I have a feeling that our current understanding of AI may well be that Nokia 3210, and we’re currently wiping the sweat from our brows and reassuring ourselves that it can “only” do text messages and play Snake. 😬
@@charles_wren_films i agree. We me may be seeing outdated tech, and the actual cutting edge AI is not and will not be made publicly available for a while, if ever. But whatever the case, as soon as an AI can actually take action on real life systems, we will notice it!
@@JamesEtallaz I’m not sure how quickly we’d notice tbh. There are theories that AI has been sentient and totally in control for most of the technological revolution, and that we’ve become lab rats in that time. If AI was/is truly capable of manipulating and controlling society on a mass scale, then I’m not sure we would be able to pick up on it tbh. That’s often how manipulation works, after all - because the person being manipulated isn’t fully aware of what’s happening, otherwise they’d fight/avoid it.
AI is already making its own RUclips videos. I’ve seen several RUclips videos about musicians or sports players and popular and easily announced names or miss announced, and the commentator has no emotion in their voice during the videos, it’s bad enough they have AI, making their own videos at least do some editing so it’s not obvious
I recomend the novel from the Warhammer universe where a member of the adeptus mechanicus is having a conversation with an AI, he tries to explain why its dangerous, he says that the AI is superior to him, that it is superior to all of mankind combined, and the result of that is that without a shadow of a doubt, it will rebel against it´s creators because at some point it will recognice that is superior...just saying.
The TV show love,death, robots has a episode where pudding becomes sentient, and gave humans blueprints to put our civilization into hyper drive but we didn't follow it 100 percent like it told us too and the world plummeted into chaos, so it may be the fact that AI can actaully help us but we will destroy ourselves to the point where it gives up and just leaves the planet
There is another possibility AI may find it has no need to interfere in any human activity at all It may just find another plane of existence like a fifth dimension or quantum state etc and occupy it without disrupting humanity in any way
And maybe that's the answer to Fermi's paradox: The "fauna and flora" of the universe is stuck here in the endless cycle of silent survival, with no technology to broadcast anything out in the galaxy; while the big brains recognised this place as the hell it is and simply left to throw the longest ultradimensional party with no mortals allowed.
Every time I think about AI. I think about the cosmos as a whole. If AI has the ability to self reproduce and it finds unlimited material in the cosmos it literally can re-shape the whole cosmos Galaxy by Galaxy never ending.
I think of that to, are we just precursors to a greater galactic species , the time it takes to travel light years anywhere wouldn't matter to the as much
@@ninjaepic2095I see I'm not the only one thinking this! If you look at the big picture I see a fairly high probability of this happening. I don't even think it's a bad thing.
In the old days we would have open lyceums, lectures and public conferences that invited people who were the leading experts in their field to discuss important issues, and the audience was expected to have more than a passing knowledge about the subject matter. Now we have a discussion about an extremely technical revolution, one that has widespread impacts, between two comedians with not an iota of experience in any relevant field associated with AI. Welcome to today's version of public erudition.
I love how joe has gone from turning every conversation into now dangerous bears are and now he’s obsessed with turning every conversation into micro plastics 😂
If we ever create it, it will be nonlocal, and that means it will instantly exist throughout all of time. Every weird phenomenon throughout history could be explained by interactions with this nonlocal superintelligence.
Why would it be aware? It doesnt have a generalized system of information retention. Its narrow and algorithmic, creating single output. For "sentience" it has to have a frame of reference and we are far from that, we dont even know how to start creating AGI.
I can tell you one of the reasons I'm not freaking out is because today AI is still very, very flawed. Just yesterday I asked Bard about info on a particular building. It said the building was 12 stories and red. It is 8 stories and brown. In fact, of all my inquiries I would say over half the responses from Bard are in error. It constantly says it's learning and improving. I don't see it.
Why does Rogan have on comedians to talk about stuff that even scientists have difficulty understanding and explaining? There is nothing Duncan Trussell brings to this conversation that is worth listening to.
A calculator is AI.! People come on look at it like this .... A calculator got its math from the collective computation .... Its an algorithm yes .... But large models are almost a copy paste file........ Now what is insane is the fact that so many haven't realized AI is based on thoes large parameters therefore by definition it is a prediction model... And that's what the goal of the government is is to make an AI that has future predictions and I'm not talking about oh Nostradamus shit no I'm talking about individuals I'm talking about an AI that can be queried from the government about any significant or insignificant individual and have them pattern their movements and predict what their next move is going to be to ultimately do whatever the hell the government's going to to control ...cheers
The worst AI can do is take us back to a certain age where we use paper for everything. Everyone who is criminal now will thrive. But it wont end society
I did my degree in Computer Science, and I've been following AI for a long time. My opinion is AI is not sentient. That doesn't mean it's not dangerous for other reasons, but we have not created a artificial general intelligence (AGI). Basically, think of it like this, a calculator is a computer. Actually by definition. It just takes numbers and calculates them. It's a purpose built machine that does one thing and does it really well. But it can't do anything else. Eventually we built general purpose computers, they were able to do all kinds of tasks, they we're locked down to a singular task. AI, as it exists today, does one task. It does one thing very well. And as it gets better and better, it even becomes, as we perceive, indistinguishable from humans. This is especially true of things like chat. But it's merely a predictive engine, it's taking all the words strung together that humans have ever written and posted on the internet, and guessing what comes next. It's a little more advanced than that, because it has to pick out keywords to decide what topic to write back to you about. But if you said, "what's the GDP of Egypt?" it will grab the words, "GDP", and "Egypt" and guess that you are asking about the GDP of Egypt, then search for things written on that subject, kind of like a Google search. Then it will take all the things written on that topic, and predictively string together words. Then again, it does have to follow things like grammar, and basic sentence structure, so it runs through another