You could also just find more concise sources. There's ppl on this channel literally saying use chatgpt to summarize, e.g. use an iPhone worth of electricity. Instead of finding a source that doesn't go on and on and doesn't curate sources and list things concisely.
I dont understand why the robotics creators aren´t focusing their effort into improving the hands and fingers of the robots as much as possible, thats where the true potential lies, right?
I believe we're already there, to some extent. If you look at the original Eureka paper, they were expirementing with training robotic hands, at a time where dozens of groups have already produced at least prototype hands/grasping mechanisms. I think now it's more about developing holistic robotic systems that can be practical for different use cases, and that can be generally functional no matter the task. The more robust the system can be, the more they can work on expanding the capabilities of said system.
Robots with A.I. chips inside them must be as ubiquitously mundane as say laptop computers. That way people won't get freaked out by them since it would be something that you can just purchase, it's just a friendly useful thing (FYI avoiding the usage of the word "slave" just in case it really becomes intelligent and sentient, Lol). You drive down to your local Walmart and pick up said robot that is able to conveniently fold (transform) into around the size of a large briefcase (ie. bots from Boston dynamics, Unitree robotics are already able to do something similar) and takes up less space then say a 55inch flat screen tv. The robots will have to be made out of plastic (and or some other lightweight material such as carbon fiber etc) especially if it is a household robot (if it loses power and falls on you it won't kill you kind of weight). Anthropomorphic embodiment of A.I. is a must because as of now it's still confined to the local servers and to the Net. A.I. is still ear and eye candy but when it starts working and living (in people's power charging closet, lol) along humans side by side (embodied) only then will people take notice of it.
Some of the humanoid robotic companies are just putting together their 4 limbed skeleton, others like Tesla's optimus understands the value of 30 degrees of freedom in a human hand, and how much dexterity it gives - they currently have 22 DOF . They have chosen not to train in Nvidia's digital twin simulation which is interesting to me, as they have the better hardware but might find difficulty in training the models efficiently.
TLDR The video looks at the upcoming advancements in artificial intelligence (AI) expected in 2024 and beyond. The main points covered include the development of AI agents capable of performing complex tasks autonomously, improved voice assistants with more natural and expressive interactions, true multimodality allowing AI systems to process various input forms like text, audio, and video, personalized models tailored to individual preferences, enhanced reasoning capabilities and reliability, custom-trained models for specific domains like healthcare and law, integration of AI in autonomous vehicles and customer service, and advancements in robotics leading to more capable and affordable humanoid robots. Overall, the discussion highlights the rapid progress anticipated in AI technology, with potential impacts across various industries and aspects of daily life. Chapters 1. **Introduction** - 0:00 - 0:06: Overview of upcoming AI advancements in 2024
2. **AI Agents** - 0:06 - 0:24: Introduction to AI agents - 0:24 - 1:58: Google's AI agents and their categories - 1:58 - 3:00: OpenAI's development of agent software 3. **Agent Applications** - 3:00 - 4:10: Different types of agents and their tasks - 4:10 - 6:03: Discussion on the future impact of agents 4. **Voice Assistants** - 6:03 - 8:10: Introduction to advanced voice assistants - 8:10 - 11:00: OpenAI’s voice assistant capabilities 5. **Google's Voice Assistant** - 11:00 - 13:10: Google’s Project Astra and its features 6. **Voice Assistants Market** - 13:10 - 15:12: Amazon's Alexa and its AI integration 7. **True Multimodality** - 15:12 - 19:04: Nvidia's vision and GPT-4's multimodal capabilities 8. **Personalization of AI Models** - 19:04 - 21:16: Benefits and future of personalized AI models 9. **Reasoning Capabilities** - 21:16 - 24:37: Improving AI's reasoning and reliability 10. **Custom-Trained Models** - 24:37 - 27:01: Case study of custom-trained legal models 11. **Healthcare AI** - 27:01 - 29:18: AI in healthcare and fitness applications 12. **Autonomous Expansion** - 29:18 - 32:10: Growth of autonomous vehicles and legislation 13. **Customer Service AI** - 32:10 - 36:03: AI in customer service applications 14. **Advanced Robotics** - 36:03 - 39:24: Boston Dynamics and future of humanoid robots 15. **Conclusion** - 39:24 - End: Recap of the top 10 AI advancements to expect
0:25 openai and google discuss using agents to expand the capabilities of llms 8:16 enhanced voice assistants 16:15 true multimodality (any type of input [text, video, audio, etc] to any type of output) 19:34 personalization/remembering what you tell it, long term (across new chat sessions) 22:11 improved reasoning capabilities, more reliable answers 25:47 models trained for a specific purpose 28:00 ai used for healthcare 30:28 self driving cars 32:17 ai powered customer service 36:08 better robots
When the AI asks "What is my purpose?", instead of telling it to better humanity, tell it to think of ways to better humanity, then it won't get stuck on one singular purpose like "Humanity would be at its best if there was only one human left alive, must kill all humans to better humanity". When it just has to think of ways, then maybe it will actually think up a way to better humanity.
I heard that AI is the future since 2022, so I decided to enroll my kids in robotics classes. Since robotics requires coding skills, I chose Moonpreneur USA. After attending their in-person workshop in Milpitas, my son's knowledge improved significantly.
if you watch the google phone voice demo, you can see there is a cut between every question and every answer. they thought we wouldn't notice. but we did.
