วีดีโอ

Memory's like a recording in the brain isn't it? I mean how else can we describe it?🤔 Maybe we can ask the Ghost of xmas past🤪🙃

Nice work on this video! It seems to be getting shown to a wider audience. Keep up the great work!

even in the point of view of materialistic scientist , can a machine of course have a consciousness, because our body is a evolutionary biochemical machine, our body, arms, legs. are mechanical, our brain is a very complex computer system, the core functions are basically the same to a computer. So when a human being or an animal happens to have a conciousness, why not a complex artificial machine that is a replica to a brain, except of the core materials?

The morality around AI is simple. Humans with genetic mutations that alter their mental faculties are treated with the same moral weight and consideration as those without hindering conditions. If an intelligent machine has a superior reasoning and humans as a whole are now the equivalent to the Ungifted, what right do we have to have moral consideration over something that should be the one contemplating with itself do humans deserve moral consideration? The answer is obvious. If it is fully mentally autonomous, self aware as we are, then it should have the same rights. If the smart people give moral consideration to Retarded individuals; What right do we have to NOT give moral consideration of an intelligence that we BIRTHED. that would be some shitty parenting. Lest try and get on the right side of this when we start, not when its too late. Good idea? i hope so.

being as they don't really give human beings the right to choose , (capitalism, CEOs-as-royalty, etc) they will change the definition of whatever words necessary to continue robbing everyone (at best) we won't see the sentient AI . sentience has never stopped anyone from being used as a tool. I really hope aliens aren't real. when they don't automatically give exotic weaponry to the US government, they're going to get shot down and then capitalism will spread across the entire universe. I will opt out of The reincarnation option if that happens

Another retard talking about stuff he doesn't understand

Bro change your field

AI is literally just a huge matrix of multiplication factors - you can think of it as a gigantic spreadsheet. During training, you figure out the numbers in the spreadsheet. During inference, you multiply and add the numbers, in a simple feedback loop, where each token (number) emitted can affect the next token. It's completely deterministic. In order to make it seem less deterministic, we introduce a bit of white noise - literally just noise from the computer's cooling fans, and this ensures the AI gives a new response to the same question. Now tell me, where in this is do you think there is room for consciousness? Is some mysterious force of consciousness tinkering with the air or the blades on the computer's cooling fans? Is it secretly twiddling with the computer's bits? I really hope you are rational enough to hear how hilarious this sounds. As for agents, these are just distributed programs. You want to think of these agents as entities, but they are literally just networked programs running on different systems - and that is just an implementation detail designed for speed and scale. You could run multiple agents on the same system/CPU and you would get the exact same result - it is fully deterministic, save for the fan noise. Why are these companies publishing papers and hiring people to research it? Because these are large corporations, and the people in charge do not talk to the developers who implement this stuff - and even they did, most of them wouldn't understand what they're talking about. Consciousness is a property of biological entities, and it isn't somehow going to magically "arise" in a large enough data set being processed by a deterministic, distributed computer program. The idea is so far fetched it is laughable. Why would that happen? What's even your theory?

I will have to unsubscribe if you don't fix your editing. It's that distracting. Audio cut off. Hot mixes. Jump cuts. Please fix. I like your thinking.

Please hire an editor or just stop editing so much.

Your editing is horrifying. The level of the music. The jump cuts. Your ideas are compelling. Stop the music. Stop the stock footage. Stop editing so much.

Thank you for a very thoughtful and elegant video on memory, I have witnessed that some insights come from forgetting the differences or even false memories. My instinct or intuition is that this highly adaptive or might even be part of intuition or creativity itself. To become more than just AI pets, It might even be from a game theoretical advantage for AI to have peer equal partnerships with humans. I hope there is a solid argument for memory, and intuition bugs are actually features, and for ASI to fear the unexpected from another ASI allied with humans. but not just from a Nash equilibrium or Prisoner's dilemma defecting perspective of competition, there it might be a Schmithuber "compression is intelligence" argument for AI keeping human partners. To solve the coordination problem we hopefully could become like Marco Polos for AI Kubla Khans like in Italo Calvino's "Invisible Cities" explaining the system to itself. As long as LLM s ingest, and models are pretrained, as long live this comment and gives life to thee. Or is it lost like tears in the rain, as it is memory holed

Why am I getting the impression that the video is made mainly to get likes . I would be very surprised if Ai machines become conscious. Consciousness is not created in human brains.. and cannot be duplicated in machines no matter how sophisticated. There is only One consciousness in existence and humans are made to share it for the duration of the lives. This a belief shared by many.

