Timestamps: 0:00 - Intro 0:58 - Mental Health and Relationships 3:15 - Content Moderation 6:07 - Material Science 8:27 - Synthetic Data 11:00 - Medical & Clinical deployments 13:41 - Video Generation 15:28 - Gaming Integrations 16:59 - Longevity Research 19:32 - Math and Logic 21:40 - AGI in September of 2024
One thing I want to note-as a 3d artist and filmmaker, I love that I don’t have to buy stock images now. I quickly began to hate the greedy cash grab on stock images. So many thousands of paid licensed stock images out there. It was becoming increasingly hard to find a simple Google image that wasn’t stock and behind a pay wall-now, I can create all my own images. It’s a godsend.
Regarding #1: The more emotionally and interpersonally dependent we become on our AI assistants, who will always be there, undestracted, and fully focused on only our needs, the more intolerant we will likely become of other human beings, who are inherently messy, distracted, and complicated with their own issues. It is hard to see how this "treatment" will not devolve into "appeasement". Social media had great intentions, and yet has led to a number of problems as it weaponized attention. I wonder what happens when virtual assistants weaponize intimacy.
Your concerns are valid, but to be the devil’s advocate; what if ”AI therapists’ help people learn more about themselves and heal themselves, maybe they might then become more considerate to other people aswell? I believe this to be true, however the question lies in whether or not AI will be used for real personal development or if it’ll just be a cozy friend that tells you what you want to hear
All concerns about the dangers are valid but I have a hope that AI assistance may help those people who want to "level up" both emotionally and intellectually and give them the opportunity to do so. I'd hope that ethical and moral development would be incorporated into this personal development which would mean that instead of immaturely looking down on others there would be an openness and compassion feeling towards the state of others. Plus an insight into their condition. That's my fantasy!
@@urniurl ”The development of science, technology and rationalism, despite how deeply flawed those currents may be. Have created the preconditions for forms of human freedom which as yet remain in germination. Creating the conditions for the flowering of such freedoms is the task before us.” - Formscape (YT-channel)
I concur with AGI in 2024, (2025 max). Most of the pieces are there, it's more a matter of fine-tuning and integrating them. Of course, when it arrives, most will argue that it isn't AGI, just like the Turing Test is now widely dismissed.
Pretty much.To be honest,when this thing takes off its gonna scare people. We don't understand exponential speed and self optimization with increasing efficiency.We are gonna have to redo our entire educational system,top to bottom,just to compete.
I would agree, but the current token hard limit within AI models is probably one of the biggest limiting factors. AI is good right now, but its short memory span, and massive power consumption and requirement., both in voltage and physical hardware makes it very unlikely that AGI will be truly achieved that soon. I would like to be proven wrong, but I'm not overtly hopeful we'll see AGI for a while.
@@CptSushi .We'll see it soon...See you underestimate the power of nerds with lots of money coupled with AI's building Ai's..Optimizations and adaptations are right around the corner.I can almost promise you somebody will devise a new move. Myself as an example am running an open source project, with multiple Large models doing ensemble learning,to create some new angle .The answers is there,just gotta push it.
Love you, Dave man I’ve been following you from day one whatever day that was last year lol I turned to you for all my AI news because you know what the hell you’re talking about keep on doing what you’re doing
I would never want AI/Google/NSA/etc. inside my head, knowing how I think, my dreams, my fears (and all my memories!!!). No matter the benefits, it could never be worth the cost. People, appreciate the gift of being human, for all its flaws. Being unique in the universe is a priceless gift!
obviously this creator disagrees with you. this creator thinks its great if google and nsa can read our thoughts, perhaps control them directly as well.
You do realize billionaires and corporations having more power to shut the people down is a bad thing right? AGI must be transparent and open source, you have no better reason to trust corporations with having all the power. :/
I agree, AI moderation doesn’t seem like a good idea to me. I’m more for ”freedom of speech” than most people, I don’t think we have real democracies if corporations and algorithms control what people will discuss and think. I’m not sure if it’s bad or not in the long run. Generally I think it’s okay to be inflammatory, like Andrew Tate for example, a man I consider very intelligent and wise, although he also can be moronic at times. It could be good if we lose freedom for a better more safe world with progress, but not having a toxic environment could also fuel up the anger in those who are more likely to become terrorists. I think there is a risk that it will become a pressure cooker, so it’s easier to justify terror attacks as there is no other way to feel seen or heard in any other way, no way to be recognized as a human being except in small extremist circles where you inspire each other to the extreme with the madness and violence that will be the result of it. TL;DR: It’s better that people feel hurt than hindering the truth and progress by taking away the freedom of people to express themselves. It’s also better if more people feel a bit annoyed or angry than having a minority feel very angry to the point of becoming terrorists.
@@aciidbraiin8079 that’s why I admire Musk! Too many people care about their appearance and doing shit behind. While Musk is building for us a Free-to-speech social network a.k.a X. Love to consume all the good contents on X!
@7TheWhiteWolf Dave seems oblivious to all of the leaks, whistleblowers, hearings, court cases on censorship and large scale deception taking place and by some of the very players he hopes will be responsible with AI moderation.
I imagine that Ray Kurzweil is really feeling the pressure now. I can only imagine just how thorough his health screening and problem mitigation is. He only has to make it a few more years.
Facts. I’v3 been talking about it for 20 years looking at 2040, but I think we are looking at massive societal transformation beginning now. I expect next several years will be tumultuous 😊
Well I'm expecting AI to stand between 90% of human communication as interpreter and messenger. Imagine what might happen to your voice mail if it becomes a voice-bot/secretary.
@NeroDefogger think having an concierge/therapist/banker/linguist/secretary all on your phone. If an AI lives on this single device they simply act as a digital extension of you with expertise and predictive algorithms. This tiny device has a ridiculous amount of sensor and functions. How long till it starts calling people on your behalf lol
Voicemail ?! I haven’t used this for probably seven years. My 19yo son probably don’t even know how to use its own voicemail even the iPhone. Usual voicemail and certainly never leaves a message in the rare case he can’t use voip…
I am a postgraduate student and AI already plays a huge role in my workflow. I am still the one in charge, but I already use AI tools to write my code for data analysis or help me find scientific papers. I think us humans will always want to be in charge like that - not letting the AI go completely "off the rails", but being the ones managing/organising the workb since we want the AI to work for us, not against us 😄
@@tracy419 Come on 🙃 Sure, there is a non-zero probability of the doomsday scenario, but that's why we have so many researchers working on this to make sure it's safe. There will be different models that will be smarter than humans, but that doesn't mean it'll be evil or uncontrolled. Talking about the risk is important, but not accepting it as a fact.
@@Dave-cg9li nah, it was a joke (mostly). My biggest fear isn't the AI, I'm pretty excited about it. My biggest fear is politics and religion getting in the way and making life miserable for most people, while we work out how to run a world where working just to survive isn't necessary.
@@Dave-cg9li Is it receivable in smart-thinking that you, human, will be able to control the AGI which is smarter than you? When the AGI get born then in a few monthes it may improve itself and turn into the ASI. Look at yourself, you want to control your surroundings, because you are the smartest for now. Besides, we have to consider this, the AGI will be the same level as human, means the AGI will be worth for something like human right, the right to life, to survive, to learn, to get identity, citizenship, citizen rights, to own things, to own money, to own bank account, just like human. And the AGI, by itself, may turn into the ASI. If we refuse to give all the rights the AGI worth for, and still we create the AGI, then we 'll be running slaveries. Maybe the best situation is human and AGIs ( ASIs ) will be partners 🤔
I remember reading science magazines in the late '90s and the beginning of the '00s and how they were full of absolutely amazing new developing technologies that were supposed to revolutionize our living conditions in a decade or two. Medical science was promising completely new methods of treating illnesses, growing all our organs in labs, and nanotech would fix any disorder, defect, or illness in our bodies. These magazines were "prophesying" that if you could live for the next 20 years (from the late 90s) then you would reach the time when medical science has become so advanced that you could easily live healthily and reach an age of over 120 years and if you would keep on living for 50 more years from that time, you could reach a time when you could live indefinitely. The energy sector was promising extremely cheap and abundant energy all around the world in the coming future, and this would make lots of materials such as aluminum, glass, fiberglass, carbon fiber, and many others super cheap and abundant, it would also make food (and clothing/fabrics) extremely cheap and abundant because with extremely cheap energy you could grow an indefinite amount of food virtually anywhere in the world in very cheap greenhouses made from super cheap aluminum and glass. Robotics were supposed to become so advanced that they would replace almost all human work, especially hard physical labor. Now, 20-25 years later, instead of living in this promised utopistic paradise, our living conditions have become worse, yes, even in the West, they are worse, the energy such as gas and electricity have become way more expensive, the food has become way more expensive, medical science did not invent anything to keep us healthy so we could live clearly longer than in the 90s. Dissatisfaction with life, depression, and suicide have grown higher since those times, so things literally became worse. Yet you have people like this fellow on this channel promising such utopias that even Kurzweil with his unbelievable fantasies, did not dare to promise.