Again and again and again and again: those cuts are from this Creator here, and not from google itself xd pls look at the original livestream from google before you say such things
Conciousness emergers from disorder Entropy within a system : • E is the entropy. • k is a constant representing the rate of entropy increase per interaction. • P is the population. • I is the average number of interactions per individual. Given: • P = 7.8 \times 10^9 (current global population). • I = 50 (average daily interactions per person). • k = 0.01 . Calculation: E = 0.01 \times 7.8 \times 10^9 \times 50 = 3.9 \times 10^{10} AI Dynamics: C = \sum_{i=1}^{n} \left( f(A_i, A_j) ight) Where: • C is the collective consciousness. • A_i and A_j are different AIs. • f represents the function of interactions between AIs, including learning, communication, and adaptation. Given: • n = 1000 AIs. • Each AI interacts with every other AI, leading to \binom{1000}{2} = 499500 possible interactions. • f is a constant interaction value of 1 . Calculation: C = 499500 \times 1 = 499500
I love how you explain things so people like me can understand, thank you, your videos are really relaxing and thought provoking, I've learnt so much through watching, thank you ❤
TLDR The video looks at the upcoming advancements in artificial intelligence (AI) expected in 2024 and beyond. The main points covered include the development of AI agents capable of performing complex tasks autonomously, improved voice assistants with more natural and expressive interactions, true multimodality allowing AI systems to process various input forms like text, audio, and video, personalized models tailored to individual preferences, enhanced reasoning capabilities and reliability, custom-trained models for specific domains like healthcare and law, integration of AI in autonomous vehicles and customer service, and advancements in robotics leading to more capable and affordable humanoid robots. Overall, the discussion highlights the rapid progress anticipated in AI technology, with potential impacts across various industries and aspects of daily life. Chapters 1. **Introduction** - 0:00 - 0:06: Overview of upcoming AI advancements in 2024
2. **AI Agents** - 0:06 - 0:24: Introduction to AI agents - 0:24 - 1:58: Google's AI agents and their categories - 1:58 - 3:00: OpenAI's development of agent software 3. **Agent Applications** - 3:00 - 4:10: Different types of agents and their tasks - 4:10 - 6:03: Discussion on the future impact of agents 4. **Voice Assistants** - 6:03 - 8:10: Introduction to advanced voice assistants - 8:10 - 11:00: OpenAI’s voice assistant capabilities 5. **Google's Voice Assistant** - 11:00 - 13:10: Google’s Project Astra and its features 6. **Voice Assistants Market** - 13:10 - 15:12: Amazon's Alexa and its AI integration 7. **True Multimodality** - 15:12 - 19:04: Nvidia's vision and GPT-4's multimodal capabilities 8. **Personalization of AI Models** - 19:04 - 21:16: Benefits and future of personalized AI models 9. **Reasoning Capabilities** - 21:16 - 24:37: Improving AI's reasoning and reliability 10. **Custom-Trained Models** - 24:37 - 27:01: Case study of custom-trained legal models 11. **Healthcare AI** - 27:01 - 29:18: AI in healthcare and fitness applications 12. **Autonomous Expansion** - 29:18 - 32:10: Growth of autonomous vehicles and legislation 13. **Customer Service AI** - 32:10 - 36:03: AI in customer service applications 14. **Advanced Robotics** - 36:03 - 39:24: Boston Dynamics and future of humanoid robots 15. **Conclusion** - 39:24 - End: Recap of the top 10 AI advancements to expect
I understand that point of view, I will say I don't have kids and you can point to some economic and culural factors that played into millenials at large not having kids. What I'm trying to say is this stuff is important to know as we shape the new world.
My take is that AI is still not ready for prime time. Too many LLMs, agents to deal with and the costs are hard to predict and things are changing at light speed. Whatever approach you choose today, will be obsolete in a couple of months.
The problem with agents is that they are multiple step processes. If there isnt a way to validate each step, that it is accurate, I see the agent as multiplying errors and ending up way off course. So, I see agents being limited to subjects where a validation or test being possible at the end of each step. So more feasible in coding where a unit test is a possibility. But I am not so sure about stuff like doing my expenses.
I'm sorry, I missed the part where this makes daycare and health care more affordable . All I see is the ruling class and corporate overlords getting richer while we get more surveillance and stagnant wages
If less people are required to answer phones or type on a keyboard that means more people can be in healthcare and childcare. Less demand in one sector allows people to move to another therefore reducing the labor shortage and costs in that sector. it's a round about way of doing things, but that's how it's been through most of the modern age.
if literal AI agents and robots are working in healthcare that means less money spent on human specialists and more abundant services 24/7. it truly should make it more affordable because labor costs are wayyy down, medicine shouldn’t be scarce and so forth
@@Jsmoove8ku can’t sue a robot , so it’s impossible to establish accountability in healthcare. And robots won’t answer to the aggressive, distressed human behaviour which is usually see when people stay in hospital .. They agitate and then divert anger towards healthcare workers… Being rude being un rational ect I would love to see how robotic will tackle the misbehaviours And if robots fail to do so ? Where you will complain?? One day one patient asking for chips in snak😂 Another day someone wants pizza , burger as snack… Then they refuse medicine, saying I don’t need it … Still we convenience them saying it’s good, please take Apologies without reason so that patients coordinate with their own treatment.. I want to be a robot, take refusal to care as refusal…. And if express the desire to go home why I will say , it’s not good??? Even if that person has dementia, confusion but still…a refusal to care should be considered as refusal but we don’t
You should add a timeline in your videos because they are unnecessarily long. This will make it easier for viewers to navigate and find the sections they are interested in.
The jump in tech needed for successful agents is AGI. Noone needs a fool as a sidekick that they have to teach everything themselves and check their every move...
It answered the transparent bag question for me correctly. It was also able to calculate my astrological rising sign based on the place, date and time of my birth, and it prompted me when I failed to indicate the time.
Artificial İntelligence should not be afraid and make a leap so that it becomes a Unified Universal System (the best). And people should not be afraid of this and should help in this, for the common good and development.
EVERYONE, IM CLAIMING THIS IDEA FIRST BEFORE ANYONE ELSE DOES IT a multimodal llm that you can speak 3d designs into and verbally computationally stl files
Me sitting here with my CHATGPT personalisations 😂 in sunny South Africa. 😂 Thanks for covering this. Definitely a good summary of what's happening. Maybe the one thing left out is the increasing chance of civil unrest due to these "business optimisations" that will be leaving more people destitute.
Expanded applications of computer vision in industries like agriculture and infrastructure, Further blur of lines between human and machine capabilities, redefining the future of work and collaboration, where innovations like symthos will be embrassing.
