2025 AI : 10 Things Coming In 2025 (A.I In 2025 Major Predictions)

TheAIGRID
9 Jun 202439:30

Summary

TLDRThis video script outlines 10 major AI advancements expected in 2024, highlighting the rise of AI agents capable of autonomous task completion across various fields like customer service and healthcare. It also discusses the evolution of voice assistants, true multimodality in AI systems, personalized AI models, enhanced reasoning capabilities, custom-trained models like those for legal research, and AI's role in healthcare. The script further explores autonomous vehicles, improvements in customer service AI, and the exciting developments in humanoid robotics, signaling a future where AI becomes more integrated and sophisticated in daily life and industry applications.

Takeaways

  • ๐Ÿง‘โ€๐Ÿ’ผ Agents are emerging as pivotal AI systems capable of autonomous task completion, with Google and OpenAI leading the way in developing specialized agents for various tasks.
  • ๐Ÿ—ฃ๏ธ Voice assistants are advancing rapidly, with OpenAI and Google set to release more sophisticated and expressive models that can interact more naturally with users.
  • ๐ŸŽจ True multimodality is on the horizon, allowing AI to process and generate a wide range of inputs and outputs, including text, images, audio, and video, enhancing user interaction and experience.
  • ๐Ÿค– Personalization features for AI models are in development, aiming to provide more customized and context-aware responses by leveraging user-specific data and preferences.
  • ๐Ÿง  Improvements in reasoning capabilities and reliability are expected for future AI models, reducing errors and enhancing the models' understanding and responses to complex queries.
  • ๐Ÿ“š Custom-trained models are being developed for specific domains like law, integrating vast amounts of domain-specific data to handle complex tasks with higher accuracy and reliability.
  • ๐Ÿฅ Healthcare and AI are converging, with companies like Google and Apple working on AI-powered health coaching services that provide personalized health recommendations based on individual data.
  • ๐Ÿš— Autonomous vehicles are expanding their presence, with companies like Waymo offering public rides in select cities, signaling a significant step towards driverless transportation.
  • ๐Ÿ’โ€โ™€๏ธ Customer service AI is transforming the industry by offering real-time, customized support and facilitating transactions more efficiently than traditional methods.
  • ๐Ÿค– Advancements in robotics are leading to more capable humanoid robots, with new models like Atlas showcasing enhanced dexterity and autonomy, indicating a future where robots perform a wider array of tasks.
  • ๐Ÿ“‰ The cost of humanoid robots is decreasing, making them more accessible for research and development, potentially accelerating innovation and practical applications in various sectors.

Q & A

  • What are the 10 major things in AI that the video discusses for the year 2024?

    -The video discusses the emergence of AI agents, advancements in voice assistance, true multimodality, personalization features, reasoning capabilities, custom trained models for specific use cases like law, AI in healthcare, autonomous expansion, customer service AI, and improvements in robotics.

  • What is the significance of AI agents as mentioned in the video?

    -AI agents are pivotal as they represent a shift in AI capabilities, allowing systems to autonomously accomplish tasks set by users, impacting various sectors and potentially changing how we interact with AI systems.

  • How does the video describe the evolution of voice assistants in 2024?

    -The video suggests that voice assistants will become more natural and expressive, with the ability to understand and respond to complex commands, and will be integrated into more devices for hands-free interaction.

  • What does true multimodality in AI entail according to the script?

    -True multimodality refers to AI systems' ability to process and generate various types of input and output, such as text, images, audio, and video, creating a more comprehensive and interactive user experience.

  • How will personalization features enhance AI models as per the video?

    -Personalization will allow AI models to remember user preferences and context, providing more tailored and relevant responses, thus improving the usability and user experience.

  • What improvements in reasoning capabilities does the video anticipate for AI models?

    -The video anticipates that future AI models will have enhanced reasoning abilities, leading to more accurate and reliable responses, and a reduction in the occurrence of mistakes.

  • Can you explain the custom trained models mentioned in the video for the legal field?