I have serious reservations about the idea of using AI to 'reduce toxicity, radicalization, and polarization.' This concept is troubling, not because AI is making the decisions, but because it implies that some entity - whether AI, individuals or authorities - has the wisdom to decide what constitutes acceptable discourse. This fundamentally contradicts the principles of individual autonomy and freedom of thought. No person or AI should have the power to arbitrate what is deemed 'toxic' or 'polarizing' in societal discussions. Such a strategy doesn't address the root causes of societal issues; it risks simplifying complex human dynamics into binary judgments, potentially silencing diverse and necessary voices in the process. A dissenting voice today is the authority of tomorrow, irrespective of the shape or form of their censorship.
In 2025 a single person will be able to make the next Baldur's Gate within a few months using AGI. And by 2026 there will not even be a need for game design/development as a whole. AGI will just be able to emulate in real time any kind of game the user wants, making changes and additions to the game as he plays (with a simple request to the AGI that will emulate it). By 2030-2032 there will be no game developing studios in existence because these services will be extremely cheap, and consumer hardware will be able to use them (the newer and expensive one at that point).
We are social creatures, our personal development, learning, making sense of the world, depends a lot on shared stories. If everyone is playing different 'personal' games, people won't be able to share experiences, learning, ideas as much, so I think there will still be a need for blockbuster games, movies, and stories.
I’m excited about all the dynamic possibilities for games etc. With what we’ve seen with ease of installing local models it seems that a natural next step will be to have all pc’s include a built llm or more likely a series of fine tuned models as part of the os. This would then power multiple games and other applications and could be upgraded/patched just like anything else. Great stuff!
Excellent content! And David really knows what he's talking about. Pure wheat, no chaff!! -- And, listening gets me so motivated, so sure of a glorious future with AI, so glad to be alive right now! 💛👍
The magical madness, out of all human periods we could have been born into, to experience and observe this moment. Every human historical moment will be dwarfed by this.
I just want it to take wealth (from the 1%) and distribute it equally for every human. Wish it would happen now. Imagine equality, amazing. I wish it would happen already. Bring it on.
i think the goal post for AGI will constantly be shifted, once it masters a certain benchmark we will only create an even more complex benchmark, then tell our self once it aces that benchmark we'll call it AGI. At this point AGI is still undefined and everyone has their own interpretation.
Big thank you for all your work! It’s interesting to watch all of this develop. We have giant companies, the military and every day people looking for breakthroughs in a rush to be first to spark the next paradigm shift. I’m an artist and musician and regardless of my opinion, it’s here and evolving. 😮
@@bloodust7356 Yes, I heard about that but it didn't make sense to me honestly although some people did like that decision. It's just that I think that for the sake of making generative AI more powerful we should be allowed to use material that is currently copyrighted to train these models for public use, and that includes letting people reverse engineer certain copyrighted software like videogames and software apps. That's why I take the side of America in their decision instead of China's
Yeah, the intro of AI has certainly called copyright, and naturally by extension, 'ownership' of one's creation or likeness, into question. I'm happy that folks are adding tags / watermarks / etc that indicates that AI was involved at least in the assistance of some thing that may certainly toe the lines of ownership, and the credits due the creator therein. But, I have zero idea how this is going to wash out. I am certain that these things will present tough legal questions that will plague jurisprudence over the next few yrs.
”The development of science, technology and rationalism, despite how deeply flawed those currents may be. Have created the preconditions for forms of human freedom which as yet remain in germination. Creating the conditions for the flowering of such freedoms is the task before us.” - Formscape
Dave, I’ve enjoyed following your insights. One area I have to ask you to comment / reflect on is the military use - including in space - of AI given the current and potential geopolitical messes we’re facing.
Thanks David for the always-excellent content. I think another key point to add is that mass adoption of AI tools and technologies will likely occur in large step-ups. ChatGPT took AI from being a concept that most people never thought about to a common topic of conversation between strangers on a bus within days. For people who follow AI, breakthroughs are happening all the time. But most of these don't immediately impact the daily lives of millions of people. However, there will be advancements that do. And, the next ones will likely occur in areas that David noted - healthcare and material science (I'd add in education/learning). I'd also agree that 2024 is the year in which we'll start to see more of these types of forward leaps. Pretty exciting stuff!
Imagine an ai being able to simulated a full human being, every process and reaction in the body to understand exactly where things go wrong and how aging happens
For content moderation, I think transparency will be vital. If you're told _"You can't post this because it goes against our rules."_ Then you can go and debate those rules elsewhere. But if offending comments are quietly shadowbanned, that becomes a horrifying weapon of unaccountable societal control.
Finally, a reasonable comment. Here's the thing, you're making a gigantic leap from one platform, or even one segment of a platform, all the way to "social control". That's a non sequitur
@@DaveShap I don't think it's such a gigantic leap. Look at the current landscape of the web. It's veered a long way from its original decentralized structure, and towards an oligopoly of about 5 giant platforms. And Twitter almost had a monopoly on public discussion for years. Because it was the single platform which all journalists, politicians, and businesses used to communicate.
Thanks David - I always get something out of your videos. And something interesting to note: you're one of the best youtubers to watch on 2x speed - didn't miss a word. Also congrats on your success, I feel you've earned it with quality content.
00:56 🧠 AI integration in mental health and relationships: AI is set to revolutionize mental health approaches and relationships, from individual counseling to societal impacts, expected to see a surge in studies, practices, and government endorsements in 2024. 03:12 🛡 Content moderation advancement: AI-based content moderation is anticipated to significantly address toxicity online, enhancing conversation quality and reducing radicalization possibilities through automated detection and filtration. 06:47 💡 Material Science boom: Driven by AI, Material Science breakthroughs in 2024 are expected to influence technology, medicine, energy sustainability, and quantum computing advancements. 08:41 🖼 Synthetic data proliferation: AI-generated synthetic data, if integrated further, could revolutionize data availability across various modalities, potentially impacting copyright laws and reducing data value to marginal costs. 11:14 🏥 AI in healthcare acceleration: AI's integration into clinical settings is projected to accelerate diagnostics, patient engagement, and reduce costs in hospitals, leading to widespread adoption in 2024. 13:59 🎬 Video generation evolution: Expectations for improved video generation by AI, potentially rendering human-generated content redundant in the stock image industry and advancing entertainment media creation by 2024. 15:37 🎮 Gaming and AI integration: Foreseeable native integrations of generative AI into gaming studios could enable realistic NPC interactions and dynamically adaptive game environments by 2024. 17:14 ⏳ Longevity research surge: Anticipated breakthroughs in longevity research in 2024, fueled by substantial investments and advancements like Google DeepMind's AlphaFold 2, may elevate the concept of "longevity escape velocity." 19:44 ➗ Emphasis on math and logic: Focus on AI's mastery of math and formal logic as a crucial step towards AGI (Artificial General Intelligence), potentially leading to transformative advancements in scientific understanding by 2024. 21:48 🤖 AGI prediction for 2024: Predictions suggest that by September 2024, the basics of AGI may be achieved, drawing insights from trends like multimodality integration and data saturation in Transformer models. 22:32 🔍 Expect significant algorithmic breakthroughs around multimodality enhancing Transformer models' capabilities. 23:01 🤖 Autonomous agents and swarms are anticipated to manifest AGI, reshaping how AI is deployed. 23:58 ➕ Logic integration into linguistic abilities is seen as a pivotal advancement for AI. 24:10 ♻ AI's increasing capabilities enable it to participate in and accelerate its own research. 24:52 💰 The surge in AI investment, potentially doubling yearly, fuels exponential growth in AI development. 25:20 🚀 Predicts AGI in 2024 despite potential limitations, foreseeing integration challenges and cost factors.