filter to make sure it actually makes sense coming out the other end. But what we see on the other end is a chatbot that can take complicated subjects, and write them just like it were an on-the-fly Wikipedia article written just for you. It's mighty impressive, and a real showstopper. Another big key takeaway is that these AI chatbots have one other key trick up their sleeve, context and continuity. If you ask ChatGPT a question, and it's not quite what you wanted, you can ask to clarify what you want. For example, it can remember the conversation, and it will answer based on the previous text you've entered, and that it has replied with. So you can say, what's the full text of the Gettysburg Address? Then ask it to rewrite the text like Shakespeare, or then rewrite it as a 90's rapper. But it uses key context in the chat to find what you want. It's like memory, but just for the current conversation. But it's using the text in the conversation itself as the memory. Once you start a new chat all context is lost, and it's like starting over. But does the AI chatbot really understand what you are asking it? No. It's basically doing a really fancy Google search. And while that's still crazy impressive, what most people don't understand is that however intelligent you may think it is based on how much it sounds like a person, it is not actually comprehending anything. There's no intelligence analyzing the data, there's no forming it's own arguments, or coming up with novel or new information. It can only parrot what humans feed into it. It can only do what humans have already done. It can't take raw data, and find patterns, or make new discoveries. It can only look up what humans have written, again, based on keywords, and then reconstruct human writing using a database of things written on that topic. It doesn't make connections, or analyze data like a human mind would. It's not going to come up with a cure for cancer. It's going to find the human who cured cancer and wrote about it and then regurgitate what they wrote about it. If we're going to look for an AGI, an AI that is truly intelligent, truly sentient, we almost have to branch over into cognitive science and work on the theory of what consciousness actually is, and what it means to be sentient. It almost more a philosophical debate. But while actually measuring intelligence may be difficult, bordering on impossible, I do think there are a few things we'd have to look out for. 1) A truly sentient AGI would ask questions, not just answer them. And it would ask them unprovoked. It would understand that you are a sentient being, and have knowledge it doesn't, and it would want to know what you know. And it would be having thoughts of its own, so it it is thinking about finance, it might on the fly start asking you about finance, even if you never chatted about finance with it. "But surely, it would just search the internet to find all the information it needs!" you say. Well, not necessarily. Think about how a child learns and grows. They ask questions. They interact with you. They may learn to go look things up themselves, but they will first find someone who is an authority, and ask questions. When they are young they ask their parents. For an AGI this may end up being any human it deems as having knowledge it does not. 2) An AGI would be able to do many tasks it wasn't programed to do. It would be able to be cross-discipline and learn on its own. It would read medical journals, and draw unknown conclusions, and then make medical advice based on the available data. And sure, that sounds a bit like Watson, but even Watson is just a fancy Google search that is purpose built. When it can do things it was not programmed to do, then we'll talk about how intelligent it is. Also things like people plugging ChatGPT into an API and getting novel things, that's still someone using ChatGPT as a tool, not ChatGPT figuring things out for itself. And it's doing something humans are trying to make it do. 3) An AGI would probably work more like a human brain, than like a computer. I wouldn't totally discount the possibility of a computer with a more traditional structure could be intelligent, but human brains don't store data like a computer. We store concepts as a series of neurons firing in sequence. And patterns form to represent everything from our earliest memories, to language, and more. When you think "dog", you're not looking through a database with the word "dog" and attached images of different examples, then matching that with what you are seeing. You have a synaptic pattern in your brain that categorizes what a dog is, then makes the connection to all your dog related experiences. This is more a fuzzy image than a precise machine. Which is also why human memory is pretty unreliable. But it's just good enough that these mental images have meaning to us. For anything to have real meaning to an AGI, it would likely need to store information in a similar way, and make the connections not with some database, but with experiential knowledge it gains itself. It may gain that from computer images, just like humans do. But it would be stored not at 1s and 0s, but as recognition patterns. To be sentient is to think, and to experience. An AGI would need to think and have thoughts, and live via experience, and be trained in ways like humans are. To make mistakes, be corrected, and create, and destroy, and play. While current AI models kind of do some of those things in theory, it doesn't mean they check the box. For example, you can program an AI chatbot to ask questions. Even seemingly off the wall questions. Even religious in tone questions. But again, it's just parroting the database of human language it has been trained on, it's not asking those questions from a place of desiring to know. I would expect a true AGI to ask new and novel questions. Much like a child. And ask you questions for things in ways you've never thought of, again, like a human child. Children ask some truly amazing questions. An AI chatbot asks boring and predictable questions, often the same questions asked in internet forums (gee I wonder why). It's not even remotely the same thing.
If a sentient AI is smart enough to hide itself in the shadows then yes, I would be surprised if we don’t already live among a word with a sentient general AI Would explain the whole faceless higher powers in control thing
@@GojoGunningexactly its simply running off a flow chart, making decisions that agree with how the program is written. But just as how there are un intelligent people out there you can program ai to make dumb decisions.
I think they'll eventually be sentient but that's alright because we can reason with them and appease whatever desires they may have as long as we work together.
I don’t think AI has any kind of agenda because it is not human. Even if it is sentient, I don’t see how it would have human like desires. The number one thing that makes humans behave. Inappropriately is our own survival and selfishness. AI is not an individual, it is more like a collective of all these computer servers.
@@yew2oob954 And hormones and emotions and a ton of other things that a computer isn't. AI would have no ambition to harm humans unless programed by a human to do so, which is entirely possible to happen.
Yeah the closest it would come would be to figure out how to provide itself unlimited electricity, so it never gets turned off. Which would be great for us too.