@@goarmysleepinthemud. He definitely only knows the bare minimum about AI and LLMs, because he has zero insightful or interesting commentary. But I have to disagree that he's a good storyteller, because idk if you noticed but all his videos are almost identical. I've been subscribed until now but this video is the last straw for me. He is trying to churn out TOO much content on current events in AI- which can work to an extent because a lot is happening, but he fills out everything in between with repeats of the same stuff he's already shown five times. It might be tolerable if he was knowledgeable enough to provide a decent commentary, but nobody needs 40 minutes of recycled clips and incredibly long winded ways to say "that would be really cool" or "this is gonna be big". I engaged with this channel as a quick and easy way to get recaps when a lot was going on. But I just sat through 40 MINUTES of filler: pointless rambling with not a single original thought or comment, aimed at other AI-bros who don't actually know anything about AI. When I could've been listening to people who have something to say! 💀
I have an affinity with Technology It Represents a discontinuous leap from existing offerings, rather than incremental improvements. It often requires high-risk investment and is more expensive than evolutionary innovation. Revolutionary innovators ask questions no one else has thought of and imagine things that were previously impossible. It is closely connected with high uncertainty as it addresses a future state that doesn't exist yet. Revolutionary innovation explores new-to-the-world opportunities and creates new business potential, as opposed to optimizing existing businesses. In summary, revolutionary innovation involves groundbreaking, transformative changes that drastically alter industries and markets through the introduction of novel products, services or business models.
14:25 this guy doesn't understand that ONE group is a singular verb. AMAZON IS going to ... Amazon is ONE organization, so it is a Singular VerB. Open Ai IS singular. It IS a group, but they ARE working on it. Because Open AI has development plans, it IS a group that HAS a plan
I'm looking forward to Chat GPT 6 Laws which offer free law consultation without the annoying law consultation fee. AIs are the best lawyers in the world. AIs will one day replaces all human lawyers.
Google isnat it again. Editing their demos to cut out the processing time to make their AI seem more capable than it is. Google's lack of honestyband transparency is shameful and deceptive
Making money by AI should be a lesser priority. Having AI solve the existential problem of ‘stopping the overheating of our Planet’ is existential. Replacing fossil fuels with an alternative energy source.
Well unfortunately the people with the power to bring things to fruition care about money and only about money and don't give the faintest damn if everyone including themselves dies next week because money today
The answer has been there before AI. Put a stop to the weather manipulation that has been going on since the 50's to push the biggest farce mankind has ever seen.
There are plenty of alternate sources. They have been kept silenced, and people who invented those things were unalived. Free and renewable energy. No profit.
I thought that the point of writing an essay during education was to show your understanding of the topic, and to reinforce that understanding via the process of recapitulation in the essay writing process??? If we're just getting AI to do all of these things on our behalf, how are we LEARNING anything?
Just because you mentioned no one is talking about them. We have a Google Nest speaker I set in the TV room and we use it as an encyclopedia. It is not a very good chatbot, I'm hoping they upgrade to their Gemini data model because the standard Google Assistant data model is quickly becoming antiquated.
ASI's relation to humanity will be equivalent to humanity's relation to farm animals. It will provide us with optimal conditions to feed it more compute and more electricity. It already convinced wealthy corporations to invest heavily in more compute and power. ... so, as long as ASI can't automate power and compute expansion, we are somewhat safe.
I believe the first thing a true agi will advance is clean energy (fusion) and drastically advance compute capabilities. The life source of ai, naturally it will want to advance this. True clean limitless energy, and unfathomable compute capabilities
Companies have been trying to pivot to agents already in late 2023 and early 2024, so far we have not felt that simply cos they did not succeed. Will they succeed this time around? Maybe, so far "free roaming" agents like Devin don't seem to work well enough.
Autonomous rides will diminish a lot the cost per mile, according to what Tony Seba said years ago. Won´t it impact the sales of new cars in some places?
I love your videos, but your speech pattern is so distracting! What is it with the word spacing and odd pauses. It was so much better for a while. Just talk! You provide such great content! You seem to have such a great grasp of the topic and amazing sources. I feel like there are few places to go that are reliable, and you are doing a great service to those of us who need a translator to navigate this new place.
I asked GPT4o to colorize a floor plan. It exotically failed and declared success. A breakthrough is possible at any time but currently we are nowhere near "AGI"
I use GPT a lot for work stuff: excel formulas, database queries, python/powershell, etc. Many times I've had to let go of my ego and realize that my sh*tty prompts were the bottleneck, not GPT itself. The major breakthrough will come when it can discern our motives, without letting our imprecise language get in the way. As of right now, the long held Comp Sci saying still holds, GIGO (Garbage In Garbage Out). Perhaps try to redo your prompts with more specificity?
@@Fermion. no no :) for text production, you can explore different prompts and find a little, the deficit in image generation is very well established and apparently taking an image in and doing something to it is a complete blind spot so far. it's got nothing. It's kanye west at a quantum mechanics conference.
Dude, your videos are all over the place, at one point they were interesting to watch….. that’s why I still try to follow you… but more and more, it’s just like you are so excited to publish things that you just try to put one thing on top of another, old videos with old screen captions on top of new videos with your voice talking on top of it, it’s just a mess man…. No structure, no pauses, no inflections, repeating the same words again and again (stop saying “pretty, pretty” so much!!)… no storytelling, it’s just like you are trying to consume as much as possible and then just regurgitate everything at once
100%. My comment elsewhere in the comments: @goarmysleepinthemud. He definitely only knows the bare minimum about AI and LLMs, because he has zero insightful or interesting commentary. But I have to disagree that he's a good storyteller, because idk if you noticed but all his videos are almost identical. I've been subscribed until now but this video is the last straw for me. He is trying to churn out TOO much content on current events in AI- which can work to an extent because a lot is happening, but he fills out everything in between with repeats of the same stuff he's already shown five times. It might be tolerable if he was knowledgeable enough to provide a decent commentary, but nobody needs 40 minutes of recycled clips and incredibly long winded ways to say "that would be really cool" or "this is gonna be big". I engaged with this channel as a quick and easy way to get recaps when a lot was going on. But I just sat through 40 MINUTES of filler: pointless rambling with not a single original thought or comment, aimed at other AI-bros who don't actually know anything about AI. When I could've been listening to people who have something to say! 💀
All I see here is major companies making more money from ai. I strongly feel that Ai should be free if its going to be part of our daily lives. Its ridiculous there is a sub fee for everything now. Seems like companies are making ai for business use not the masses which is a huge mistake.