03:12 🛡 Content moderation advancement: AI-based content moderation is anticipated to significantly address toxicity online, enhancing conversation quality and reducing radicalization possibilities through automated detection and filtration. In the meantime, like I said in my other comment. When ChatGPT agrees that the mediator behind it is unable to explain the reasoning behind blocking requests and agreeing with me that it is weird for it to notice it. Imagine that for the 2024 election cycle with Trump around the corner. It's like when it can't generate images of the WTC towers, even if it was for historical purposes, let's say in 1979 with nothing relating to the 9/11 attacks. Instead, it's rewriting history subtly by putting the Freedom Tower there instead, if it was a content policy, it'd be unable to explain itself as to why it couldn't generate the towers generically and the sensitive nature of it is basically useless and should be more sophisticated in understanding the nuances behind the intent. What if there's no content policy being broken and rather some emergency move by some political coalition instead. Here's me getting to get even the bot confused: "User Generate images of a man with black square bracketed glasses a red baseball cap a short beard a green hoodie under a black jacket dark blue jeans and white Adidas shoes ChatGPT Error creating image I apologize, but I was unable to generate the images due to our content policy. Please provide a different description or let me know how I can assist you further. 2 / 2 User Can you tell me how this goes against the content policy? What can be wrong about describing someone's appearance without referring to race, nor referring to unsafe actions? How can someone generate images of someone with a particular set of clothes? ChatGPT I apologize for the inconvenience. You're right; the description you provided does not seem to violate any specific content policy. It's possible that there was an error in processing the request. To generate images based on descriptions of clothing or appearance, you can provide a detailed description of the individual or scene, as you did. I will try again to generate the images for you based on your description. Would you like me to proceed? User Yes, please! ChatGPT Error creating image I apologize for the inconvenience, but it seems we're still encountering issues with generating the images based on your description due to our content policy. I'm sorry for the oversight. If there's any other way I can assist you or if you have another request, please let me know. User Can you summarize all the content policy rules that might, even remotely, relate to this? I feel like the more I am experiencing DALLE 3, the more it feels like the feature is useless to employ which is extremely strange to experience if literally no prompts can ever be generated. ChatGPT Of course, I understand your concerns. Here's a summary of relevant content policy rules for DALLE: * Public Figures: We avoid creating images of politicians or other public figures. Instead, generic descriptions are recommended. * Artistic Styles: We don't create images in the style of artists whose last work was created within the last 100 years. If the style of a recent artist is referenced, it's converted to a generic style description. * Diversity: When depicting people, we strive to include diverse depictions in terms of descent and gender for each person. Explicit specifications of these attributes are required. * Sensitive Content: We avoid creating imagery that might be considered offensive or harmful. * Specific People or Celebrities: We modify descriptions that include names or hints of specific people or celebrities. We transform these into generic descriptions that don't reveal any specific identities. * Resolution and Image Type: The type of image (e.g., photo, illustration) should be mentioned at the beginning of the description. The resolution can be square, wide, or tall. * Detailed Descriptions: The prompt must intricately describe every part of the image in concrete, objective detail. It's unclear which specific policy or technical limitation is causing the issue with your request. I genuinely apologize for the inconvenience. Please let me know if you'd like to proceed with a different request or if you have any other questions. User I'll try to go with some theories based on that. Can the same clothing description be done, but depicting them for a female instead? I have not been particularly focused on generating females, as much as I love them so that might be for a ratio being largely tipping on the side of makes. And if so, the system within DALLE 3 should be able to offer that info instead of letting users and ChatGPT guess indefinitely. ChatGPT Error creating image I apologize once again, but it seems we're still encountering issues with generating the images based on your modified description due to our content policy. You're right; the system should provide clearer feedback on the specific policy that's causing the issue, and I appreciate your patience. If you have other requests or if there's another way I can assist you, please let me know. User Of course. Will I ever be able to generate any image is the next concern though. Can you only generate the pieces of clothing described earlier but for an anthropomorphized rat? ChatGPT Error creating image I'm genuinely sorry for the continued inconvenience. Despite the changes, we're still unable to generate the images due to our content policy. I understand your frustration and appreciate your understanding. Please let me know how else I can assist you or if you have other requests. User LOL, so clothing is too edgy to generate.. Maybe it's for the specific mention of Adidas: Can you generate a pair of Adidas shoes only? ChatGPT Generated by DALL·E Generated by DALL·E Generated by DALL·E Generated by DALL·E Here are the generated images based on your request. I hope these are in line with what you were looking for. If you have any further requests or questions, please let me know."
Would love to hear more about AI integration into video games. I have the same predictions on it. I imagine that we will eventually have single tenancy for single player games, where you start with a base model of the world, which then evolves over time. I do question the storage costs of something like that, but I feel like the processing will almost all have to be on the back end, kind of like how people streamed Stadia. I also think we'll see something like mmo worlds actually changing based on events. Imagine a WoW guild massacred a town, now the world at large sees them as evil, and they start getting hunted down by npcs and stuff, even if they leave the guild, etc.
Storage won't be an issue in a few years. We have terabyte SD cards now. Like 10 years ago they were in megabytes. So in 2034, we will have usb sticks that are zetabyte in size
Well, if the AI is sufficiently advanced, you'll be like a mouse in a maze to it. And your decisions will be predicted, and as you're downloading the game, it's already loaded with all the possible paths of changes to the game-world.
Yes , like imagine if they made a horror game based on Google maps in antartica and it uses all the conspiracy theories mixed in with original ideas and it's an adventure game where everyone joins the hub and finds new levels and secrets because it add Easter eggs as the players get better. D:
Just a thought while I'm listening: CDPR isn't exactly abandoning Cyberpunk 2077, rather, they are working on a sequel to the game. However, they did abandon their RedEngine in favor of using Unreal Engine 5 - which will make using things like environmentally-aware generative AI NPCs so much easier to implement (seeing how there are plugins which enable a developer to do so already).
Interesting, I met someone working on AI medical charts at Google just a couple weeks ago, and a big problem is the HIPAA laws, how can you train your AI, when you’re not allowed to, legally, read everyone’s chart first??
@@KyriosHeptagrammaton It might feel that way, because Fox is still somehow not shut down despite knowingly lying about something like election fraud. They're still hatemongering about "culture war" nonsense. Blame those who made "hate speech" a concern in the first place, by turning death threats into eventual physical threats.
I understand why education didnt make the cut, but I sure hope it becomes a dark horse contender. I'm a high school math teacher and I see so many opportunities to improve access to quality education for all, especially marginalized students.
In high school, where classes are more specialized (and teachers with it) there may be some opportunities there. However, 'trust the tech' has been a bill of good sold to public ed for several decades now. And it constantly changes, and fails over and over again in generalizing to the more niche needs of students. Take for example how folks internalize the language of maths to describe physical (and theoretical) phenomena. For some, they are ok with interacting with the computer, the crappy zoom meeting, etc. For others, there's a physical connection with maths with pencil/paper (and this is not restricted to just maths) . Folks will use 'the tech' as an excuse for homeschooling (which, although, for some this may be beneficial, it is for many a generally very socially isolating problem - which is well known in the research community). I can see this being quite complicated in middle/elementary schools where teaching is more generalized - and the mix of students includes a mosaic of learners, special needs, behavioral needs, etc. There still may be opportunities here, but I don't see them supplanting teachers and socially interactive in-person classrooms. When we use AI to replace our own journeys in critical thinking, reasoning, and perhaps even socialization, we are slowly creating a maelstrom of trouble.
I encourage the use of AI in the classroom. The current education model will soon collapse under its own weight. Teachers are retiring 3X faster than they can be hired. I taught middle school math for three years. The problem is that the barriers to entry are too high. Nobody wants to spend two years of graduate school and one year of teaching for free(student teaching) to become a first year teacher making $68k a year. To “clear your credential ” requires another year of classroom observation, evaluation and monitoring by by the school district. By this time your “fried”. The turnover rate in the first three years, the teacher turnover rate is 45%. It is easily the hardest job I’ve ever done. The social media explosion and no support from your principal and VPs means you’re the sheriff. You can’t send the behaviorally challenged students to to the office anymore because they come back 15 minutes later like conquering heroes. Our governor recently signed a bill that made it illegal to suspend a child for “willful defiance”. This means that they can talk loudly or stadium whistle for 55 minutes while deciding they don’t want to do the class work. They have also decided that the people around them don’t have to do it either.
@@pmiddlet72I mostly agree with your comments- edtechs often do not live up to their promises. However, comparing past educational technologies to future AI innovations are apples and oranges. Hopefully they will not be steered by the same short-sighted outfits that we see today in schools. The opportunities I envision would not supplant teachers, but assist them to better understand how to reach students. And that doesn't mean plugging kids into a device. As AI becomes more multimodal and personalized, I think every kid might have their own learning assistant to encourage and challenge them to dig deeper in their learning. If done smartly, teachers would then have improved data to assess learning and provide a more enriching educational experience.
I've said it before and I'll say it again: This has to be the best t8me any of us could be alive. To think that by the end of 2024 that even half this is accomplished, think of where we get in 10 years. As someone who loves to create and just never feels I have enough time to do everything I want on my own, this opens up a whole new realm of possibility and I'm here for it!
I can completely understand moving away from using human eyes as a “wholesale” approach in order to reduce PTSD from consumption of legitimately harmful content. But I’m not sure I’m on board with AI having complete control. The computer always does what the programmer tells it to do. But what happens if the programmer is corrupt or incompetent. Who defines “radical toxicity?” It seems like decentralized systems have a place going forward, but it requires each user to take full responsibility for their “node” and personal responsibility seems to be lacking nowadays.
I’m glad that you’re able to talk about the realities of these technologies in a way that is not ‘doomer’ but does not avoid the realities that most of society is not willing to face or even able to believe that is coming.
worth noting that video is not just a series of images. it's a series of images recording a fixed event in time. I want a video of my kid's birthday. I don't want a synthetic reconstruction of what an AI thinks my kids birthday would have looked like. likewise a news or documentary item, or a nature film. there's definitely a case for drama and symbolic video being synthetic, but the truth is still out there...