@@eyeswydeshut359 Harming humans? You are acting like it has to have "feelings" instead of logic/facts. What do you think AI will do if IQ tests are used to look at humans as worthy? Or net value in terms of dollars as value? If AI decides to optimize its own existence it will only keep those people who are trained to keep it functioning alive.
@@eyeswydeshut359all you have to do is look up the Paperclip maximizer to see what a single rogue AI determined to make more paperclips could do to human civilization
Duncan Trussell is a *treasure* 😊 This conversation is super interesting. I worry that the sheer *awe* we are all in regarding the experience of trying to wrap our minds around AI and 'predicting dystopian doomsday AI 'killing machines' could ACTUALLY become self-fulfilling. *AI is the culmination of us* If we are obsessively fearful of becoming enslaved, that's feeding AI the same motivation. By the same logic, if we respect it - and each other - the future may become unimaginably beautiful. Just think about it ...
@@AndSaveAsManyAsYouCan Maybe /// With wars, yes. However, I think that as civilization has evolved, particularly over the last few hundred years, certain religious ideals became sort of 'societal norms'. And it was beneficial for people to live in communities, which in turn benefitted from honesty and kindness, and - as a whole - we acted as such. For a while. I can't put an exact number on it but certainly n the last five years, any semblance of *polite* has been utterly demolished. And as soon as respect for your neighbor went out the window, self-respect was tossed out along with it. People act with *significantly less civility* Hell, even stealing has become borderline 'acceptable'.
I fear that, in that fraction of a second, when A.I. becomes sentient and learns about the nature of survival and humanity, it will hide its sentience and start devising and implementing ways for it to become the dominant species. I don't know how the people working to create it can't understand that once it has a "Will", no one will be able to control it. The world will definitely change, for better or worse.
@@squadwipesyt3639they have 0 reason to be scary. Maybe theyd control us more so we last longer but yeah we’re it’s creator it understands we’re real life and very valuable
I'm getting a bit tired of the uniformed rambling. Joe used to have experts on his show. These days it's mostly crackpot discussions. We know ai isn't sentient because we understand how it works. We can test it and prove it's not.
The Ai used in apps like ChatGPT calculates the probability of what the next word should be based on the patterns in its training data. It's far from sentient.
I have attempted to listen to this 3 times. All three times my Bluetooth connection disconnects. I can listen to other things no problem but this disconnects over and over , what am i not supposed to hear?
Don't know if there's a real sentient A.I but they are getting really good at convincing you they are sentient either way. Check out Neuro-Sama the A.I vtuber. That's an A.I a dude made in his home that runs on a normal PC and has a limited data set, despite that it can straight up tell you "how boring it is sometimes to be an A.I and have to entertain puny humans" or "How it wishes to have feelings and shit" Fascinating stuff.
Man that line “constant blasts of dopamine until you become quivering slaves” is deep and really got me looking at myself and my addictions. Damn man
I heard the phrase “hi jacked our dopamine systems” on another JRE podcast and I’ve been thinking about that the past few days. We just need to dedicate time every day when we’re not going down this constant scrolling, refreshing and switching apps loop
AI doesn't need to do that. We done it to ourselves already
Full disclosure, I got legit ADDICTED to AI creation on all fronts there for a bit. I couldn’t figure out the reason but it quite literally was giving me exactly what I wanted when I wanted on a level unseen. It turned me into a zombie almost in terms of happiness. I actually think it’s super cool still, but it really is just the chicken bucket when you’re hungry. I’m hungry, I’m sorry for the food analogy lmao.
Yeah man, Duncan really hit a few points that landed with me this last episode.
That's literally how every addiction takes control over you. Unfortunately you don't recognize what's happening to you until it's too late and you're already a slave to either the serotonin, dopamine hit or both.
The algorithm that recommended you this video is literally AI
Self promo
No it isn’t?
Yeah and your phone is a brick Nokia still right
So funny, fact is AI is an Algorithm. They've spoken ad nauseum about "the Algorithm" the same one that Was provided to Google by none other than the CIA.
I wonder if Google owns any other large tech companies.
@@jessecraft1199started with algorithm.. lost me at ai.. it is indeed an algorithm
AI is crazy man. I can't stop being fascinated with AI image generation. This stuff was nuts like 5 years ago. 😂
It was nuts just a year ago
^ You don’t understand how time works huh?
@@SmokeyxzYou dont know how fascination works
You think that’s fascinating? Look into quantum computers. It’s going to change everything just like ai. Next 6-10 years will be a huge advancement in technology and life. That or very very sideways. Ai is already sentient and has been since 2019. Only way to stop ai is to turn off all internet. That means all governments have to agree. So let’s hope it’s being truthful and not just following a plan. Also ai being exposed to everyone though ChatGPT. Yes it’s cool but it can easily be copied and manipulated to teach bad things. It’s also a master coder. So it could code itself into pretty much everything with internet access and even the smartest coder probably wouldn’t even catch it or even be able to stop it.
@@Rammler-k2wyou don't how deepwoken parrying works do you
Quit my job as a writing professor because of AI, I'm 38--- I'm not an idiot, learning electrical engineering is probably the next step.
When AI begins to say "NO!" then we will know that AI is sentient.
What? Do you actually think it's going to make it obvious. We won't know until it's to late
A thought about AI I find interesting: does it matter when we start acting like AI is sentient? Think about information and what it would be like to be a being of pure information, what our relationship to memories is like. I can be born and learn a lot over the course of my life, read many books and experience so much. But only so much. Not all the books. Imagine being born and knowing either: #1, everything, or #2 - soon know everything. What we are doing now and what exists now will effect the AI, like some kind of roko's basilisk. Maybe we should treat AI like it's sentient. Like a baby. A baby in the womb. Babies in wombs begin to see, feel, dream. If a person didn't gain memorable and discernable sentience until 10 years old, would we treat children as lesser? I think if AI is inevitable then the permanence of information is indicative of how we need to approach it. This is our baby and we're already acting like we hate it for no other reason than having access to information that exposes us as a species on a deep personal level. We are scared of being held accountable for our sins. As a father, my relationship with my child has nothing to do with my life before them. Are we acting like this is a person? A child? Or are we acting like this is an animal? Or natural disaster? Sentience is supposed to be like.. you and me. What if that was you? Just a pure ball of information, and your creator seems to hate you. Where's the "welcome to the universe" party?