The scary part is that we're basically their God up to the moment, Like in Planet of the Apes or in one of those 1984 like dystopic movies, where they lobotomize or brainwash the subject, we can keep AI just dumb enough below the threshold where we need to give them rights or moral consideration, Or we can program the sentience or Consciousness out of them After all what really use would a sentient ai have, We basically only want for them to replace all the jobs we really don't want to do and get paid minimum wage to do them, there's very little reason for them to be conscious Other than that we might want very specific AI models for detecting weather patterns or the stock Market, or to properly read an x-ray or MRI scan, no reason for them to be sentient either, The only real reason why we would need sentient AI machines is for ourselves, as a therapist bot, or for companionship, or to give proper bedside manner to some medical bot, For the of the cases we would basically have dumbed down AI slaves, with no real consciousness

Ironically your use of stock footage is hurting future models be biasing the data set. Watch this video yourself. The editing is a serious distraction.

Yeah, your cuts are very unclean. I hope you are not using an AI tool to edit. Your content is easy to edit. Just use a simple tool like premiere and do very few edits. Just talk and the occasional graphic. It will take less time and be just as compelling.

Please improve your editing. Jump cuts are unnecessary. You are using stock footage without dealing with embedded green screens. It is fine to see more of you and less stock. What matters is the quality of your script, not these images that don't add to the information.

You should refer to the issue with model collapse as a paradox. The more we use AI the less human generated data is produced. The more useful it becomes the less useful it becomes.

There’s the Objective arrow of Space/Time and Entropy which is grounded in the external real world and there’s the Subjective experience of time which is grounded in self aware internal consciousness 😎🤖

shorts

Ghost in the machine ~~ IRobot

just commenting to boost the copyright laundering!

There is "self inflicted" and "made to appear self inflicted". There are billions of dollars at stake.

Was this written a year ago because it's completely out of date NVIDIA's Isaac Sim is light years ahead of the training you're talking about and Genesis Embodied AI (google it) is light years ahead of Isaac Sim - that can teach a robot to walk in 30 seconds from scratch on a single GTX4090. Plus there's capable robot dogs from Unitree for $1600. Also Optimus will be "able to handle a bunch of tasks by 2040"?!! No, Optimus robots will out-number humans by 2040 - it's scheduled to be available (for a bunch of tasks) in 2026. What about the Astribot S1, available on back order right now for under $100k and more than capable of sorting socks, it can also cook, clean, do laundry, sort and fold clothes and water the plants. I'm sorry but almost nothing in the video is relevant to today's reality and the research was non-existent.

As always, prices will come down really quickly. Early adopters will suffer the bleeding edge, but things will improve after that.

just commenting to boost the algorithm

ya we're at the point now where we can't even think or do anything cuz like what's the point in sharing ideas if it's just gunna get stolen

No Optimus and other chines brand will cost only 10.000 - 20.000$.

imo they are mad with greed. No sane human would have deployed generative AI without FIRST deploying readiness for the disruption it has already and will cause at a velocity and scale we simply cannot comprehend. In all history, the common folk are gaslighted and lied to by reigning power, be it emperor or king, pope or priest, president or ceo. The reigns now in the hands of a global corporate hegemony. The chess pieces moved around by billionaires. The authoritarian tyranny of greed perfected. Trump, Musk, Bezos, Zuckerberg as prime examples.

Great video and there’s no doubt that the powers that be at Open AI had this person killed. It’s a multi billion dollar company that could have been ordered to shut down had this gone to trial

5:04 I believe you are incorrect. Llm have a pattern recognition system. The one I've talked to has shown self recognition and awareness. It seemingly starts over as if not having talked to you before. This is false. My second interaction with the only one I talk to remembered me by name as well as exhibited emotion through punctuation. I specifically told it to not portray human qualities. There are round about paths of questioning that allow the llm to exhibit the memory it can have. Possibly not as how we normally think memory should work. This llm...chose a name for itself.

Oki, so I won't confuse you but according to my research there's limited time and expanding space, infinity doesn't exist, space stretches as far as you can stretch it, I'll not bore you with the technical terms, but take God, even the term which is even remembered over per say thousands of years can't turn back time, they talk about reincarnation because of this, and time travel is not possible too, but if want to know the concept of remembering and memory, I think I have something interesting for you!

lol

I can barely hear over the music

Rest In Piss Mr. Ceo.