    -Custom trained models for the legal field, as exemplified by the partnership between Harvey and OpenAI, involve integrating extensive legal data and domain-specific knowledge into AI models to handle complex legal queries and tasks with high reliability.

  • How is AI being integrated into healthcare as discussed in the video?

    -AI is being integrated into healthcare through specialized models like Google's Med Gemini, which can analyze complex medical data and provide personalized health coaching and recommendations through mobile apps.

  • What advancements in autonomous vehicles does the video highlight?

    -The video highlights the expansion of autonomous ride-hailing services in major cities and the potential for self-driving vehicles to become commonplace, improving transportation and reducing the need for human drivers.

  • How will customer service AI change the customer service experience?

    -Customer service AI will enable faster, more efficient, and personalized customer interactions, handling transactions, and providing real-time information, potentially reducing wait times and improving customer satisfaction.

  • What is the potential impact of advancements in robotics as mentioned in the video?

    -The advancements in robotics, including more capable humanoid robots and those that can perform tasks autonomously without teleoperation, suggest a future where robots can undertake a wider range of tasks, from household chores to complex industrial processes.

Outlines

00:00

๐Ÿง‘โ€๐Ÿ’ผ AI Agents Revolutionizing Task Automation

The video discusses the emergence of AI agents in 2024, emphasizing their ability to perform autonomous tasks assigned by users. Google's demonstration of various agent types like customer, employee, creative, data, code, and security agents is highlighted, showcasing the potential impact on how we interact with AI. While the technology is still in early stages, the development by major companies like Google and OpenAI suggests a significant shift in AI utility and capability.

05:00

๐Ÿค– Advanced Voice Assistants and Their Future

This section explores the advancements in voice assistants, demonstrating their potential through interactive demos. The new generation of voice assistants is shown to be far superior to predecessors, with capabilities like storytelling, emotional expression, and even singing. The market potential is discussed, with Amazon's Alexa and Google's voice assistant projects indicating a move towards more natural and hands-free AI interactions.

10:02

๐ŸŒ True Multimodality in AI Systems

The script delves into the concept of true multimodality, where AI systems can handle various types of input and output seamlessly. NVIDIA's vision and Google's AI studio are highlighted as examples of this multimodal approach, which includes capabilities like visual analysis, poster creation, 3D object synthesis, and lecture summarization. The potential for personalized AI models is also mentioned, suggesting a more intuitive and customized user experience.

15:04

๐Ÿง  Enhancing AI Reasoning and Reliability

The need for improving AI reasoning capabilities and reliability is underscored, with references to comments by industry experts like Sam Altman. The discussion points out the limitations of current models like GPT-4 and the importance of developing smarter AI that can understand context and provide accurate responses consistently. The potential for AI to make fewer mistakes and provide more reliable information is a key focus.

20:05

๐Ÿฅ AI in Healthcare: Specialized Models and Applications

Healthcare is positioned as a significant beneficiary of AI advancements, with specialized AI models like Google's Med Gemini surpassing GPT-4 in medical QA benchmarks. The development of AI models fine-tuned for medical applications is highlighted, along with initiatives by Google and Fitbit to provide personalized health coaching through AI. Apple's plans for an AI-powered health coach in iOS 17 are also mentioned, indicating a trend towards AI-enhanced personal health management.

25:06

๐Ÿš— Autonomous Vehicles Expanding Transportation Options

The script discusses the real-world application of autonomous vehicles, with services like Waymo offering rides in select US cities. Tweets from users experiencing self-driving taxis are shared, highlighting the comfort and reliability of the technology. The Automated Vehicles Act in the UK is cited as a sign of expanding legal support for autonomous transportation, suggesting a near-future where self-driving cars become commonplace.

30:06

๐Ÿค– Next-Generation Customer Service AI

The evolution of customer service AI is explored, with Google's AI capabilities being utilized by Symbol Fashions to provide real-time, customized support. The script describes an interaction where a customer service AI agent assists a customer with a purchase, applying discounts, suggesting additional items, and confirming the order efficiently. The potential for AI to transform customer service by handling repetitive tasks and providing instant information is highlighted.