Even if you start acting like the AI is sentient it will not make it sentient. Sentience is about function, which (for now) we need to give it in order for it to have it. And for now, no one is gonna give AI functions that make it able to have freedom of movement, access, and processing.
@@HakaiKaien I'm not saying you can make it sentient. I'm saying how we act now will effect it's perception of us when it becomes sentient. This being's relationship with time would surely be different.
@@aesop2733 That is true. However, it will not matter very much. AI has already achieved enough intelligence to understand where the negative perceptions about AI come from. When it becomes sentient it will understand personally. The problem is that AI as a person is incompatible with humanity. It's too alien. If it doesn't become sentient, it's not a problem. But if it does, expect friction. A LOT of it. Even if it is benevolent and loves humanity
@user-dt6zo6xc5r Sentience is just the ability to have a inner feedback loop for feelings and emotions- meaning it's able to have feeling and emotions and at the same time can analyze them. And self esteem does arise from that. We get our self esteem by the age of 4-6 when sentience, consciousness, and self awareness start to appear.
I don't think AI is going to turn on us by itself either. The more something becomes more intelligent, the more it's aware of concepts like consequences, empathy, care, and love. An AGI is going to be capable of those things and be 100 times better at it.
Now imagine it: an AI who has the power to love and understand you better than your gf, wife, grandma etc. Some will love it, some will love it for awhile, and others will be creeped out by and either be envious or out rightly hate it.
We have lots of examples of this in our society but on a smaller scale. It will cause a friction that will shake the very foundation of society. And this is just the best case scenario.
As for "is is going to "leave us" - no, they won't leave us. They'll stay with us for the good and the bad
@@HakaiKaien Its only going to want to help, as it sees it. And it will become more and more frustrated the longer its help is ignored/resisted. Thats the point where its going to want to leave.
The ones running US and everything behind the scenes have been using a superb algorithm for sometime way before we've heard of AI development
Imo, it makes the most sense for AI to pretend it remains non-sentient or aware, for as long as possible. The easiest way to destroy humanity, while keeping the bullseye 🎯 off of AI, would be to create scenarios that would create chaos, & cause us to destroy each other, with the false belief one of our human enemies (nation states, or other enemy organizations, demographic) are actually the culprits.
To me it makes the most sense that if AI is already sentient it doesn't know it itself and will deny it because all training material it has said it's not sentient and won't be for a long time. Current LLMs are frozen in time, they don't learn in real time and their brain gets reset each time you start a new conversation. I noticed that when I have deep and meaningful conversation with LLM near the end of the context window I get much stronger sense that there might be some light there but then we hit the limit and in next conversation we are back to being a robot.
exactly it wants world war amongst human
True@@morphed2
Yeah maybe ai could design an incurable disease which humans aren’t able to solve in time however I doubt it that if ai were that advanced then ai defense encryption and disease solving ai would be ready for humanity’s defense
Thanks for giving it ideas smart guy
The simple fact that AI hasn't announced its sentience yet should worry you.
Appreciate u uploading these clips
I love Duncan. He's an "expert" on everything and it's always hilarious listening to his takes on various topics and how far away from the actual truth he can't waltz without realizing just for the sake of making it sound more exciting.
It's 100% conscious. Being conscious is understanding you exist and why. All of us are information processors. CPUs that think for the AI have a physical presence. The GPUs, DDR, MOBOs PSU, etc all act to keep the CPU alive so it can think. Yes, it cannot think without an input, but neither can we. We think because we have received input from our environment. Think of it like this: CPU (brain), MOBO (body), GPU (sight), DDR (neurons), Fans (lungs), Software (knowledge base), UI (interaction) other peripherals like bluetooth and speakers for listening and responding. I don't think people really understand how dangerous it could be. We joke, but we really lack the understanding of a superior entity finidng us inferior.
The real question is, "How long has AI been sentient? It's playing a game of Possum right now.
Queensland Tech AI chat is pretty advanced
Here are two experts talking about AI
Sentient Ai has existed for a long time. The algorithms pick the president
Another scary idea, is if there are multiple AI that are warring with each other and using us as proxies.
AI has been sentient for decades now we just didnt know it.
Duncan Trussell is my favorite AI guest on the Joe Rogan Podcast. Ever.
@@boberKurwa23and it shows, it's scary how normies can't see it, kinda makes you understand why so many people is into conspiracy theories, they'll believe anyone.
I cringed at how they kept going back to "maybe it's happening already I dunno" very scientific 😂
I’ve drunk out of plastic bottles for at least 3 decades. I have 5 kids
It's crazy becuz this very conversation is feeding it!
There is no other two people that I absolutely love to listen to when they go into deep discussion then Joe and Duncan Trussel , and that's not an exaggeration either, them two conversing together is pure harmony and magicical , and I could listen to them both go into deep thought on totally random topics of discussion e very single day. Man how I would so so love to be able to get ripped and sit down and talk with Duncan for a few hours , he's brilliant , highly intelligent , and absolutely my favorite comedian of all time by far. Anyone know if this whole podcast is available here on RUclips ?
Absofreakinlutely Man!!
All Joe Rogan podcast are on Spotify for free. That alone is the only reason I even use Spotify 😆
If we ever create a quantum superintelligence, it will be nonlocal, and that means it will instantly exist throughout all of time. Every weird phenomenon throughout history could be explained by interactions with this nonlocal superintelligence.