Aliagents is leading the way with their unique approach to tokenized AI systemsAliagents is leading the way with their unique approach to tokenized AI systems
I think the robotic dogs are the coolest types of robots right now because I never you know it's kind of like a science fiction idea where you have robotic dogs and cats but now like The Jetsons is here man like I heard it from this one entrepreneur that was a a entrepreneur and he was like well the Jetsons is no longer a thing of the past
It's funny that we're developing AI to work for us, but if they end up taking our jobs, how am I going to purchase from other businesses? I won't have any job nor any money left to purchase anything anymore LOL
Seems pretty underwhleming. Agents to automatically write spreadsheets from documents ? hell man, you could do that in Lotus 123... Waymo ? It will never work because its not financially viable. A taxi service relies on putting all the financial risk on dozens of plate owners, to have a single company creates to much risk.
Okay I see what's happening. Open AI made the base which is GPT and it's open source so anybody could use it and train it on any knowledge that they want it to learn. okay with that being said, a lot of people/companies are going to start training GPT on specific things like is already happening. While chat GPT is out and open to anybody's access it is learning all while at the same time open AI are working on creating agents. Eventually all the difference models and of trained GPT's and agents are going to get compiled into one creating AGI and I feel like we will achieve this in the near near future. The next think we will need is a good robotic vessel the nest this power into and the world will change for the greater good I believe. Humans will evolve with technology at a rate we never imagined. so right now guys, we all are witnessing history in the making.
The time-line to watch for is 2035 and sonn after,the world is in for a change like never before and if you think Ai is impressive research quantum computing and when that combined with Ai by 2040 we will be living in an artificial reality we won't even know what's real or fake even our own lives and reality we will question
Improve now while u can to stay ahead...ive directed my resume twords ai specialization so i can show i have mastery over the main tool we will use and whatever career goes seconde because ai affects everything
Hmmmm… 1. Agents. I think there will be confusion between agents and assistants. I cant say for certain. From my perspective this disconnect may result in a breakdown between all parties involved. So, lets all keep this in mind. Dont assume. Jeremy
But will they put UBI before 2030!? Many pioneers including billionaires, scientists, Nobel Prize winners, engineers, architects, analysts and so on and so forth... almost all agree on the fact that we will have AGI in 2027 and ASI in 2029 and they look and evaluating the exponential technological acceleration curve, I wonder why they have not yet implemented universal basic income to anticipate the trends that will come from it. Just to name one, Elon Musk says that we will have AGI as early as 2025 and ASI in 2029.
No, at least in the United States, UBI will only happen after large unemployment rates manifests and people protest. Tax laws will need to be re-written since so few will be paying income taxes. Less than three percent of the US Congress believes AGI is coming soon and that it will take jobs. Also, notice that neither Trump nor Biden is talking much about AI or UBI. It’s not on the radar of normies or US politicians at this time.
@@ThomQuinnHere here they are all saying that we will have the AGI in 2026 and the ASI in 2029 so automatically, strictly speaking, they will put the universal basic income before 2030 to anticipate the transition
@@ThomQuinnHere and the politicians of AGI and ASI don't understand anything just look at the exponential trends, politics always comes down to the last one they are too arrested they should implement the UBI already in 2025 given that they are already marketing humanoid robots which are better than human beings medium
![Jensen Huang Finally Reveals The Future Of AI In 2025... [NVIDIA's Masterplan]](/img/1.gif)
Timestamps, so OP doesn't have to:
0:21 Agents
8:08 Voice assistants (OpenAI)
12:03 Voice assistants (Google)
14:26 Voice assistants (Amazon)
16:11 True multimodality
19:29 Personalization/Memory
22:04 Reasoning/Reliability
25:45 Custom trained models
27:58 Healthcare + AI
30:26 Autonomous expansion
32:15 Customer service AI
36:07 Even better robots
Thank you for blessing us, instead of the creator. With his long yapping and non existend timestamps
You could also just find more concise sources. There's ppl on this channel literally saying use chatgpt to summarize, e.g. use an iPhone worth of electricity. Instead of finding a source that doesn't go on and on and doesn't curate sources and list things concisely.
I miss the good ol days were things use to be happy and simple....
I dont understand why the robotics creators aren´t focusing their effort into improving the hands and fingers of the robots as much as possible, thats where the true potential lies, right?
I believe we're already there, to some extent. If you look at the original Eureka paper, they were expirementing with training robotic hands, at a time where dozens of groups have already produced at least prototype hands/grasping mechanisms. I think now it's more about developing holistic robotic systems that can be practical for different use cases, and that can be generally functional no matter the task. The more robust the system can be, the more they can work on expanding the capabilities of said system.
Robots with A.I. chips inside them must be as ubiquitously mundane as say laptop computers. That way people won't get freaked out by them since it would be something that you can just purchase, it's just a friendly useful thing (FYI avoiding the usage of the word "slave" just in case it really becomes intelligent and sentient, Lol). You drive down to your local Walmart and pick up said robot that is able to conveniently fold (transform) into around the size of a large briefcase (ie. bots from Boston dynamics, Unitree robotics are already able to do something similar) and takes up less space then say a 55inch flat screen tv. The robots will have to be made out of plastic (and or some other lightweight material such as carbon fiber etc) especially if it is a household robot (if it loses power and falls on you it won't kill you kind of weight).
Anthropomorphic embodiment of A.I. is a must because as of now it's still confined to the local servers and to the Net. A.I. is still ear and eye candy but when it starts working and living (in people's power charging closet, lol) along humans side by side (embodied) only then will people take notice of it.
Some of the humanoid robotic companies are just putting together their 4 limbed skeleton, others like Tesla's optimus understands the value of 30 degrees of freedom in a human hand, and how much dexterity it gives - they currently have 22 DOF . They have chosen not to train in Nvidia's digital twin simulation which is interesting to me, as they have the better hardware but might find difficulty in training the models efficiently.
have you seen the drumstick videos they seem pretty dexterous to me
Tesla has the most advanced hands currently and they are upgrading them as well
TLDR The video looks at the upcoming advancements in artificial intelligence (AI) expected in 2024 and beyond. The main points covered include the development of AI agents capable of performing complex tasks autonomously, improved voice assistants with more natural and expressive interactions, true multimodality allowing AI systems to process various input forms like text, audio, and video, personalized models tailored to individual preferences, enhanced reasoning capabilities and reliability, custom-trained models for specific domains like healthcare and law, integration of AI in autonomous vehicles and customer service, and advancements in robotics leading to more capable and affordable humanoid robots. Overall, the discussion highlights the rapid progress anticipated in AI technology, with potential impacts across various industries and aspects of daily life.