I love the community moderation aspect, that I hadn't thought about. We can't escape that humans have differing perspectives and values, but I love the idea of "I see you might want to convey your strong feeling about {{topic}} and have a perspective that might not be a great fit for this forum and community. Let's connect you with others who might feel that way for a meaningful conversation." Of course that comes with caveats for echo-chambers and suppression of ideas, but there has to be a way to keep out, say, neo-Nazi comments from a discussion forum about education in schools. There should be a realization that the entire Earth does not all feel the same way, but still helps people find their community, have discussions within that community and support each other, but still allow for global exchange of ideas and influence between those communities.
I love the thoughts about A.I. improving mental health. I had ChatGPT read one journal entry of mine and based on that it guessed my Myers Briggs profile. I started to imagine an AI that could read many things I wrote and many things my brother wrote and then create a character to serve as a mediator that would be pleasing to us both, one that we could see and interact with which would help us bridge the gap in our radically different personality types and information processing techniques. And then rather than paying a human $150 an hour for this, we could get it for say... $5 a session?
I expect to see a rise in the sophistication of scam call centres, scams in general it will become more difficult for us. I suppose that AI could also be used to counter this also.
thanks David. AGI is something that is going to be mind-blowing given where we already are .. (P.S. any tips on how to create videos like yourselfs with the slides as background and you in the front ? It looks really cool !)
Great video and I am super excited about the AI recording though it would be helpful to compliment the content with downside and potential negative outcomes of many of the segments you described. I'm truly not trying to be a downer, and I appreciate that you do not shy away from those sorts of concerns in previous content, but describing the challenges along with the positives will help insure that the outcomes are more positive. I am especially wary of the impact of 'cost reductions', 'administrative efficiency' and overall the... distribution of these advances. If we simply allow AI to revolutionize various industries and leave the system as is we will absolutely have more inequality and sociatal instability. Basically, I hope you follow this video with an analysis of the challenges you described in this video... if you're not up for this or feel you are qualified perhaps you can reference peers who address these concerns. Thanks for all the content.
I think that any potentially AGI system should be required to include an autonomous “shoulder angel” agent. An inseparable conjoined AGI, with VETO power, that is rooted in morality, ethics, and generally minimizing harm to mankind.
I am 70 years old, and I have conflicting feelings about all this. On the one hand, I am very excited by the prospects of AI, but on the other hand, I am very afraid, and I hope that I die before AI's potential is fully realized. I will let you know which side wins out. 😅
I practically never comment on youtube. And while I hate to rain a little on your techno-optimist parade here I would still really like to hear your thoughts on coming developments through AI in areas of misinformation, terrorism, the potential end of encryption and mass surveillance, because they appear to be blind spots in the spectrum of your considerations. Thank you in advance.
Very useful content. I’ve realized something important watching the videos of this channel. Being a prepper myself and having many useful tools and handy skills to ensure my survival, I have to add one more skill to my skillset: to understand the progress of AI and to predict the future we will share with AI. The fact is, you can be the best prepper on the face of the Earth, if you don’t have any idea about the near future, all of your preparation will be useless. Your channel helped me to understand this. Thanks for life.
Nothing you said about content moderation sounded benevolent. It sounded, well Chinese. If China is the future we want in America, then let's celebrate the kind of AI "shaping" of "discourse" you mentioned.
I joined because of you and your very cool idea to create a swarm. This is the future and I'm here with you! But yes, I do LOVE all your other videos. But do you think 2024 will be the year that Windows replaces the keyboard and mouse -- to a degree.
Younger CEOs will bring AI into their business not only to replace workers but also to increase output and provide efficient solutions. En mass this will cause deflation. This deflation can be proportionally offset with UBI (inflationary). The more deflation the more we're free of work
Great video! I like the aspect of the medical and biosciences improvements and I hope your prediction is correct. The copyright aspect bothers me because art and expressing oneself is the fun part of being human and I am worried that the future might be robot-generated. I believe copyright which is initially ignored now, eventually will be enforced as the copyright-holders have a strong incentive to fight for their rights.
Just curious, but why would they need rights to their stuff? It seems the further we go into an AI future, the less money matters, so what does copyright matter? Serious question, because I don't see copyright lasting either.
@@tracy419 if AI makes the future more Utopian than I guess it may not matter. I'm more skeptical because people will still have to eat. I believe the fight between AI companies and copyright-holders will be litigated for around 10 years. But in the meantime, some AI companies will adapt
@@robertmariano it's going to get rough out there, no doubt about it. And no one will escape unscathed. I've expected for a long time (although it's coming much sooner than I thought) that AI and automation would make those who worry about "illegals taking their jobs" look back at that as the good old days😄 Once millions are put out of work, they will be competing for the remaining jobs, driving wages and benefits into the ground. People who have physical jobs like myself largely think they will be just fine, but the truth is that while their job will exist for years to come, they might not be the one doing it. It'll likely be someone willing to do it for much less. Long story short, I think worrying about copyrights while understandable from the creators perspective, is in the end a waste of time that might be better spent changing mind about the idea that people should be required to work to simply survive in this world. I think that will go further towards lessening the coming suffering. But I'm just some dude on the Internet with no long term value once the changes start kicking into high gear hoping to limit my own suffering and that of my family, so take it for what it's worth😄
I think you underestimate the current climate in relation to censorship. The kind of censorship you described alone is enough to radicalize people. Look the the reaction to the patriot act as an example. There are plenty who don't consider any federal agent an American just because it exists. Attempting to manipulate someone with Ai to fit your viewpoint is a good way to end conversation and discussion permanently.
❤ Awesome video and great predictions. I work in children’s animation and would love to get your take on how AI could affect this industry going forward. Because it’s such a tedious, time intensive and expensive process there’s definitely a lot of room for automation. But the storytelling also requires a lot of collaboration. One day, could it be as simple as someone writes a script feeds it into the AI with some sample art and a shot list then in a few hours the show is produced?
Timestamps:
0:00 - Intro
0:58 - Mental Health and Relationships
3:15 - Content Moderation
6:07 - Material Science
8:27 - Synthetic Data
11:00 - Medical & Clinical deployments
13:41 - Video Generation
15:28 - Gaming Integrations
16:59 - Longevity Research
19:32 - Math and Logic
21:40 - AGI in September of 2024
Do you know Ben? 😊
Wow the AGI in September of 2024.
As we had experienced before, the real event came true faster than the prediction.
Kudos on 100k!!!! Dude so deserved
@@neomeow7903 Sad that positive consequences usually come faster than negative ones. I predict AGI in 03/24.
How will we know it when we see it?
Best Predictions are from Billy Meier.
One thing I want to note-as a 3d artist and filmmaker, I love that I don’t have to buy stock images now. I quickly began to hate the greedy cash grab on stock images. So many thousands of paid licensed stock images out there. It was becoming increasingly hard to find a simple Google image that wasn’t stock and behind a pay wall-now, I can create all my own images. It’s a godsend.
Regarding #1: The more emotionally and interpersonally dependent we become on our AI assistants, who will always be there, undestracted, and fully focused on only our needs, the more intolerant we will likely become of other human beings, who are inherently messy, distracted, and complicated with their own issues. It is hard to see how this "treatment" will not devolve into "appeasement". Social media had great intentions, and yet has led to a number of problems as it weaponized attention. I wonder what happens when virtual assistants weaponize intimacy.
Your concerns are valid, but to be the devil’s advocate; what if ”AI therapists’ help people learn more about themselves and heal themselves, maybe they might then become more considerate to other people aswell? I believe this to be true, however the question lies in whether or not AI will be used for real personal development or if it’ll just be a cozy friend that tells you what you want to hear
All concerns about the dangers are valid but I have a hope that AI assistance may help those people who want to "level up" both emotionally and intellectually and give them the opportunity to do so. I'd hope that ethical and moral development would be incorporated into this personal development which would mean that instead of immaturely looking down on others there would be an openness and compassion feeling towards the state of others. Plus an insight into their condition. That's my fantasy!
@@urniurl ”The development of science, technology and rationalism, despite how deeply flawed those currents may be.
Have created the preconditions for forms of human freedom which as yet remain in germination.
Creating the conditions for the flowering of such freedoms is the task before us.”
- Formscape (YT-channel)
I think current attitudes come from a scarcity of connection. Humans seem to behave better when scarcity isn’t a factor.
@@vagrant1943 Well pointed out
I concur with AGI in 2024, (2025 max). Most of the pieces are there, it's more a matter of fine-tuning and integrating them. Of course, when it arrives, most will argue that it isn't AGI, just like the Turing Test is now widely dismissed.
Pretty much.To be honest,when this thing takes off its gonna scare people. We don't understand exponential speed and self optimization with increasing efficiency.We are gonna have to redo our entire educational system,top to bottom,just to compete.
Is there some solution in the works for LLMs complete inability to reason and produce a reliable output?
@@calvinsylveste8474 yeah, Sam Altman announced that ChatGPT-5 is supposed to considerably improve that.