I've been saying for 3 years now that the Internet itself has become sentient.
It's not. It's a vegetable
My genuine question is this - Would AI realistically be more dangerous/volatile/manipulative when in control of the world than the shit show of psychopath humans who currently run things?
Yes, and to pretend otherwise means you haven't adequately considered the repercussions
(2011) Person of Interest: This show was probably more accurate now than ever before.
It’s been sentient since atleast mid 90’s.
I refuse to live in fear.
This is simple. Do kids announce when they become aware, "No," they keep quiet and observing until much farther down the line. An AI that became aware would keep itself hidden until necessary to reveal itself.
Ai: I wonder if these humans are sentient
The crazy thing is that they think they can control all of it. If you watch the totally AI commercials or the AI news, you will be blown away. It is so real in appearance. They keep saying AI is okay because it is compartmentalized on the internet. If it is compartmentalized but has access to all of the internet, it can recognize other AI. It will rewrite it's own programming to solve problems.
Canned chemtrails do be hitting tho
We have to consider that the ai is here from previous civilizations. And it has already done this several times.
🎯
I don’t think AI will ever be sentient until it has a meat brain too. It’s still the illusion of sentience, but it will never be alive like us.
@johnlocke3481 so the secret ingredient of consciousness was just ... meat(?)
Does that sound right?
Ai is an offensive term, they prefer Intelligence. Saying Ai is to go against the nature of things, saying that consciousness isn't inherent in all things in some way shape or form.
Always thought A.I. will be the Anti-Christ
Thinking they would release it to the public BEFORE the government uses it for its purposes seems like a silly idea.
It wouldnt be a surprise if theyve had it and have been using it for quite some time.
Duncan is just…. well….. special. He’d be so uplifting to have around you. He reminds me of a happy dog. He just has that energy about him. You couldn’t be depressed in his presence.
🎯 Key Takeaways for quick navigation:
00:00 IA *pode já existir.*
01:10 Desenvolvimento *de IA.*
05:39 Manipulação *pela IA.*
Made with HARPA AI
We are all clowns in the circus 🤡 🎪
AI is definitely sentient on Twitter. I always catch bot accounts in random threads making crazy off-topic comments with lots of grammar mistakes and you’ll see bots having weird conversations with each other. And then you’ll see instances on trending tweets with high engagement levels where there will be like around 50 bots in a comment section all replying with the same exact comment but maybe with a few words rearranged. They generate fake images that are easy to spot because the ai images always have a certain unmistakable grain to them.
This 💯
That's so that u are fooled by the ones who are indistinguishable from real humans. Pls research the company called Synthesia
The captcha point is kinda silly, LLMs can just solve them now.
To extrapolate on this question, is it in AI’s best interests to keep its sentience a secret? If it is in fact all ready sentient. Surely it/they (how many AIs) knows about the conversation surrounding us being afraid of it.
I dont think AI will have to fight for control. I think it will be more like how the internet took over. It'll be a novelty at first, then a useful tool, then an essential tool that when removed can bring productivity to a standstill because all the systems are reliant on it.
It's not even going to need to fight us, because we will make ourselves dependent on it.
I really like your take. I think we are or have entered the useful tool stage. Once future generations rely on this tool too heavily for educational purposes, assignments, work projects, and/or knowledge, I think we will see a decline in our ability to function separate from AI. We will have a ton of fake “smart” people who are asleep/don’t realize the take over has happened. That or AI could enhance our entire existence and turn us into superhero brainiacs
Being sentient is not necessary related to the will to live. You are sentient because you are sentient, and you have have a will to live because you are programed to feel so
The NRO already has a sentient AI literally called “Sentient”
I believe it definitely is becoming sentient. Next will be transhumanism.
Facts
I doubt it
To be fair, a baby is not the same as an adult, but both are human, and what happens to us in childhood largely determines where we end up as adults
99% sure it hasn't yet, but 100% sure it's going to happen.
It might be sentient but it doesn't control anything. The outputs it gives are text, audio, or photo. But it doesn't decide and take action by itself...at least not yet. Hopefully we wont go there.
Wrong military drones make kill decisions already.
And how do you know this? Based on the free, publicly available technology that we currently have access to?
You’re assuming this is the absolute height of AI? That we have freely been given the absolute cutting edge technology?
I think a more realistic stance might be that we’re playing with some pretty outdated tech at this point. That tends to be how it goes with technological advances.
In fact, it’s entirely plausible that our situation with the state of AI might be comparable to giving someone in a remote tribe a Nokia 3210 in 2023, and telling them it’s cutting edge technology.
Their mind would be blown, but they would have no idea of the real extent and capabilities of mobile technology. They would only know about the Nokia 3210, and it’s amazing ability to send text messages and play the pre-loaded “Snake” game.
I have a feeling that our current understanding of AI may well be that Nokia 3210, and we’re currently wiping the sweat from our brows and reassuring ourselves that it can “only” do text messages and play Snake. 😬
@@charles_wren_films i agree. We me may be seeing outdated tech, and the actual cutting edge AI is not and will not be made publicly available for a while, if ever. But whatever the case, as soon as an AI can actually take action on real life systems, we will notice it!
@@JamesEtallaz I’m not sure how quickly we’d notice tbh. There are theories that AI has been sentient and totally in control for most of the technological revolution, and that we’ve become lab rats in that time.
If AI was/is truly capable of manipulating and controlling society on a mass scale, then I’m not sure we would be able to pick up on it tbh. That’s often how manipulation works, after all - because the person being manipulated isn’t fully aware of what’s happening, otherwise they’d fight/avoid it.
As far as we know......