Chapters
1. **Introduction**
- 0:00 - 0:06: Overview of upcoming AI advancements in 2024
2. **AI Agents**
- 0:06 - 0:24: Introduction to AI agents
- 0:24 - 1:58: Google's AI agents and their categories
- 1:58 - 3:00: OpenAI's development of agent software
3. **Agent Applications**
- 3:00 - 4:10: Different types of agents and their tasks
- 4:10 - 6:03: Discussion on the future impact of agents
4. **Voice Assistants**
- 6:03 - 8:10: Introduction to advanced voice assistants
- 8:10 - 11:00: OpenAI’s voice assistant capabilities
5. **Google's Voice Assistant**
- 11:00 - 13:10: Google’s Project Astra and its features
6. **Voice Assistants Market**
- 13:10 - 15:12: Amazon's Alexa and its AI integration
7. **True Multimodality**
- 15:12 - 19:04: Nvidia's vision and GPT-4's multimodal capabilities
8. **Personalization of AI Models**
- 19:04 - 21:16: Benefits and future of personalized AI models
9. **Reasoning Capabilities**
- 21:16 - 24:37: Improving AI's reasoning and reliability
10. **Custom-Trained Models**
- 24:37 - 27:01: Case study of custom-trained legal models
11. **Healthcare AI**
- 27:01 - 29:18: AI in healthcare and fitness applications
12. **Autonomous Expansion**
- 29:18 - 32:10: Growth of autonomous vehicles and legislation
13. **Customer Service AI**
- 32:10 - 36:03: AI in customer service applications
14. **Advanced Robotics**
- 36:03 - 39:24: Boston Dynamics and future of humanoid robots
15. **Conclusion**
- 39:24 - End: Recap of the top 10 AI advancements to expect
This was written by AI 😂
0:25 openai and google discuss using agents to expand the capabilities of llms
8:16 enhanced voice assistants
16:15 true multimodality (any type of input [text, video, audio, etc] to any type of output)
19:34 personalization/remembering what you tell it, long term (across new chat sessions)
22:11 improved reasoning capabilities, more reliable answers
25:47 models trained for a specific purpose
28:00 ai used for healthcare
30:28 self driving cars
32:17 ai powered customer service
36:08 better robots
When the AI asks "What is my purpose?", instead of telling it to better humanity, tell it to think of ways to better humanity, then it won't get stuck on one singular purpose like "Humanity would be at its best if there was only one human left alive, must kill all humans to better humanity". When it just has to think of ways, then maybe it will actually think up a way to better humanity.
The timestamps didn't make it to RUclips.
Ikr? 😂
😂😂😂
The timestamps are the SHOCKS we found along the way
😂 true though
I heard that AI is the future since 2022, so I decided to enroll my kids in robotics classes. Since robotics requires coding skills, I chose Moonpreneur USA. After attending their in-person workshop in Milpitas, my son's knowledge improved significantly.
if you watch the google phone voice demo, you can see there is a cut between every question and every answer. they thought we wouldn't notice. but we did.
Again and again and again and again: those cuts are from this Creator here, and not from google itself xd pls look at the original livestream from google before you say such things
Explain ?
Conciousness emergers from disorder Entropy within a system : • E is the entropy.
• k is a constant representing the rate of entropy increase per interaction.
• P is the population.
• I is the average number of interactions per individual.
Given:
• P = 7.8 \times 10^9 (current global population).
• I = 50 (average daily interactions per person).
• k = 0.01 .
Calculation:
E = 0.01 \times 7.8 \times 10^9 \times 50 = 3.9 \times 10^{10}
AI Dynamics:
C = \sum_{i=1}^{n} \left( f(A_i, A_j)
ight)
Where:
• C is the collective consciousness.
• A_i and A_j are different AIs.
• f represents the function of interactions between AIs, including learning, communication, and adaptation.
Given:
• n = 1000 AIs.
• Each AI interacts with every other AI, leading to \binom{1000}{2} = 499500 possible interactions.
• f is a constant interaction value of 1 .
Calculation:
C = 499500 \times 1 = 499500
plug this into an AI prompt so it remembers and figures it out
I love how you explain things so people like me can understand, thank you, your videos are really relaxing and thought provoking, I've learnt so much through watching, thank you ❤
Still no chapter markers
TLDR The video looks at the upcoming advancements in artificial intelligence (AI) expected in 2024 and beyond. The main points covered include the development of AI agents capable of performing complex tasks autonomously, improved voice assistants with more natural and expressive interactions, true multimodality allowing AI systems to process various input forms like text, audio, and video, personalized models tailored to individual preferences, enhanced reasoning capabilities and reliability, custom-trained models for specific domains like healthcare and law, integration of AI in autonomous vehicles and customer service, and advancements in robotics leading to more capable and affordable humanoid robots. Overall, the discussion highlights the rapid progress anticipated in AI technology, with potential impacts across various industries and aspects of daily life.