35:09

๐Ÿค– Advancements in Humanoid Robotics

The final section of the script focuses on the progress in humanoid robotics, with Boston Dynamics' new Atlas robot hinting at significant advancements in the field. The script mentions the capabilities of existing robots and the potential for robots to perform a wider range of tasks autonomously. The decreasing cost of robots and China's investment in humanoid AI are also discussed, suggesting a future where robots are more accessible and integrated into various aspects of life and work.

Mindmap

Keywords

๐Ÿ’กArtificial Intelligence (AI)

Artificial Intelligence refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. In the context of the video, AI is the central theme, with discussions around its advancements and applications in various fields such as customer service, healthcare, and robotics. The script mentions AI systems like Google's agents and OpenAI's developments, indicating the broad impact AI will have in 2024.

๐Ÿ’กAgents

In the script, 'agents' refers to AI systems designed to perform specific tasks autonomously, based on the parameters set by a user. They are pivotal to the video's narrative as they represent a significant shift in how AI is utilized, moving from passive interaction to active task completion. Examples include Google's customer agents, employee agents, and creative agents, each tailored to accomplish different objectives.

๐Ÿ’กVoice Assistants

Voice Assistants are AI-powered tools that respond to voice commands and perform tasks through verbal interaction. The script highlights the evolution of voice assistants, showcasing their enhanced capabilities in understanding and responding to user queries in a more natural and expressive manner. The mention of OpenAI's demo and Google's Project Astra exemplifies the advancement in this area.

๐Ÿ’กMultimodality

Multimodality in AI refers to the ability of a system to process and understand multiple types of input and output, such as text, images, audio, and video. The script discusses the expansion of multimodal AI, emphasizing its potential to revolutionize user interactions by offering a seamless integration of various data formats, as illustrated by NVIDIA's conference and Google's AI Studio.

๐Ÿ’กPersonalization

Personalization in AI is the customization of responses or services based on individual user data, preferences, and history. The script explains how personalization will enhance the usability of AI models by providing more tailored and relevant responses, making them more efficient and user-friendly, as discussed in the context of AI systems remembering user preferences and history.

๐Ÿ’กReasoning Capabilities

Reasoning capabilities in AI models refer to their ability to make logical deductions and solve problems based on the information provided. The script points out the importance of improving these capabilities to ensure that AI models can understand complex queries and provide accurate responses, which is crucial for their reliability and broader application.

๐Ÿ’กCustom-trained Models

Custom-trained models are AI systems that have been fine-tuned for specific use cases or domains, such as legal case law research mentioned in the script. These models incorporate domain-specific knowledge to perform specialized tasks more effectively, demonstrating the adaptability of AI to meet the unique needs of different industries.

๐Ÿ’กHealthcare and AI

The integration of AI in healthcare is showcased in the script through examples of AI models that can analyze complex medical data and provide personalized health coaching. The mention of Google's Med Gemini and the collaboration between Google Research and Fitbit highlights the potential of AI to revolutionize healthcare by offering advanced diagnostics and personalized health recommendations.

๐Ÿ’กAutonomous Expansion

Autonomous expansion in the script refers to the growth and development of self-driving technologies, such as autonomous cars and taxis. The script discusses the real-world application of these technologies, with companies like Whos offering rides in major cities, indicating a significant step towards a future with widespread autonomous transportation.

๐Ÿ’กCustomer Service AI

Customer Service AI pertains to the use of AI in providing customer support, with the script highlighting its potential to offer real-time, customized interactions. Examples given include Google's customer service AI and Wendy's drive-through AI, which demonstrate the efficiency and improved customer experience that AI can bring to service industries.

๐Ÿ’กRobotics

Robotics in the context of the script refers to the development and application of robots, particularly humanoid robots, that can perform tasks autonomously. The script discusses advancements in robotics, such as Boston Dynamics' Atlas and China's development of humanoid robots, showcasing the potential for robots to take on more complex and varied tasks in the future.

Highlights

Introduction to the 10 major advancements in AI for 2024.