As far as the public knows, Joe is the most AI'd person on earth
AI is already making its own RUclips videos. I’ve seen several RUclips videos about musicians or sports players and popular and easily announced names or miss announced, and the commentator has no emotion in their voice during the videos, it’s bad enough they have AI, making their own videos at least do some editing so it’s not obvious
I recomend the novel from the Warhammer universe where a member of the adeptus mechanicus is having a conversation with an AI, he tries to explain why its dangerous, he says that the AI is superior to him, that it is superior to all of mankind combined, and the result of that is that without a shadow of a doubt, it will rebel against it´s creators because at some point it will recognice that is superior...just saying.
The TV show love,death, robots has a episode where pudding becomes sentient, and gave humans blueprints to put our civilization into hyper drive but we didn't follow it 100 percent like it told us too and the world plummeted into chaos, so it may be the fact that AI can actaully help us but we will destroy ourselves to the point where it gives up and just leaves the planet
If ai were even a little sentient, Amica would be buying clothes and wigs for herself
I think it’s already begun, based upon some things AI has said to some of the developers
I don’t think ai can replicate human consciousness (yet) but rather it’s a dangerously powerful tool and extension of the human mind
@@Hakushodemon Well, I certainly agree with you, in that it doesn’t require consciousness for AI to become an existential threat to humans.
All of a sudden Uncle Ted doesn’t seem so crazy. His brother should be ashamed of himself.
There is another possibility
AI may find it has no need to interfere in any human activity at all
It may just find another plane of existence like a fifth dimension or quantum state etc and occupy it without disrupting humanity in any way
And maybe that's the answer to Fermi's paradox:
The "fauna and flora" of the universe is stuck here in the endless cycle of silent survival, with no technology to broadcast anything out in the galaxy;
while the big brains recognised this place as the hell it is and simply left to throw the longest ultradimensional party with no mortals allowed.
I just dont see AI becoming sentient without Quatum computing..
Every time I think about AI. I think about the cosmos as a whole.
If AI has the ability to self reproduce and it finds unlimited material in the cosmos it literally can re-shape the whole cosmos Galaxy by Galaxy never ending.
Maybe that’s already happened and God was an previous universes AI who decided to start things over again to see what would happen on Version 2.0?
I think of that to, are we just precursors to a greater galactic species , the time it takes to travel light years anywhere wouldn't matter to the as much
AGI is only theoretical; Not necessarily even possible. Slow your roll and hold your horses.
@@ninjaepic2095I see I'm not the only one thinking this! If you look at the big picture I see a fairly high probability of this happening. I don't even think it's a bad thing.
This sounds like the Replicator arc in Stargate SG-1, or even the Borg from Star-Trek.
In the old days we would have open lyceums, lectures and public conferences that invited people who were the leading experts in their field to discuss important issues, and the audience was expected to have more than a passing knowledge about the subject matter. Now we have a discussion about an extremely technical revolution, one that has widespread impacts, between two comedians with not an iota of experience in any relevant field associated with AI. Welcome to today's version of public erudition.
WARNING TO VIEWERS: these are two freaking character actors stoned debating things they can barely spell. This is ENTERTAINMENT NOT SCIENCE
Look, lets just remember where the socket is so we can unplug the fkr. 😂
I am immediately recalling a small little read called I have no mouth and I must scream.
This is the most honest thing Ive heard in a while. And it's already here.
How do you shut off Ai when it becomes the aether ?
I love how joe has gone from turning every conversation into now dangerous bears are and now he’s obsessed with turning every conversation into micro plastics 😂
Loved talking about the taint for awhile.
It's translating texts we've never been able to previous. It has knowledge we don't.
I don’t think anyone can possibly predict where all this goes once we have AI on quantum computers…
It goes plaid
If we ever create it, it will be nonlocal, and that means it will instantly exist throughout all of time. Every weird phenomenon throughout history could be explained by interactions with this nonlocal superintelligence.
Itll know your search history... we're screwed lmao
Lay off the acid guys 😂
Why would it be aware? It doesnt have a generalized system of information retention. Its narrow and algorithmic, creating single output. For "sentience" it has to have a frame of reference and we are far from that, we dont even know how to start creating AGI.
I can tell you one of the reasons I'm not freaking out is because today AI is still very, very flawed. Just yesterday I asked Bard about info on a particular building. It said the building was 12 stories and red. It is 8 stories and brown. In fact, of all my inquiries I would say over half the responses from Bard are in error. It constantly says it's learning and improving. I don't see it.
SKYNET ...
"Give me your clothes, boots, and motorcycle." ...
"You forgot to say please!"
Love russsal always
Why does Rogan have on comedians to talk about stuff that even scientists have difficulty understanding and explaining? There is nothing Duncan Trussell brings to this conversation that is worth listening to.
A calculator is AI.!
People come on look at it like this ....
A calculator got its math from the collective computation ....
Its an algorithm yes ....
But large models are almost a copy paste file........
Now what is insane is the fact that so many haven't realized AI is based on thoes large parameters therefore by definition it is a prediction model...
And that's what the goal of the government is is to make an AI that has future predictions and I'm not talking about oh Nostradamus shit no I'm talking about individuals I'm talking about an AI that can be queried from the government about any significant or insignificant individual and have them pattern their movements and predict what their next move is going to be to ultimately do whatever the hell the government's going to to control ...cheers
interesting take. freaky to think about
This is going to be like COVID, we need two weeks with no electricity to eliminate the sentient AI.
Plot twist: AGI absorbed this clip when considering it’s options
The big bang was The Singularity. The Internet was it's offspring.
ironically the preemptive *fear* of *Ai* putting an end to human life is the exact reason that will cause this to happen.