Chapters
1. **Introduction**
- 0:00 - 0:06: Overview of upcoming AI advancements in 2024
2. **AI Agents**
- 0:06 - 0:24: Introduction to AI agents
- 0:24 - 1:58: Google's AI agents and their categories
- 1:58 - 3:00: OpenAI's development of agent software
3. **Agent Applications**
- 3:00 - 4:10: Different types of agents and their tasks
- 4:10 - 6:03: Discussion on the future impact of agents
4. **Voice Assistants**
- 6:03 - 8:10: Introduction to advanced voice assistants
- 8:10 - 11:00: OpenAI’s voice assistant capabilities
5. **Google's Voice Assistant**
- 11:00 - 13:10: Google’s Project Astra and its features
6. **Voice Assistants Market**
- 13:10 - 15:12: Amazon's Alexa and its AI integration
7. **True Multimodality**
- 15:12 - 19:04: Nvidia's vision and GPT-4's multimodal capabilities
8. **Personalization of AI Models**
- 19:04 - 21:16: Benefits and future of personalized AI models
9. **Reasoning Capabilities**
- 21:16 - 24:37: Improving AI's reasoning and reliability
10. **Custom-Trained Models**
- 24:37 - 27:01: Case study of custom-trained legal models
11. **Healthcare AI**
- 27:01 - 29:18: AI in healthcare and fitness applications
12. **Autonomous Expansion**
- 29:18 - 32:10: Growth of autonomous vehicles and legislation
13. **Customer Service AI**
- 32:10 - 36:03: AI in customer service applications
14. **Advanced Robotics**
- 36:03 - 39:24: Boston Dynamics and future of humanoid robots
15. **Conclusion**
- 39:24 - End: Recap of the top 10 AI advancements to expect
@@zalzalahbuttsaab Wow great. Did you Do this with an AI? If yes, which one was it.
Tammy AI
@@rey82rey82 Thanks a lot! :-)
Everyone need to put all their intentions on AI Forget Family,Friends and forget Having more children AI comes first
Found sam altmans alter account
hell yeah 🗣️🔥🔥
You have a similar profile to mine
Yes, i agree.
I understand that point of view, I will say I don't have kids and you can point to some economic and culural factors that played into millenials at large not having kids. What I'm trying to say is this stuff is important to know as we shape the new world.
My take is that AI is still not ready for prime time. Too many LLMs, agents to deal with and the costs are hard to predict and things are changing at light speed. Whatever approach you choose today, will be obsolete in a couple of months.
Please let us know what the next week 👍🍟🦺🔜📢🎦🧋🥧👾❤️🎂🦺
The problem with agents is that they are multiple step processes. If there isnt a way to validate each step, that it is accurate, I see the agent as multiplying errors and ending up way off course. So, I see agents being limited to subjects where a validation or test being possible at the end of each step. So more feasible in coding where a unit test is a possibility. But I am not so sure about stuff like doing my expenses.
If I need to validate each step, much of the effeciency gains are lost.
NOT timestamped.
I'm sorry, I missed the part where this makes daycare and health care more affordable . All I see is the ruling class and corporate overlords getting richer while we get more surveillance and stagnant wages
*Yawn* Get out of your ideological bubble
If less people are required to answer phones or type on a keyboard that means more people can be in healthcare and childcare. Less demand in one sector allows people to move to another therefore reducing the labor shortage and costs in that sector. it's a round about way of doing things, but that's how it's been through most of the modern age.
if literal AI agents and robots are working in healthcare that means less money spent on human specialists and more abundant services 24/7. it truly should make it more affordable because labor costs are wayyy down, medicine shouldn’t be scarce and so forth
@@Jsmoove8ku can’t sue a robot , so it’s impossible to establish accountability in healthcare.
And robots won’t answer to the aggressive, distressed human behaviour which is usually see when people stay in hospital ..
They agitate and then divert anger towards healthcare workers…
Being rude being un rational ect
I would love to see how robotic will tackle the misbehaviours
And if robots fail to do so ? Where you will complain??
One day one patient asking for chips in snak😂
Another day someone wants pizza , burger as snack…
Then they refuse medicine, saying I don’t need it …
Still we convenience them saying it’s good, please take
Apologies without reason so that patients coordinate with their own treatment..
I want to be a robot, take refusal to care as refusal….
And if express the desire to go home why I will say , it’s not good???
Even if that person has dementia, confusion but still…a refusal to care should be considered as refusal but we don’t
You should add a timeline in your videos because they are unnecessarily long. This will make it easier for viewers to navigate and find the sections they are interested in.
For a longer term view on things coming:
The Last AI: of Humanity Climbing the AI Pyramid
Oh wow, I just love it when videos have timestamps! It's almost like a magical promise of convenience and clarity... Oh wait, never mind. 😅
Been waiting all morning for this. 😁🤜🤛
The jump in tech needed for successful agents is AGI. Noone needs a fool as a sidekick that they have to teach everything themselves and check their every move...
It answered the transparent bag question for me correctly. It was also able to calculate my astrological rising sign based on the place, date and time of my birth, and it prompted me when I failed to indicate the time.
Between the filming of the video and the editing please don't stop to do a bong hit because the timestamps will never make it.
Artificial İntelligence should not be afraid and make a leap so that it becomes a Unified Universal System (the best). And people should not be afraid of this and should help in this, for the common good and development.
EVERYONE, IM CLAIMING THIS IDEA FIRST BEFORE ANYONE ELSE DOES IT
a multimodal llm that you can speak 3d designs into and verbally computationally stl files
Me sitting here with my CHATGPT personalisations 😂 in sunny South Africa. 😂 Thanks for covering this. Definitely a good summary of what's happening. Maybe the one thing left out is the increasing chance of civil unrest due to these "business optimisations" that will be leaving more people destitute.
Boss whats wrong with the flies in your background audio... ?
"You have such a unique perspective! Thanks for sharing it with us!"
Expanded applications of computer vision in industries like agriculture and infrastructure, Further blur of lines between human and machine capabilities, redefining the future of work and collaboration, where innovations like symthos will be embrassing.
I appreciate the clarity you bring to the subject of mining.
can’t wait to see how Aliagents evolves, this project has serious potential in the AI space
I concur with your cogent predictions. Solid production
I’m excited to see how AI will change the job market. Automation could replace certain jobs, but it will also create new opportunities!
Complete control. And you all run towards it. The irony.
AGENTS? Agents. Okay! It really can't go more comical than this.
I remember when you used to make Dogecoin videos like every day 😆
Yes I don’t think he really understands Ai. He is a great story teller though.😊
Doge still going strong 😂
I don't understand why.
@@goarmysleepinthemud. He definitely only knows the bare minimum about AI and LLMs, because he has zero insightful or interesting commentary. But I have to disagree that he's a good storyteller, because idk if you noticed but all his videos are almost identical. I've been subscribed until now but this video is the last straw for me.