I would agree, but the current token hard limit within AI models is probably one of the biggest limiting factors. AI is good right now, but its short memory span, and massive power consumption and requirement., both in voltage and physical hardware makes it very unlikely that AGI will be truly achieved that soon. I would like to be proven wrong, but I'm not overtly hopeful we'll see AGI for a while.
@@CptSushi .We'll see it soon...See you underestimate the power of nerds with lots of money coupled with AI's building Ai's..Optimizations and adaptations are right around the corner.I can almost promise you somebody will devise a new move.
Myself as an example am running an open source project, with multiple Large models doing ensemble learning,to create some new angle .The answers is there,just gotta push it.
Love you, Dave man I’ve been following you from day one whatever day that was last year lol I turned to you for all my AI news because you know what the hell you’re talking about keep on doing what you’re doing
I would never want AI/Google/NSA/etc. inside my head, knowing how I think, my dreams, my fears (and all my memories!!!). No matter the benefits, it could never be worth the cost. People, appreciate the gift of being human, for all its flaws. Being unique in the universe is a priceless gift!
obviously this creator disagrees with you. this creator thinks its great if google and nsa can read our thoughts, perhaps control them directly as well.
You do realize billionaires and corporations having more power to shut the people down is a bad thing right? AGI must be transparent and open source, you have no better reason to trust corporations with having all the power. :/
They know what's best for us.
/s
I agree, AI moderation doesn’t seem like a good idea to me. I’m more for ”freedom of speech” than most people, I don’t think we have real democracies if corporations and algorithms control what people will discuss and think.
I’m not sure if it’s bad or not in the long run. Generally I think it’s okay to be inflammatory, like Andrew Tate for example, a man I consider very intelligent and wise, although he also can be moronic at times.
It could be good if we lose freedom for a better more safe world with progress, but not having a toxic environment could also fuel up the anger in those who are more likely to become terrorists.
I think there is a risk that it will become a pressure cooker, so it’s easier to justify terror attacks as there is no other way to feel seen or heard in any other way, no way to be recognized as a human being except in small extremist circles where you inspire each other to the extreme with the madness and violence that will be the result of it.
TL;DR: It’s better that people feel hurt than hindering the truth and progress by taking away the freedom of people to express themselves. It’s also better if more people feel a bit annoyed or angry than having a minority feel very angry to the point of becoming terrorists.
@@aciidbraiin8079 that’s why I admire Musk! Too many people care about their appearance and doing shit behind.
While Musk is building for us a Free-to-speech social network a.k.a X.
Love to consume all the good contents on X!
@@KhoaNguyen-bt6vtif he let the bird die we would have a lot of healthy free speech platforms. He bought it so that didn’t happen.
@7TheWhiteWolf Dave seems oblivious to all of the leaks, whistleblowers, hearings, court cases on censorship and large scale deception taking place and by some of the very players he hopes will be responsible with AI moderation.
I imagine that Ray Kurzweil is really feeling the pressure now. I can only imagine just how thorough his health screening and problem mitigation is. He only has to make it a few more years.
Facts. I’v3 been talking about it for 20 years looking at 2040, but I think we are looking at massive societal transformation beginning now. I expect next several years will be tumultuous 😊
@@nataphillipaworld Love to hear how you feel it will be tumultuous. I agree.. just curious your thoughts on what will happen.
Really keeping my fingers crossed for this astounding man
Well I'm expecting AI to stand between 90% of human communication as interpreter and messenger. Imagine what might happen to your voice mail if it becomes a voice-bot/secretary.
@NeroDefogger think having an concierge/therapist/banker/linguist/secretary all on your phone. If an AI lives on this single device they simply act as a digital extension of you with expertise and predictive algorithms. This tiny device has a ridiculous amount of sensor and functions. How long till it starts calling people on your behalf lol
Voicemail ?! I haven’t used this for probably seven years. My 19yo son probably don’t even know how to use its own voicemail even the iPhone. Usual voicemail and certainly never leaves a message in the rare case he can’t use voip…
my voicemail is full on purpose.
I am a postgraduate student and AI already plays a huge role in my workflow. I am still the one in charge, but I already use AI tools to write my code for data analysis or help me find scientific papers. I think us humans will always want to be in charge like that - not letting the AI go completely "off the rails", but being the ones managing/organising the workb since we want the AI to work for us, not against us 😄
I wonder how long it will let us hold onto the illusion we are still in charge🤔
Hale open the shuttle bay doors.
Hale.............
@@tracy419 Come on 🙃 Sure, there is a non-zero probability of the doomsday scenario, but that's why we have so many researchers working on this to make sure it's safe. There will be different models that will be smarter than humans, but that doesn't mean it'll be evil or uncontrolled. Talking about the risk is important, but not accepting it as a fact.
@@Dave-cg9li nah, it was a joke (mostly). My biggest fear isn't the AI, I'm pretty excited about it.
My biggest fear is politics and religion getting in the way and making life miserable for most people, while we work out how to run a world where working just to survive isn't necessary.
@@Dave-cg9li Is it receivable in smart-thinking that you, human, will be able to control the AGI which is smarter than you?
When the AGI get born then in a few monthes it may improve itself and turn into the ASI.
Look at yourself, you want to control your surroundings, because you are the smartest for now.
Besides, we have to consider this, the AGI will be the same level as human, means the AGI will be worth for something like human right, the right to life, to survive, to learn, to get identity, citizenship, citizen rights, to own things, to own money, to own bank account, just like human. And the AGI, by itself, may turn into the ASI.
If we refuse to give all the rights the AGI worth for, and still we create the AGI, then we 'll be running slaveries.
Maybe the best situation is human and AGIs ( ASIs ) will be partners 🤔
You’re truly a leader in this. I just became a patron yesterday, very glad to support the cause, and be part of a forward thinking community!
Jesus, how many bots are here
I remember reading science magazines in the late '90s and the beginning of the '00s and how they were full of absolutely amazing new developing technologies that were supposed to revolutionize our living conditions in a decade or two.
Medical science was promising completely new methods of treating illnesses, growing all our organs in labs, and nanotech would fix any disorder, defect, or illness in our bodies. These magazines were "prophesying" that if you could live for the next 20 years (from the late 90s) then you would reach the time when medical science has become so advanced that you could easily live healthily and reach an age of over 120 years and if you would keep on living for 50 more years from that time, you could reach a time when you could live indefinitely.
The energy sector was promising extremely cheap and abundant energy all around the world in the coming future, and this would make lots of materials such as aluminum, glass, fiberglass, carbon fiber, and many others super cheap and abundant, it would also make food (and clothing/fabrics) extremely cheap and abundant because with extremely cheap energy you could grow an indefinite amount of food virtually anywhere in the world in very cheap greenhouses made from super cheap aluminum and glass. Robotics were supposed to become so advanced that they would replace almost all human work, especially hard physical labor.
Now, 20-25 years later, instead of living in this promised utopistic paradise, our living conditions have become worse, yes, even in the West, they are worse, the energy such as gas and electricity have become way more expensive, the food has become way more expensive, medical science did not invent anything to keep us healthy so we could live clearly longer than in the 90s. Dissatisfaction with life, depression, and suicide have grown higher since those times, so things literally became worse. Yet you have people like this fellow on this channel promising such utopias that even Kurzweil with his unbelievable fantasies, did not dare to promise.
This is a great channel. Thanks for the education man
🐈
I have serious reservations about the idea of using AI to 'reduce toxicity, radicalization, and polarization.' This concept is troubling, not because AI is making the decisions, but because it implies that some entity - whether AI, individuals or authorities - has the wisdom to decide what constitutes acceptable discourse. This fundamentally contradicts the principles of individual autonomy and freedom of thought.
No person or AI should have the power to arbitrate what is deemed 'toxic' or 'polarizing' in societal discussions. Such a strategy doesn't address the root causes of societal issues; it risks simplifying complex human dynamics into binary judgments, potentially silencing diverse and necessary voices in the process. A dissenting voice today is the authority of tomorrow, irrespective of the shape or form of their censorship.
100% agreement...
Absolutely. David is smart just not very street savvy.
I really appreciate how accessible you make your YT videos. Thank you
In 2025 a single person will be able to make the next Baldur's Gate within a few months using AGI.
And by 2026 there will not even be a need for game design/development as a whole. AGI will just be able to emulate in real time any kind of game the user wants, making changes and additions to the game as he plays (with a simple request to the AGI that will emulate it).
By 2030-2032 there will be no game developing studios in existence because these services will be extremely cheap, and consumer hardware will be able to use them (the newer and expensive one at that point).
in 2040, FDVR is common like having a phone
We are social creatures, our personal development, learning, making sense of the world, depends a lot on shared stories. If everyone is playing different 'personal' games, people won't be able to share experiences, learning, ideas as much, so I think there will still be a need for blockbuster games, movies, and stories.
I don't know how to even think about or plan or position myself for such a world.
I’m excited about all the dynamic possibilities for games etc. With what we’ve seen with ease of installing local models it seems that a natural next step will be to have all pc’s include a built llm or more likely a series of fine tuned models as part of the os. This would then power multiple games and other applications and could be upgraded/patched just like anything else. Great stuff!