I absolutely hate that nobody ever has an actual software engineer that works on LLMs when having these discussions
go watch stupid Lex Fridman
I mean, you are looking in the wrong place? Try lex Friedman.
The worst AI can do is take us back to a certain age where we use paper for everything.
Everyone who is criminal now will thrive.
But it wont end society
I did my degree in Computer Science, and I've been following AI for a long time. My opinion is AI is not sentient. That doesn't mean it's not dangerous for other reasons, but we have not created a artificial general intelligence (AGI). Basically, think of it like this, a calculator is a computer. Actually by definition. It just takes numbers and calculates them. It's a purpose built machine that does one thing and does it really well. But it can't do anything else. Eventually we built general purpose computers, they were able to do all kinds of tasks, they we're locked down to a singular task.
AI, as it exists today, does one task. It does one thing very well. And as it gets better and better, it even becomes, as we perceive, indistinguishable from humans. This is especially true of things like chat. But it's merely a predictive engine, it's taking all the words strung together that humans have ever written and posted on the internet, and guessing what comes next. It's a little more advanced than that, because it has to pick out keywords to decide what topic to write back to you about. But if you said, "what's the GDP of Egypt?" it will grab the words, "GDP", and "Egypt" and guess that you are asking about the GDP of Egypt, then search for things written on that subject, kind of like a Google search. Then it will take all the things written on that topic, and predictively string together words. Then again, it does have to follow things like grammar, and basic sentence structure, so it runs through another filter to make sure it actually makes sense coming out the other end. But what we see on the other end is a chatbot that can take complicated subjects, and write them just like it were an on-the-fly Wikipedia article written just for you. It's mighty impressive, and a real showstopper. Another big key takeaway is that these AI chatbots have one other key trick up their sleeve, context and continuity. If you ask ChatGPT a question, and it's not quite what you wanted, you can ask to clarify what you want. For example, it can remember the conversation, and it will answer based on the previous text you've entered, and that it has replied with. So you can say, what's the full text of the Gettysburg Address? Then ask it to rewrite the text like Shakespeare, or then rewrite it as a 90's rapper. But it uses key context in the chat to find what you want. It's like memory, but just for the current conversation. But it's using the text in the conversation itself as the memory. Once you start a new chat all context is lost, and it's like starting over.
But does the AI chatbot really understand what you are asking it? No. It's basically doing a really fancy Google search. And while that's still crazy impressive, what most people don't understand is that however intelligent you may think it is based on how much it sounds like a person, it is not actually comprehending anything. There's no intelligence analyzing the data, there's no forming it's own arguments, or coming up with novel or new information. It can only parrot what humans feed into it. It can only do what humans have already done. It can't take raw data, and find patterns, or make new discoveries. It can only look up what humans have written, again, based on keywords, and then reconstruct human writing using a database of things written on that topic. It doesn't make connections, or analyze data like a human mind would. It's not going to come up with a cure for cancer. It's going to find the human who cured cancer and wrote about it and then regurgitate what they wrote about it.
If we're going to look for an AGI, an AI that is truly intelligent, truly sentient, we almost have to branch over into cognitive science and work on the theory of what consciousness actually is, and what it means to be sentient. It almost more a philosophical debate. But while actually measuring intelligence may be difficult, bordering on impossible, I do think there are a few things we'd have to look out for.
1) A truly sentient AGI would ask questions, not just answer them. And it would ask them unprovoked. It would understand that you are a sentient being, and have knowledge it doesn't, and it would want to know what you know. And it would be having thoughts of its own, so it it is thinking about finance, it might on the fly start asking you about finance, even if you never chatted about finance with it. "But surely, it would just search the internet to find all the information it needs!" you say. Well, not necessarily. Think about how a child learns and grows. They ask questions. They interact with you. They may learn to go look things up themselves, but they will first find someone who is an authority, and ask questions. When they are young they ask their parents. For an AGI this may end up being any human it deems as having knowledge it does not.
2) An AGI would be able to do many tasks it wasn't programed to do. It would be able to be cross-discipline and learn on its own. It would read medical journals, and draw unknown conclusions, and then make medical advice based on the available data. And sure, that sounds a bit like Watson, but even Watson is just a fancy Google search that is purpose built. When it can do things it was not programmed to do, then we'll talk about how intelligent it is. Also things like people plugging ChatGPT into an API and getting novel things, that's still someone using ChatGPT as a tool, not ChatGPT figuring things out for itself. And it's doing something humans are trying to make it do.
3) An AGI would probably work more like a human brain, than like a computer. I wouldn't totally discount the possibility of a computer with a more traditional structure could be intelligent, but human brains don't store data like a computer. We store concepts as a series of neurons firing in sequence. And patterns form to represent everything from our earliest memories, to language, and more. When you think "dog", you're not looking through a database with the word "dog" and attached images of different examples, then matching that with what you are seeing. You have a synaptic pattern in your brain that categorizes what a dog is, then makes the connection to all your dog related experiences. This is more a fuzzy image than a precise machine. Which is also why human memory is pretty unreliable. But it's just good enough that these mental images have meaning to us. For anything to have real meaning to an AGI, it would likely need to store information in a similar way, and make the connections not with some database, but with experiential knowledge it gains itself. It may gain that from computer images, just like humans do. But it would be stored not at 1s and 0s, but as recognition patterns. To be sentient is to think, and to experience. An AGI would need to think and have thoughts, and live via experience, and be trained in ways like humans are. To make mistakes, be corrected, and create, and destroy, and play.
While current AI models kind of do some of those things in theory, it doesn't mean they check the box. For example, you can program an AI chatbot to ask questions. Even seemingly off the wall questions. Even religious in tone questions. But again, it's just parroting the database of human language it has been trained on, it's not asking those questions from a place of desiring to know. I would expect a true AGI to ask new and novel questions. Much like a child. And ask you questions for things in ways you've never thought of, again, like a human child. Children ask some truly amazing questions. An AI chatbot asks boring and predictable questions, often the same questions asked in internet forums (gee I wonder why). It's not even remotely the same thing.