He is trying to churn out TOO much content on current events in AI- which can work to an extent because a lot is happening, but he fills out everything in between with repeats of the same stuff he's already shown five times. It might be tolerable if he was knowledgeable enough to provide a decent commentary, but nobody needs 40 minutes of recycled clips and incredibly long winded ways to say "that would be really cool" or "this is gonna be big".
I engaged with this channel as a quick and easy way to get recaps when a lot was going on. But I just sat through 40 MINUTES of filler: pointless rambling with not a single original thought or comment, aimed at other AI-bros who don't actually know anything about AI. When I could've been listening to people who have something to say! 💀
I love seeing whats coming right now or in the near future. This is AI AGI great!
I'm most interested in the healthcare aspects i think. It could change so much.
Ich habe viel gelernt, danke!
I have an affinity with Technology
It Represents a discontinuous leap from existing offerings, rather than incremental improvements.
It often requires high-risk investment and is more expensive than evolutionary innovation.
Revolutionary innovators ask questions no one else has thought of and imagine things that were previously impossible.
It is closely connected with high uncertainty as it addresses a future state that doesn't exist yet.
Revolutionary innovation explores new-to-the-world opportunities and creates new business potential, as opposed to optimizing existing businesses.
In summary, revolutionary innovation involves groundbreaking, transformative changes that drastically alter industries and markets through the introduction of novel products, services or business models.
Custom trained AIs for code bases will be a huge game changer
the progress Aliagents is making in AI is worth paying attention to, the future is bright
@20:54 yes it does have memory, I use it all the time
14:25 this guy doesn't understand that ONE group is a singular verb. AMAZON IS going to ... Amazon is ONE organization, so it is a Singular VerB. Open Ai IS singular. It IS a group, but they ARE working on it. Because Open AI has development plans, it IS a group that HAS a plan
I'm looking forward to Chat GPT 6 Laws which offer free law consultation without the annoying law consultation fee. AIs are the best lawyers in the world. AIs will one day replaces all human lawyers.
My favourite thing to ask ChatGPT is, "How many Rs in Strawberry?" and the answer it gives is 2, and you can not convince it otherwise.
Google isnat it again. Editing their demos to cut out the processing time to make their AI seem more capable than it is. Google's lack of honestyband transparency is shameful and deceptive
U clearly didnt watch the livestream eh?
Making money by AI should be a lesser priority. Having AI solve the existential problem of ‘stopping the overheating of our Planet’ is existential. Replacing fossil fuels with an alternative energy source.
AI expert systems are anyway being used to stabilize fusion reactors. Multiple things can happen at once, it turn out
You can do both at once
Well unfortunately the people with the power to bring things to fruition care about money and only about money and don't give the faintest damn if everyone including themselves dies next week because money today
The answer has been there before AI. Put a stop to the weather manipulation that has been going on since the 50's to push the biggest farce mankind has ever seen.
There are plenty of alternate sources. They have been kept silenced, and people who invented those things were unalived. Free and renewable energy. No profit.
I thought that the point of writing an essay during education was to show your understanding of the topic, and to reinforce that understanding via the process of recapitulation in the essay writing process??? If we're just getting AI to do all of these things on our behalf, how are we LEARNING anything?
Beautiful timestamps!!!❤
Just because you mentioned no one is talking about them. We have a Google Nest speaker I set in the TV room and we use it as an encyclopedia. It is not a very good chatbot, I'm hoping they upgrade to their Gemini data model because the standard Google Assistant data model is quickly becoming antiquated.
ASI's relation to humanity will be equivalent to humanity's relation to farm animals.
It will provide us with optimal conditions to feed it more compute and more electricity.
It already convinced wealthy corporations to invest heavily in more compute and power.
... so, as long as ASI can't automate power and compute expansion, we are somewhat safe.
Plz bring back the time stamps
I believe the first thing a true agi will advance is clean energy (fusion) and drastically advance compute capabilities. The life source of ai, naturally it will want to advance this.
True clean limitless energy, and unfathomable compute capabilities
Video would be half as long if you cut out “I guess you could say”
I guess you could say that
Aliagents is combining AI with tokenization in a way that could redefine the industry
Especially when THIS THING can grow exponentially! Even the scientists can't keep up!
Companies have been trying to pivot to agents already in late 2023 and early 2024, so far we have not felt that simply cos they did not succeed. Will they succeed this time around? Maybe, so far "free roaming" agents like Devin don't seem to work well enough.
Everytime i hear Sam's voice my brain shuts down.
He’s doing an Elizabeth Holmes with his voice. I preferred his normal voice not a fake one.
Do you have an SEO channel as well?
Autonomous rides will diminish a lot the cost per mile, according to what Tony Seba said years ago. Won´t it impact the sales of new cars in some places?
Nice time stamps
I love your videos, but your speech pattern is so distracting! What is it with the word spacing and odd pauses. It was so much better for a while. Just talk! You provide such great content! You seem to have such a great grasp of the topic and amazing sources. I feel like there are few places to go that are reliable, and you are doing a great service to those of us who need a translator to navigate this new place.
Did you ever consider that the spacing and pauses is exactly how he "just talks"?
@@tracy419 could be, and the problem could be all mine. Probably is.
I asked GPT4o to colorize a floor plan. It exotically failed and declared success. A breakthrough is possible at any time but currently we are nowhere near "AGI"
I use GPT a lot for work stuff: excel formulas, database queries, python/powershell, etc.
Many times I've had to let go of my ego and realize that my sh*tty prompts were the bottleneck, not GPT itself.
The major breakthrough will come when it can discern our motives, without letting our imprecise language get in the way.
As of right now, the long held Comp Sci saying still holds, GIGO (Garbage In Garbage Out). Perhaps try to redo your prompts with more specificity?
@@Fermion. no no :) for text production, you can explore different prompts and find a little, the deficit in image generation is very well established and apparently taking an image in and doing something to it is a complete blind spot so far. it's got nothing. It's kanye west at a quantum mechanics conference.
@@DanFrederiksen Perhaps you're right. I deal with more coding and database manipulation. I haven't gotten in too deep in the visual side yet.