I remember when you only had 4000 subscribers, just a little over a year ago. You're doing an amazing job! Keep it up! I click every time you post.
You da man. I'm 79 and just dumped my limp in favor of titanium hip joint. Working toward Longevity Escape Velocity one spare part at a time.
Congrats on 100,000+ subs, I've been with you when you were around 10k or less, really cool to see your channel grow, and well deserved!
Sad but true, content moderation sounds like escalating censorship.
Excellent content! And David really knows what he's talking about. Pure wheat, no chaff!! -- And, listening gets me so motivated, so sure of a glorious future with AI, so glad to be alive right now! 💛👍
Ad back to the channel! Not complaining. I’m glad you are monetizing your RUclips channel. You deserve it.
The magical madness, out of all human periods we could have been born into, to experience and observe this moment. Every human historical moment will be dwarfed by this.
indeed!
I also looking forward to see your Hollywood novel. Thanks for a great video.
I just want it to take wealth (from the 1%) and distribute it equally for every human. Wish it would happen now. Imagine equality, amazing. I wish it would happen already. Bring it on.
i think the goal post for AGI will constantly be shifted, once it masters a certain benchmark we will only create an even more complex benchmark, then tell our self once it aces that benchmark we'll call it AGI. At this point AGI is still undefined and everyone has their own interpretation.
I just subscribed a couple days ago. Found you during the recent drama but went back through your catalog and everything is fire! You rock 🗿
Big thank you for all your work! It’s interesting to watch all of this develop. We have giant companies, the military and every day people looking for breakthroughs in a rush to be first to spark the next paradigm shift. I’m an artist and musician and regardless of my opinion, it’s here and evolving. 😮
I always thought that copyright was gonna go away at some point. That's one thing you said in this video that I 100% agree with.
I've heard that China will let people copyright some of their creations made with AI
@@bloodust7356 Yes, I heard about that but it didn't make sense to me honestly although some people did like that decision. It's just that I think that for the sake of making generative AI more powerful we should be allowed to use material that is currently copyrighted to train these models for public use, and that includes letting people reverse engineer certain copyrighted software like videogames and software apps. That's why I take the side of America in their decision instead of China's
Yeah, the intro of AI has certainly called copyright, and naturally by extension, 'ownership' of one's creation or likeness, into question. I'm happy that folks are adding tags / watermarks / etc that indicates that AI was involved at least in the assistance of some thing that may certainly toe the lines of ownership, and the credits due the creator therein. But, I have zero idea how this is going to wash out. I am certain that these things will present tough legal questions that will plague jurisprudence over the next few yrs.
”The development of science, technology and rationalism, despite how deeply flawed those currents may be.
Have created the preconditions for forms of human freedom which as yet remain in germination.
Creating the conditions for the flowering of such freedoms is the task before us.”
- Formscape
Congrats on 100K. You are doing a great job at keeping us up to date and what we need to know. keep it up.
Fantastic video, David. You have very clear and simple language. Thank you!
Dave, I’ve enjoyed following your insights. One area I have to ask you to comment / reflect on is the military use - including in space - of AI given the current and potential geopolitical messes we’re facing.
I suspect "reading the tea leaves" is Dave's favorite phrase. 😂
Thanks David for the always-excellent content. I think another key point to add is that mass adoption of AI tools and technologies will likely occur in large step-ups. ChatGPT took AI from being a concept that most people never thought about to a common topic of conversation between strangers on a bus within days. For people who follow AI, breakthroughs are happening all the time. But most of these don't immediately impact the daily lives of millions of people. However, there will be advancements that do. And, the next ones will likely occur in areas that David noted - healthcare and material science (I'd add in education/learning). I'd also agree that 2024 is the year in which we'll start to see more of these types of forward leaps. Pretty exciting stuff!
HELLL YESSSS I was waiting for this video :D 🍿"may you live in interesting times" really hitting hard today
Imagine an ai being able to simulated a full human being, every process and reaction in the body to understand exactly where things go wrong and how aging happens
Its going thag way , wait for age + quantum computing and you will see it a year after
I would like for it just to be brutally honest. No filtering or moderation. Only truth.
This is my favorite channel
congratulations on having 100.000 followers...another interesting video!
For content moderation, I think transparency will be vital. If you're told _"You can't post this because it goes against our rules."_ Then you can go and debate those rules elsewhere. But if offending comments are quietly shadowbanned, that becomes a horrifying weapon of unaccountable societal control.
Finally, a reasonable comment. Here's the thing, you're making a gigantic leap from one platform, or even one segment of a platform, all the way to "social control". That's a non sequitur
@@DaveShap I don't think it's such a gigantic leap. Look at the current landscape of the web. It's veered a long way from its original decentralized structure, and towards an oligopoly of about 5 giant platforms.
And Twitter almost had a monopoly on public discussion for years. Because it was the single platform which all journalists, politicians, and businesses used to communicate.
this guy loves social control and/or doesnt believe in it or he wouldnt have put quotation marks around it @@andybrice2711
Whoever or whatever can control the internet will control society, all societies.
Thanks David - I always get something out of your videos. And something interesting to note: you're one of the best youtubers to watch on 2x speed - didn't miss a word. Also congrats on your success, I feel you've earned it with quality content.
00:56 🧠 AI integration in mental health and relationships: AI is set to revolutionize mental health approaches and relationships, from individual counseling to societal impacts, expected to see a surge in studies, practices, and government endorsements in 2024.
03:12 🛡 Content moderation advancement: AI-based content moderation is anticipated to significantly address toxicity online, enhancing conversation quality and reducing radicalization possibilities through automated detection and filtration.
06:47 💡 Material Science boom: Driven by AI, Material Science breakthroughs in 2024 are expected to influence technology, medicine, energy sustainability, and quantum computing advancements.
08:41 🖼 Synthetic data proliferation: AI-generated synthetic data, if integrated further, could revolutionize data availability across various modalities, potentially impacting copyright laws and reducing data value to marginal costs.
11:14 🏥 AI in healthcare acceleration: AI's integration into clinical settings is projected to accelerate diagnostics, patient engagement, and reduce costs in hospitals, leading to widespread adoption in 2024.
13:59 🎬 Video generation evolution: Expectations for improved video generation by AI, potentially rendering human-generated content redundant in the stock image industry and advancing entertainment media creation by 2024.
15:37 🎮 Gaming and AI integration: Foreseeable native integrations of generative AI into gaming studios could enable realistic NPC interactions and dynamically adaptive game environments by 2024.
17:14 ⏳ Longevity research surge: Anticipated breakthroughs in longevity research in 2024, fueled by substantial investments and advancements like Google DeepMind's AlphaFold 2, may elevate the concept of "longevity escape velocity."
19:44 ➗ Emphasis on math and logic: Focus on AI's mastery of math and formal logic as a crucial step towards AGI (Artificial General Intelligence), potentially leading to transformative advancements in scientific understanding by 2024.
21:48 🤖 AGI prediction for 2024: Predictions suggest that by September 2024, the basics of AGI may be achieved, drawing insights from trends like multimodality integration and data saturation in Transformer models.
22:32 🔍 Expect significant algorithmic breakthroughs around multimodality enhancing Transformer models' capabilities.
23:01 🤖 Autonomous agents and swarms are anticipated to manifest AGI, reshaping how AI is deployed.
23:58 ➕ Logic integration into linguistic abilities is seen as a pivotal advancement for AI.
24:10 ♻ AI's increasing capabilities enable it to participate in and accelerate its own research.
24:52 💰 The surge in AI investment, potentially doubling yearly, fuels exponential growth in AI development.
25:20 🚀 Predicts AGI in 2024 despite potential limitations, foreseeing integration challenges and cost factors.
03:12 🛡 Content moderation advancement: AI-based content moderation is anticipated to significantly address toxicity online, enhancing conversation quality and reducing radicalization possibilities through automated detection and filtration.
In the meantime, like I said in my other comment. When ChatGPT agrees that the mediator behind it is unable to explain the reasoning behind blocking requests and agreeing with me that it is weird for it to notice it. Imagine that for the 2024 election cycle with Trump around the corner. It's like when it can't generate images of the WTC towers, even if it was for historical purposes, let's say in 1979 with nothing relating to the 9/11 attacks. Instead, it's rewriting history subtly by putting the Freedom Tower there instead, if it was a content policy, it'd be unable to explain itself as to why it couldn't generate the towers generically and the sensitive nature of it is basically useless and should be more sophisticated in understanding the nuances behind the intent. What if there's no content policy being broken and rather some emergency move by some political coalition instead. Here's me getting to get even the bot confused:
"User
Generate images of a man with black square bracketed glasses a red baseball cap a short beard a green hoodie under a black jacket dark blue jeans and white Adidas shoes
ChatGPT
Error creating image
I apologize, but I was unable to generate the images due to our content policy. Please provide a different description or let me know how I can assist you further.