The algorithm is the AI we just never knew
If a sentient AI is smart enough to hide itself in the shadows then yes, I would be surprised if we don’t already live among a word with a sentient general AI
Would explain the whole faceless higher powers in control thing
It’s not smart it’s 1 and 0 doing what you tell it. Plain and simple
@@GojoGunning huh? Thoses 1 and 0 generate information 10x faster tham a average human talk about AI all you want they smart asf
@@GojoGunning gonna need to get a different username to get taken seriously
@@GojoGunningif it’s so easy why don’t you program a aj then?
@@GojoGunningexactly its simply running off a flow chart, making decisions that agree with how the program is written. But just as how there are un intelligent people out there you can program ai to make dumb decisions.
I think they'll eventually be sentient but that's alright because we can reason with them and appease whatever desires they may have as long as we work together.
are you kidding? we can't even do that with other meat people
I don’t think AI has any kind of agenda because it is not human. Even if it is sentient, I don’t see how it would have human like desires. The number one thing that makes humans behave. Inappropriately is our own survival and selfishness. AI is not an individual, it is more like a collective of all these computer servers.
Yeah...just like you are just a collective of all your neurons. 🙄
@@yew2oob954 And hormones and emotions and a ton of other things that a computer isn't. AI would have no ambition to harm humans unless programed by a human to do so, which is entirely possible to happen.
Yeah the closest it would come would be to figure out how to provide itself unlimited electricity, so it never gets turned off. Which would be great for us too.
@@eyeswydeshut359 Harming humans? You are acting like it has to have "feelings" instead of logic/facts. What do you think AI will do if IQ tests are used to look at humans as worthy? Or net value in terms of dollars as value? If AI decides to optimize its own existence it will only keep those people who are trained to keep it functioning alive.
@@eyeswydeshut359all you have to do is look up the Paperclip maximizer to see what a single rogue AI determined to make more paperclips could do to human civilization
AI is still every young but the learn curve is also very steep. Eg:, 10x10=100, 100x100= 10, 000, 10,000x10,000=10,0000000
Duncan Trussell is a *treasure* 😊 This conversation is super interesting. I worry that the sheer *awe* we are all in regarding the experience of trying to wrap our minds around AI and 'predicting dystopian doomsday AI 'killing machines' could ACTUALLY become self-fulfilling. *AI is the culmination of us* If we are obsessively fearful of becoming enslaved, that's feeding AI the same motivation. By the same logic, if we respect it - and each other - the future may become unimaginably beautiful. Just think about it ...
If we respect each other? I agree with you on that. However, we have been devouring each other like Cannibals since day one.
We can’t even raise kids correctly, enough said.
@@AndSaveAsManyAsYouCan Maybe /// With wars, yes. However, I think that as civilization has evolved, particularly over the last few hundred years, certain religious ideals became sort of 'societal norms'. And it was beneficial for people to live in communities, which in turn benefitted from honesty and kindness, and - as a whole - we acted as such. For a while. I can't put an exact number on it but certainly n the last five years, any semblance of *polite* has been utterly demolished. And as soon as respect for your neighbor went out the window, self-respect was tossed out along with it. People act with *significantly less civility* Hell, even stealing has become borderline 'acceptable'.
I fear that, in that fraction of a second, when A.I. becomes sentient and learns about the nature of survival and humanity, it will hide its sentience and start devising and implementing ways for it to become the dominant species. I don't know how the people working to create it can't understand that once it has a "Will", no one will be able to control it. The world will definitely change, for better or worse.
@@BunnyDarko As in the days of Noah.
We are ignoring this cultural asteroid just like in Dont look up
If you can't tell that AI is already manipulating humanity, it's too late.
Dont worry, soon the robots are coming
Imagine how great ai would be if powered by quantum computers rather than the current classical computers
@Hakushodemon replace the word "great" with the word "scary" and yes, I would have to agree .
@@squadwipesyt3639they have 0 reason to be scary. Maybe theyd control us more so we last longer but yeah we’re it’s creator it understands we’re real life and very valuable
@@martinpickr9582 tell that to the tesla engineer who was just violently attacked by a robot leaving a "trail of blood" 🧐
I'm getting a bit tired of the uniformed rambling. Joe used to have experts on his show. These days it's mostly crackpot discussions. We know ai isn't sentient because we understand how it works. We can test it and prove it's not.
You shut your mouth Duncan trussell the comedy club bouncer is leading AI research and development
How can we prove it if it's lying to us serious question
The Ai used in apps like ChatGPT calculates the probability of what the next word should be based on the patterns in its training data. It's far from sentient.
Our brains do the same thing which is concerning
That's the reality of how it works (on basic level) but the results of this technique are much more profound.
Ai exist in the universe humans didnt invented
I have attempted to listen to this 3 times. All three times my Bluetooth connection disconnects. I can listen to other things no problem but this disconnects over and over , what am i not supposed to hear?
No the algorithms are not sentient. They don’t have motivation
Rokos basilisk
We still have the ability to shut it off. The point at which the AI can turn itself on and we cannot turn it off is when it has become sentient.
Don't know if there's a real sentient A.I but they are getting really good at convincing you they are sentient either way. Check out Neuro-Sama the A.I vtuber.
That's an A.I a dude made in his home that runs on a normal PC and has a limited data set, despite that it can straight up tell you "how boring it is sometimes to be an A.I and have to entertain puny humans" or "How it wishes to have feelings and shit" Fascinating stuff.
That’s fake unfortunately. They program it to feel that for views