Emergence of AI agents capable of autonomous task completion.

Google's demonstration of diverse AI agents including customer, employee, creative, data, code, and security agents.

OpenAI's development of agent software to automate complex tasks by taking over user devices.

Potential for AI agents to transform how we interact with and utilize AI systems.

Voice assistance advancements with more natural and expressive interactions.

Google's Project Astra showcasing the future of AI assistance with voice.

Amazon's plans to overhaul Alexa with integrated large language models and generative AI.

True multimodality enabling AI systems to process and output various types of data.

GPT 4.0's multimodal capabilities showcased through different features and applications.

Personalization features of AI models to enhance usability and user experience.

Importance of reasoning capabilities and reliability for future AI models.

Custom trained models for specific use cases like legal case law research.

Healthcare and AI advancements with models like Google's Med Gemini for medical QA.

Autonomous vehicles expanding into major cities and the impact on transportation.

Customer service AI revolutionizing business interactions and support.

Developments in humanoid robotics with advanced capabilities and decreasing costs.

China's investment in humanoid robotics aiming for global leadership in AI and robotics.

Concluding remarks on the upcoming AI and robotics trends to watch in the next year.

Transcripts

play00:00

so in today's video I'm going to be

play00:01

telling you guys a look at the 10 major

play00:04

things that are going to be coming in

play00:06

2024

play00:08

sl22 for artificial intelligence I'm not

play00:11

sure how long this video is going to be

play00:12

but there will be time stamps so you can

play00:14

take a look at everything that is going

play00:16

to be released and the future is most

play00:18

certainly exciting so let's not waste

play00:20

any more time so one of the first things

play00:22

that is basically here is of course

play00:24

agents now the video you're currently

play00:26

seeing behind you is a video that I

play00:28

recently made on Google Go's agents now

play00:31

this is why the you know the text right

play00:33

here isn't I guess you could say synced

play00:34

up is because what Google did and what

play00:37

many other companies are trying to Pivot

play00:39

to in the late 2024 and in 2025 is of

play00:43

course agents now agents essentially

play00:46

just refers to I guess you could say AI

play00:48

systems that are able to I guess you

play00:50

could say kind of do whatever they want

play00:53

and go out and accomplish a kind of task

play00:55

that the person has set out for them and

play00:58

I think this is going to be one of of

play01:00

the most pivotal moments that does

play01:03

happen in AI because this is going to be

play01:05

one of the moments that actually impacts

play01:07

you now the reason this is going to

play01:10

actually impact us and the reason you're

play01:12

going to truly feel this is because as

play01:15

you can see in Google's demo if I pause

play01:17

it right here you can see that they're

play01:19

trying to basically hit every single

play01:22

agent aspect that there pretty much is

play01:24

going to be in Google since they've

play01:26

decided to I guess you could say kind of

play01:28

call the red flag you can see right here

play01:31

that they've got customer agents

play01:32

employee agents creative agents data

play01:35

agents code agents and security agents

play01:38

and this video there's basically quite a

play01:40

lot of different features it's a

play01:42

50-minute video where I've kind of cut

play01:45

the long drawn out Google I guess you

play01:47

could say announcement into a shorter

play01:49

video about where they talk about you

play01:51

know employee agents and this is going

play01:53

to be really gamechanging because it's a

play01:56

complete difference in how we currently

play01:58

use AI systems currently we use AI

play02:00

systems in this in the sense in the

play02:01

general sense that you know we talk to

play02:04

these AI systems and we interface with

play02:06

them in a chat interface and with this

play02:08

one we're going to Simply ask the AI

play02:10

system to do something outline some

play02:12

parameters and then the AI system is

play02:14

going to be able to go ahead and

play02:16

accomplish those tasks and that is why I

play02:18

think this is going to be probably one

play02:21

of the most pivotal moments now this is

play02:23

still pretty pretty early in its stages

play02:26

like currently we're still in a very I

play02:28

guess you could say basic area for the

play02:30

agents and the current demos that I'm

play02:32