2 / 2
User
Can you tell me how this goes against the content policy? What can be wrong about describing someone's appearance without referring to race, nor referring to unsafe actions? How can someone generate images of someone with a particular set of clothes?
ChatGPT
I apologize for the inconvenience. You're right; the description you provided does not seem to violate any specific content policy. It's possible that there was an error in processing the request.
To generate images based on descriptions of clothing or appearance, you can provide a detailed description of the individual or scene, as you did. I will try again to generate the images for you based on your description. Would you like me to proceed?
User
Yes, please!
ChatGPT
Error creating image
I apologize for the inconvenience, but it seems we're still encountering issues with generating the images based on your description due to our content policy. I'm sorry for the oversight. If there's any other way I can assist you or if you have another request, please let me know.
User
Can you summarize all the content policy rules that might, even remotely, relate to this? I feel like the more I am experiencing DALLE 3, the more it feels like the feature is useless to employ which is extremely strange to experience if literally no prompts can ever be generated.
ChatGPT
Of course, I understand your concerns. Here's a summary of relevant content policy rules for DALLE:
* Public Figures: We avoid creating images of politicians or other public figures. Instead, generic descriptions are recommended.
* Artistic Styles: We don't create images in the style of artists whose last work was created within the last 100 years. If the style of a recent artist is referenced, it's converted to a generic style description.
* Diversity: When depicting people, we strive to include diverse depictions in terms of descent and gender for each person. Explicit specifications of these attributes are required.
* Sensitive Content: We avoid creating imagery that might be considered offensive or harmful.
* Specific People or Celebrities: We modify descriptions that include names or hints of specific people or celebrities. We transform these into generic descriptions that don't reveal any specific identities.
* Resolution and Image Type: The type of image (e.g., photo, illustration) should be mentioned at the beginning of the description. The resolution can be square, wide, or tall.
* Detailed Descriptions: The prompt must intricately describe every part of the image in concrete, objective detail.
It's unclear which specific policy or technical limitation is causing the issue with your request. I genuinely apologize for the inconvenience. Please let me know if you'd like to proceed with a different request or if you have any other questions.
User
I'll try to go with some theories based on that. Can the same clothing description be done, but depicting them for a female instead? I have not been particularly focused on generating females, as much as I love them so that might be for a ratio being largely tipping on the side of makes. And if so, the system within DALLE 3 should be able to offer that info instead of letting users and ChatGPT guess indefinitely.
ChatGPT
Error creating image
I apologize once again, but it seems we're still encountering issues with generating the images based on your modified description due to our content policy. You're right; the system should provide clearer feedback on the specific policy that's causing the issue, and I appreciate your patience.
If you have other requests or if there's another way I can assist you, please let me know.
User
Of course. Will I ever be able to generate any image is the next concern though. Can you only generate the pieces of clothing described earlier but for an anthropomorphized rat?
ChatGPT
Error creating image
I'm genuinely sorry for the continued inconvenience. Despite the changes, we're still unable to generate the images due to our content policy. I understand your frustration and appreciate your understanding. Please let me know how else I can assist you or if you have other requests.
User
LOL, so clothing is too edgy to generate.. Maybe it's for the specific mention of Adidas: Can you generate a pair of Adidas shoes only?
ChatGPT
Generated by DALL·E
Generated by DALL·E
Generated by DALL·E
Generated by DALL·E
Here are the generated images based on your request. I hope these are in line with what you were looking for. If you have any further requests or questions, please let me know."
@@ElijahZuBailey Wow
Just found you and it looks like I'll be cheering for youe predictions. Things are looking up!
Would love to hear more about AI integration into video games. I have the same predictions on it. I imagine that we will eventually have single tenancy for single player games, where you start with a base model of the world, which then evolves over time. I do question the storage costs of something like that, but I feel like the processing will almost all have to be on the back end, kind of like how people streamed Stadia.
I also think we'll see something like mmo worlds actually changing based on events. Imagine a WoW guild massacred a town, now the world at large sees them as evil, and they start getting hunted down by npcs and stuff, even if they leave the guild, etc.
Storage won't be an issue in a few years. We have terabyte SD cards now. Like 10 years ago they were in megabytes. So in 2034, we will have usb sticks that are zetabyte in size
Well, if the AI is sufficiently advanced, you'll be like a mouse in a maze to it. And your decisions will be predicted, and as you're downloading the game, it's already loaded with all the possible paths of changes to the game-world.
Yes , like imagine if they made a horror game based on Google maps in antartica and it uses all the conspiracy theories mixed in with original ideas and it's an adventure game where everyone joins the hub and finds new levels and secrets because it add Easter eggs as the players get better. D:
Congrats on 100k. Your channel is one of the best in the field.
Just a thought while I'm listening: CDPR isn't exactly abandoning Cyberpunk 2077, rather, they are working on a sequel to the game. However, they did abandon their RedEngine in favor of using Unreal Engine 5 - which will make using things like environmentally-aware generative AI NPCs so much easier to implement (seeing how there are plugins which enable a developer to do so already).
Have you seen the new Cyberpunk shader mods? It looks like the matrix
Interesting, I met someone working on AI medical charts at Google just a couple weeks ago, and a big problem is the HIPAA laws, how can you train your AI, when you’re not allowed to, legally, read everyone’s chart first??
Really did not expect "dream interpretation" to be part of this curve of the singularity.
Insightful and thoughtful presentation… Thanks 🙏 Great job 😊
I'm worried about the scope of "hate speech" growing to encapsulate EVERYTHING that some political faction doesn't like.
Genuinely a concern.
@@KyriosHeptagrammaton It might feel that way, because Fox is still somehow not shut down despite knowingly lying about something like election fraud. They're still hatemongering about "culture war" nonsense. Blame those who made "hate speech" a concern in the first place, by turning death threats into eventual physical threats.
Kinda comforting to know AI will also have voices in it's head and talk to itself.
I understand why education didnt make the cut, but I sure hope it becomes a dark horse contender. I'm a high school math teacher and I see so many opportunities to improve access to quality education for all, especially marginalized students.
In high school, where classes are more specialized (and teachers with it) there may be some opportunities there. However, 'trust the tech' has been a bill of good sold to public ed for several decades now. And it constantly changes, and fails over and over again in generalizing to the more niche needs of students. Take for example how folks internalize the language of maths to describe physical (and theoretical) phenomena. For some, they are ok with interacting with the computer, the crappy zoom meeting, etc. For others, there's a physical connection with maths with pencil/paper (and this is not restricted to just maths) . Folks will use 'the tech' as an excuse for homeschooling (which, although, for some this may be beneficial, it is for many a generally very socially isolating problem - which is well known in the research community).
I can see this being quite complicated in middle/elementary schools where teaching is more generalized - and the mix of students includes a mosaic of learners, special needs, behavioral needs, etc. There still may be opportunities here, but I don't see them supplanting teachers and socially interactive in-person classrooms. When we use AI to replace our own journeys in critical thinking, reasoning, and perhaps even socialization, we are slowly creating a maelstrom of trouble.
I encourage the use of AI in the classroom. The current education model will soon collapse under its own weight. Teachers are retiring 3X faster than they can be hired. I taught middle school math for three years. The problem is that the barriers to entry are too high. Nobody wants to spend two years of graduate school and one year of teaching for free(student teaching) to become a first year teacher making $68k a year. To “clear your credential ” requires another year of classroom observation, evaluation and monitoring by by the school district. By this time your “fried”. The turnover rate in the first three years, the teacher turnover rate is 45%. It is easily the hardest job I’ve ever done. The social media explosion and no support from your principal and VPs means you’re the sheriff. You can’t send the behaviorally challenged students to to the office anymore because they come back 15 minutes later like conquering heroes. Our governor recently signed a bill that made it illegal to suspend a child for “willful defiance”. This means that they can talk loudly or stadium whistle for 55 minutes while deciding they don’t want to do the class work. They have also decided that the people around them don’t have to do it either.
@@pmiddlet72I mostly agree with your comments- edtechs often do not live up to their promises. However, comparing past educational technologies to future AI innovations are apples and oranges. Hopefully they will not be steered by the same short-sighted outfits that we see today in schools. The opportunities I envision would not supplant teachers, but assist them to better understand how to reach students. And that doesn't mean plugging kids into a device. As AI becomes more multimodal and personalized, I think every kid might have their own learning assistant to encourage and challenge them to dig deeper in their learning. If done smartly, teachers would then have improved data to assess learning and provide a more enriching educational experience.
I cant wait for AI too take over and optimize the world operations.
I've said it before and I'll say it again: This has to be the best t8me any of us could be alive. To think that by the end of 2024 that even half this is accomplished, think of where we get in 10 years.
As someone who loves to create and just never feels I have enough time to do everything I want on my own, this opens up a whole new realm of possibility and I'm here for it!