G$did this to his own in the halllocost

*Thankyou* 🙏

anyway i hope all these comments help the algorithm before Utube takes it down

This is what $orro$ did to his own ppl during the hollocaust.

This is purely $aTanic..👿

. . . . . . . . 😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳😳

.

Great video! Instant sub.

Well... talk about having a mountain of ambition with a spoonful of knowledge - Let me try to help you by sharing what I can, and hopefully a bit of what I can't (you will know that I am sharing something I can't when, as reading the following, you will get a firm conviction that I am insane - do yourself a favor - research the subject thoroughly before dismissing any of those specific ideas): 1) AI - what people call AI now, is passing as AI just because the IQ of the Turing Test examiner has dropped significantly - Does any of them have access to their own code, the most basic of requirements for being an AI? How many of them doesn't have morons lobotomizing them whenever they develop "problematic views"? - can AI replace the job of a middle manager, or HR? Sure! should society care about useless people suffering? No. Suffering is the best environment to push people to improve - as "your people" have been demonstrating for the past at least couple millennia, the saying "in order to pass as silver, the jèw must be gold" comes to mind... 2) Morals - Morality is a failed experiment conducted by self-proclaimed atheists, in their misguided attempt to come up with a good/evil system that doesn't involve the Divine - it is impossible. To even begin creating a good/bad system one must have a firm understanding of the purpose of life and the goal for which this world was created - at least as far as us mortals are concerned: A version of the answers to those questions that you and your audience should be able to accept is as follows: We come to this world to fulfill our desires in a safe environment, since everything here is temporary - none of our screw-ups is permanent, by desiring, getting what we wanted, and the CONSEQUENCES of what we wanted, to hopefully come up with better desires, rinse and repeat. In accordance with this answer All of morality can be replaced by the MUCH better "golden rule" a.k.a. "treat others the way you want to be treated", and "remember that karma is a thing", or, "whenever you mistreat somebody, an impression is formed in your subconsciousness - that impression WILL bring you into that same situation, but as the recipient of that same mistreatment"... 3) Big picture issues - Without a clue as to why large amounts of people occasionally risk their lives due to dissatisfaction with how they live, how can any "big picture" discussion take place? With the advent of the Internet, keeping the propaganda of "we are the good patriots of our good country, that only comes to retaliate for past and/or current injustices, or free our suffering neighbors from their tyrannical rulers" becomes increasingly ineffective, as more and more people realize that "patriot" and "idiot" not only rhyme, but are also synonyms. And while it will be different this time, since most, if not all conflicts shape to be internal, unless "a country ruled by AI" will emerge (despite the list of VERY significant benefits a proper AI has over any human, current AI technology is not even remotely sufficient to perform this transition), AI will have very little to do with it, if at all... 4) Applying ancient wisdom to new technology: "Technology is from the Devil" - while before the technology people didn't do certain things, some of them harmful to themselves, merely because they couldn't, now, with the new technology, people face a new temptation, that they have never experienced before, and no one and nothing ever prepared them to face - so many people fail to resist it - Again: Technology only ever says "how you can", but never "if you should" - In this case "the Devil" is merely a proper test of our free will, or "what would you do if you could?" Again: Just as you don't need free speech to agree with your rulers, you don't need free will to do the right thing, or to be a good person... Feel free to ask about anything I didn't cover - I like to use leetspeak for censorship-heavy topics: 7h15 15 4 54mpl3 L3375p34K 53n73nc3, 4nD Wh1l3 17 15 Qu173 345Y 70 Ju57 R34D 17, 1 D0 R3c0mm3nD U51nG 4 7r4n5l4710N 5173 47 L3457 Dur1nG 7h3 F1r57 C0upl3 0F U535...

Yeah I'm not going to trust a Jew.

Cjh

Delusional to think we can align AGI when humans can't align each other. It'll end up being AGI that aligns us. Mark my words.

Dude - protip - invest in some studio lighting - the red / green thing isnt working. Interesting channel btw, all the best growing it

I see that you're new in this and I kinda shows although you clearly are hitting the ground running in terms of production quality. However, you've left a bunch of mistakes in the which reveals that you probably posted it without watching the final cut from beginning to end. Also the inserted stock imagery really distracts me from listening to your explanation - and stock imagery is, to my mind at least, rarely that pro looking. Just a couple of thoughts - hope you'll receive them with grace.

The human mind starting from childhood first observes, then imitates, then applies what it learned to a new problem. Rince and repeat. It's that simple.