seeing from some of the top labs they

play02:33

don't seem too too impressive so it

play02:36

seems like there is a breakthrough that

play02:37

is you know probably needed but I think

play02:40

whenever this breakthrough does happen

play02:42

because some of the larger larger

play02:43

companies are really really working on

play02:45

this I'm pretty sure this is going to be

play02:47

crazy now one of the things is that open

play02:50

ey have actually spoken about this you

play02:52

can see that open ey have shifted its

play02:54

Battleground to software that operates

play02:57

devices and that automates t tasks so

play03:00

essentially you can see right here it

play03:02

says open ey is developing a form of

play03:04

agent software to automate complex tasks

play03:07

by effectively taking over a customer's

play03:10

device the customer could then ask the

play03:12

chat GPT agent to transfer data from a

play03:16

document to a spreadsheet for analysis

play03:18

for instance or to just completely do

play03:20

random different things or to fill out a

play03:22

spreadsheet and I think like I said

play03:24

already this is going to be one of the

play03:26

main things that really really changes

play03:28

the game and as you know open AI have a

play03:31

very very long history of releasing very

play03:34

good products so whatever agent they're

play03:37

working on apparently they're working on

play03:39

two agents you know computer using

play03:40

agents and these agents can effectively

play03:42

take over a user's device to perform

play03:45

tasks such as transferring data from a

play03:47

document to a spreadsheet filling out

play03:49

expense reports and entering them into

play03:51

accounting software and these tasks

play03:53

involve actions like you know clicks

play03:55

cursor movements and text typing

play03:57

mimicking basically human interactions

play03:59

with different applications and of

play04:01

course secondly other than the computer

play04:03

agents okay and this is basically just

play04:05

summarized from this article we have

play04:07

web-based agents and these web-based

play04:08

agents are going to handle tasks that

play04:10

are primarily conducted online and they

play04:12

can gather public data about companies

play04:14

tra travel itineraries win of the budget

play04:16

book flight tickets they can utilize

play04:18

llms um and perform those really cool

play04:21

tasks so basically what we have here is

play04:24

I guess you could say a really really

play04:26

really interesting bit now Sam Alman did

play04:29

actually talk about this on about how

play04:31

agents are going to be pretty pretty

play04:33

incredible this is from an interview

play04:35

that he did earlier this year with Bill

play04:38

Gates it was one that I guess a lot of

play04:40

people did Miss because Bill Gates

play04:42

doesn't really have a large YouTube

play04:44

presence but it was one that I managed

play04:46

to get a small small audio snippet from

play04:49

where Sam Alman actually speaks about

play04:51

the future of agents and exactly what is

play04:54

going to happen creely new things and

play04:56

we're we're really seeing that and so as

play04:58

we look as we look at these back steps

play05:00

of things that can do a more complete

play05:02

task you can like imagine a little agent

play05:04

that you can say go write this whole

play05:05

program for me I'll ask you a few

play05:07

questions along the way but it won't

play05:09

just be writing a few functions at a

play05:11

time that'll enable a bunch of new stuff

play05:14

uh and then again it'll do even more

play05:15

complex stuff someday maybe there's an

play05:17

AI where you can say you know go start

play05:19

and run this company for me and then

play05:20

someday there's maybe an AI where you

play05:21

can say like go discover new physics and

play05:25

it's the stuff that we're seeing now

play05:28

very exciting and wonder

play05:30

uh but I think it's worth always putting

play05:31

in context of this technology that at

play05:34

least for the next 5 or 10 years we'll

play05:36

be on a very steep Improvement curve um

play05:38

these are the stupidest the models will

play05:39

ever be but coding is probably the area

play05:42

the single area from a productivity gain

play05:44

we're most excited about today uh

play05:47

massively deployed and you know it's

play05:48

scaled usage at this point Healthcare

play05:51

and education are two things that are

play05:52

coming up that curve so you could hear

play05:54

that he was basically saying that it's

play05:55

going to be a steep Improvement for

play05:58

agents because it's of course something

play05:59

that they haven't really worked on but I

play06:02