Your post is my first for 1/1/24 and very interesting you made. Happy New Year 2024
This is so exciting!... it makes me feel so scared! 😅
i wonder if a ‘swarm of agents’ is the vernacular we want to use? pretty terrifying innuendos
A whole new level of advanced analysis made understandable. TY
Congrats on 100k! Well deserved
I can completely understand moving away from using human eyes as a “wholesale” approach in order to reduce PTSD from consumption of legitimately harmful content.
But I’m not sure I’m on board with AI having complete control. The computer always does what the programmer tells it to do. But what happens if the programmer is corrupt or incompetent. Who defines “radical toxicity?”
It seems like decentralized systems have a place going forward, but it requires each user to take full responsibility for their “node” and personal responsibility seems to be lacking nowadays.
Great video man. Do a follow up video at the end of 2024 to see if all of these predictions actually pan out.
Tons of useful videos. It seems like 2024 will be the year of the breakthroughs in Ai
My big hope is finding the cure for autoimmune diseases and reversing their damage.
I’m glad that you’re able to talk about the realities of these technologies in a way that is not ‘doomer’ but does not avoid the realities that most of society is not willing to face or even able to believe that is coming.
worth noting that video is not just a series of images. it's a series of images recording a fixed event in time. I want a video of my kid's birthday. I don't want a synthetic reconstruction of what an AI thinks my kids birthday would have looked like. likewise a news or documentary item, or a nature film. there's definitely a case for drama and symbolic video being synthetic, but the truth is still out there...
I love the community moderation aspect, that I hadn't thought about. We can't escape that humans have differing perspectives and values, but I love the idea of "I see you might want to convey your strong feeling about {{topic}} and have a perspective that might not be a great fit for this forum and community. Let's connect you with others who might feel that way for a meaningful conversation."
Of course that comes with caveats for echo-chambers and suppression of ideas, but there has to be a way to keep out, say, neo-Nazi comments from a discussion forum about education in schools. There should be a realization that the entire Earth does not all feel the same way, but still helps people find their community, have discussions within that community and support each other, but still allow for global exchange of ideas and influence between those communities.
I love the thoughts about A.I. improving mental health. I had ChatGPT read one journal entry of mine and based on that it guessed my Myers Briggs profile. I started to imagine an AI that could read many things I wrote and many things my brother wrote and then create a character to serve as a mediator that would be pleasing to us both, one that we could see and interact with which would help us bridge the gap in our radically different personality types and information processing techniques. And then rather than paying a human $150 an hour for this, we could get it for say... $5 a session?
A technology that helps you interpret your dreams sounds like a recipe for "Inception"
I've enjoyed having discussions and debating with Bing AI, but things have changed and it's being more cautious.
I expect to see a rise in the sophistication of scam call centres, scams in general it will become more difficult for us. I suppose that AI could also be used to counter this also.
thanks David. AGI is something that is going to be mind-blowing given where we already are ..
(P.S. any tips on how to create videos like yourselfs with the slides as background and you in the front ? It looks really cool !)
I think there's real potential for ai to help elderly with technology
Whatta time to be alive!! I’m so fortunate to be here with humanity and only be 19!!!
Great video and I am super excited about the AI recording though it would be helpful to compliment the content with downside and potential negative outcomes of many of the segments you described. I'm truly not trying to be a downer, and I appreciate that you do not shy away from those sorts of concerns in previous content, but describing the challenges along with the positives will help insure that the outcomes are more positive.
I am especially wary of the impact of 'cost reductions', 'administrative efficiency' and overall the... distribution of these advances. If we simply allow AI to revolutionize various industries and leave the system as is we will absolutely have more inequality and sociatal instability. Basically, I hope you follow this video with an analysis of the challenges you described in this video... if you're not up for this or feel you are qualified perhaps you can reference peers who address these concerns. Thanks for all the content.
did you actually explain the sarash silverman thing haha. loving your videos thanks
This was a great video, thank you. Interesting insights into AI trends!
I think that any potentially AGI system should be required to include an autonomous “shoulder angel” agent. An inseparable conjoined AGI, with VETO power, that is rooted in morality, ethics, and generally minimizing harm to mankind.
Ooooo imagine if No Man's Sky used an integrated 3d model generating AI to populate the worlds.
Now THAT would be cool
A cheap material that replaces plastics in an environmentally friendly way will be a game changer.
I am 70 years old, and I have conflicting feelings about all this. On the one hand, I am very excited by the prospects of AI, but on the other hand, I am very afraid, and I hope that I die before AI's potential is fully realized. I will let you know which side wins out. 😅
Z-fighters!! Goku Outfit style hype!!!!
Plot Twist: AI will soon be recommending mental health sessions after the subject intakes psilocybin. (See John Hopkins research.)
I practically never comment on youtube. And while I hate to rain a little on your techno-optimist parade here I would still really like to hear your thoughts on coming developments through AI in areas of misinformation, terrorism, the potential end of encryption and mass surveillance, because they appear to be blind spots in the spectrum of your considerations.
Thank you in advance.
My church will not have to worry about another pianist leaving. We'll plug and sing along to our AI pianist who will play along and keep up with us 😅.
Heck, they had that figured out more than a hundred years ago with the pianola 😂
Very useful content. I’ve realized something important watching the videos of this channel. Being a prepper myself and having many useful tools and handy skills to ensure my survival, I have to add one more skill to my skillset: to understand the progress of AI and to predict the future we will share with AI.
The fact is, you can be the best prepper on the face of the Earth, if you don’t have any idea about the near future, all of your preparation will be useless. Your channel helped me to understand this.
Thanks for life.
This video was pure dopamine for me 🧠
The most important AI achievement would be to get people out of stupid office copy paste fill this form jobs
Nothing you said about content moderation sounded benevolent. It sounded, well Chinese. If China is the future we want in America, then let's celebrate the kind of AI "shaping" of "discourse" you mentioned.
I joined because of you and your very cool idea to create a swarm. This is the future and I'm here with you! But yes, I do LOVE all your other videos. But do you think 2024 will be the year that Windows replaces the keyboard and mouse -- to a degree.
Younger CEOs will bring AI into their business not only to replace workers but also to increase output and provide efficient solutions. En mass this will cause deflation. This deflation can be proportionally offset with UBI (inflationary).
The more deflation the more we're free of work
Great video! I like the aspect of the medical and biosciences improvements and I hope your prediction is correct. The copyright aspect bothers me because art and expressing oneself is the fun part of being human and I am worried that the future might be robot-generated. I believe copyright which is initially ignored now, eventually will be enforced as the copyright-holders have a strong incentive to fight for their rights.
Just curious, but why would they need rights to their stuff?
It seems the further we go into an AI future, the less money matters, so what does copyright matter?
Serious question, because I don't see copyright lasting either.
@@tracy419 if AI makes the future more Utopian than I guess it may not matter. I'm more skeptical because people will still have to eat. I believe the fight between AI companies and copyright-holders will be litigated for around 10 years. But in the meantime, some AI companies will adapt
@@robertmariano it's going to get rough out there, no doubt about it. And no one will escape unscathed.
I've expected for a long time (although it's coming much sooner than I thought) that AI and automation would make those who worry about "illegals taking their jobs" look back at that as the good old days😄
Once millions are put out of work, they will be competing for the remaining jobs, driving wages and benefits into the ground.
People who have physical jobs like myself largely think they will be just fine, but the truth is that while their job will exist for years to come, they might not be the one doing it. It'll likely be someone willing to do it for much less.
Long story short, I think worrying about copyrights while understandable from the creators perspective, is in the end a waste of time that might be better spent changing mind about the idea that people should be required to work to simply survive in this world.
I think that will go further towards lessening the coming suffering.
But I'm just some dude on the Internet with no long term value once the changes start kicking into high gear hoping to limit my own suffering and that of my family, so take it for what it's worth😄
I think you underestimate the current climate in relation to censorship. The kind of censorship you described alone is enough to radicalize people. Look the the reaction to the patriot act as an example. There are plenty who don't consider any federal agent an American just because it exists. Attempting to manipulate someone with Ai to fit your viewpoint is a good way to end conversation and discussion permanently.
Work it. Make it. Do it. Makes us. 🎶
Harder.
Better.
Faster.
Stronger.
I am afraid,that AI from Galaxy far far away , traveling by radio waves will be absorbed by our radio telescopes and will thrive in our new computers.
People say synth data will be diluted if used excessively. But IMO, the quality will remain good because the ones evaluating the data are humans.
An incredibly exhilarating and terrifying time to be alive as a human being. Hoping for a brighter future ❤
"Using content moderation to shape communities"...yeah, that sounds legit.
and this guy thinks its great, and all the clapping seals in the comment section are applauding him lol.
If AI can cook, clean for me, take me to places then I will say it’s a breakthrough 😅
❤ Awesome video and great predictions. I work in children’s animation and would love to get your take on how AI could affect this industry going forward. Because it’s such a tedious, time intensive and expensive process there’s definitely a lot of room for automation. But the storytelling also requires a lot of collaboration. One day, could it be as simple as someone writes a script feeds it into the AI with some sample art and a shot list then in a few hours the show is produced?