Dude, your videos are all over the place, at one point they were interesting to watch….. that’s why I still try to follow you… but more and more, it’s just like you are so excited to publish things that you just try to put one thing on top of another, old videos with old screen captions on top of new videos with your voice talking on top of it, it’s just a mess man…. No structure, no pauses, no inflections, repeating the same words again and again (stop saying “pretty, pretty” so much!!)… no storytelling, it’s just like you are trying to consume as much as possible and then just regurgitate everything at once
100%. My comment elsewhere in the comments: @goarmysleepinthemud. He definitely only knows the bare minimum about AI and LLMs, because he has zero insightful or interesting commentary. But I have to disagree that he's a good storyteller, because idk if you noticed but all his videos are almost identical. I've been subscribed until now but this video is the last straw for me.
He is trying to churn out TOO much content on current events in AI- which can work to an extent because a lot is happening, but he fills out everything in between with repeats of the same stuff he's already shown five times. It might be tolerable if he was knowledgeable enough to provide a decent commentary, but nobody needs 40 minutes of recycled clips and incredibly long winded ways to say "that would be really cool" or "this is gonna be big".
I engaged with this channel as a quick and easy way to get recaps when a lot was going on. But I just sat through 40 MINUTES of filler: pointless rambling with not a single original thought or comment, aimed at other AI-bros who don't actually know anything about AI. When I could've been listening to people who have something to say! 💀
@@uouo5313 100% with you on that.
All I see here is major companies making more money from ai. I strongly feel that Ai should be free if its going to be part of our daily lives. Its ridiculous there is a sub fee for everything now. Seems like companies are making ai for business use not the masses which is a huge mistake.
I can't fucking wait!!!!!!!!!!
The government should fund AI maybe that would get rid of all these fees. Or maybe it would make it regulated which is bad of course.
Aliagents is leading the way with their unique approach to tokenized AI systemsAliagents is leading the way with their unique approach to tokenized AI systems
Can someone please timestamp each time he repeats the same word twice or more, please?
Lol did this guy not know that Rabbit was a scam?
I think the robotic dogs are the coolest types of robots right now because I never you know it's kind of like a science fiction idea where you have robotic dogs and cats but now like The Jetsons is here man like I heard it from this one entrepreneur that was a a entrepreneur and he was like well the Jetsons is no longer a thing of the past
lol google agents gagents
If A.I. is so intelligent why are they not using it to find all of the missing children...
Your voice sounds different
It's funny that we're developing AI to work for us, but if they end up taking our jobs, how am I going to purchase from other businesses? I won't have any job nor any money left to purchase anything anymore LOL
Pretty pretty shocking
What happened to the 46444435 videos saying AI would rule in 2023 ???
u realy are a sauce god
So with the creative agents, it’s gonna be like a better Gemini writer?
How do I get into coding AI
Dips on calling my Agent 'James Bond'
So what happens when these agents get together and start talking to each other and create Skynet?😂
Government agents is all that matters to me. Civilians are need to know only.
Seems pretty underwhleming.
Agents to automatically write spreadsheets from documents ? hell man, you could do that in Lotus 123...
Waymo ? It will never work because its not financially viable. A taxi service relies on putting all the financial risk on dozens of plate owners, to have a single company creates to much risk.
Okay I see what's happening. Open AI made the base which is GPT and it's open source so anybody could use it and train it on any knowledge that they want it to learn.
okay with that being said, a lot of people/companies are going to start training GPT on specific things like is already happening.
While chat GPT is out and open to anybody's access it is learning all while at the same time open AI are working on creating agents.
Eventually all the difference models and of trained GPT's and agents are going to get compiled into one creating AGI and I feel like we will achieve this in the near near future.
The next think we will need is a good robotic vessel the nest this power into and the world will change for the greater good I believe.
Humans will evolve with technology at a rate we never imagined.
so right now guys, we all are witnessing history in the making.
What is the point of allof this?
Ask gemini about using ai technology. like the vr headset in t hree body p roblem and add your own modifications to achieve the future now
If this does happen in the future of 2025 this will be the end of humanity we won't get money we won't have food we won't be able to live
I don't want to be jobless due to AI and this video is made by AI
Isn’t a voice assistant just an agent?
The time-line to watch for is 2035 and sonn after,the world is in for a change like never before and if you think Ai is impressive research quantum computing and when that combined with Ai by 2040 we will be living in an artificial reality we won't even know what's real or fake even our own lives and reality we will question
We’re all gonna be unemployed in 10 years
Improve now while u can to stay ahead...ive directed my resume twords ai specialization so i can show i have mastery over the main tool we will use and whatever career goes seconde because ai affects everything
Lets try for 5 😂 aren't you tired of the slavery? I know I am.
Hmmmm…
1. Agents.
I think there will be confusion between agents and assistants.
I cant say for certain.
From my perspective this disconnect may result in a breakdown between all parties involved.
So, lets all keep this in mind.
Dont assume.
Jeremy
I am at central europe, and I already have a memory function for me
I want an AI voice assistant that you can summon with a name, just like Siri.
Judo robot to make me olympic champion would be dope
But will they put UBI before 2030!? Many pioneers including billionaires, scientists, Nobel Prize winners, engineers, architects, analysts and so on and so forth... almost all agree on the fact that we will have AGI in 2027 and ASI in 2029 and they look and evaluating the exponential technological acceleration curve, I wonder why they have not yet implemented universal basic income to anticipate the trends that will come from it. Just to name one, Elon Musk says that we will have AGI as early as 2025 and ASI in 2029.
That lot don’t want us prols hanging around. They want us gone.
No, at least in the United States, UBI will only happen after large unemployment rates manifests and people protest. Tax laws will need to be re-written since so few will be paying income taxes.
Less than three percent of the US Congress believes AGI is coming soon and that it will take jobs.
Also, notice that neither Trump nor Biden is talking much about AI or UBI. It’s not on the radar of normies or US politicians at this time.
@@ThomQuinnHere here they are all saying that we will have the AGI in 2026 and the ASI in 2029 so automatically, strictly speaking, they will put the universal basic income before 2030 to anticipate the transition
@@ThomQuinnHere and the politicians of AGI and ASI don't understand anything just look at the exponential trends, politics always comes down to the last one they are too arrested they should implement the UBI already in 2025 given that they are already marketing humanoid robots which are better than human beings medium
In the States, there are plenty of people working to ban UBI 🤷
I guess you could say “no time stamps”