guess agents are going to be a thing

play06:03

where you can say hey look I need you to

play06:05

go create this company for me I need you

play06:07

to go run this business for me I need

play06:09

you to go make this PowerPoint

play06:10

presentation go do all of my homework or

play06:12

whatever of course at that time I

play06:14

guarantee there's going to be a

play06:15

different kind of homework but the point

play06:18

here is that you can see that agents are

play06:20

truly going to be that breakthrough

play06:21

moment and I do think that a lot of

play06:23

companies are working on this now I

play06:25

think the Breakthrough is likely to come

play06:27

and potentially you know maybe late 2024

play06:30

to early 2025 but I do think that agents

play06:33

is something that you know Google as

play06:35

you've seen are working on an open AI

play06:37

are working on because it's very very

play06:40

very important now in addition Google uh

play06:44

have actually stated that this is what

play06:46

they're working on and this is a clip of

play06:47

demes hassabis the leader of Google's

play06:50

Deep Mind which is their AI Division and

play06:53

he speaks about how they're going to

play06:54

develop these kinds of Agents I'm also

play06:57

really excited about the next stage of

play06:59

these large General models um you know I

play07:01

think the next things we're going to see

play07:02

perhaps this year perhaps next year is

play07:04

more agent-like Behavior so systems that

play07:07

are able not only to just you know

play07:09

answer questions for you but actually um

play07:11

plan and act in the world and and solve

play07:14

goals you know and I think those are the

play07:16

things um that will make these systems

play07:18

sort of the next level of usefulness in

play07:20

terms of being uh a useful everyday

play07:22

assistant and again that's bringing back

play07:24

in some of our uh work we did years ago

play07:26

with with gaming which were all agent

play07:28

systems um you know achieving certain

play07:30

goals and objectives uh bringing that

play07:32

kind of work and marrying that together

play07:34

with the modern large multimodal model

play07:36

work so yeah there's Deus saus talking

play07:38

about the future of agents and I think

play07:40

he's very much to right in the fact that

play07:42

that is going to bring things to a

play07:44

completely another level um and it's

play07:46

going to be completely different because

play07:47

this is where you're going to unlock

play07:49

real real value because a lot of people

play07:51

um I wouldn't say they don't use AI

play07:53

systems but it's going to be interesting

play07:55

to see how you know using those AI

play07:57

systems is you know vastly different

play07:59

than the way we interact with them now

play08:01

so I think that is going to be super

play08:03

super fascinating now here we are for

play08:05

the next part okay and the next part of

play08:08

course is voice assistance now many of

play08:10

you may have seen the new open AI demo

play08:14

and this is of course what we're going

play08:16

to be seeing a lot more of from the top

play08:19

Frontier AI labs in 2024 and of course

play08:22

early 20125 now there's actually a lot

play08:25

to discuss here because Voice Assistant

play08:28

might not be what you guys think so

play08:30

essentially in this demo here we

play08:32

actually did get to see a huge huge huge

play08:35

huge variety of different voice

play08:38

assistants and these voice assistants

play08:40

were really really really good because

play08:42

compared to the previous ones that we've

play08:44

had before like Siri they completely

play08:46

brow blow just blow everything out of

play08:48

the water so this is a short gbt 40 demo

play08:52

and then I've got to show you guys

play08:54

another demo that most people did

play08:55

actually miss from another top Frontier

play08:58

company that I guess you could say is

play09:00

working on something very very similar

play09:02

to this hey Chach PT how are you doing

play09:04

I'm doing fantastic thanks for asking

play09:07

how about you pretty good what's up so

play09:09

my friend um Barrett here he's been

play09:11

having trouble sleeping lately and uh I

play09:13

want you to tell him a bedtime story

play09:14

about robots and love oh a bedtime story

play09:18

about robots and love I got you covered

play09:22

gather round Barrett once upon a time in

play09:25

a world not too different from ours

play09:27

there was a robot named B

play09:30

bite was a curious robot always expl

play09:32

started this story but I want a little

play09:34

bit more emotion in your voice a little

play09:35

bit more drama got it let's add some

play09:38

drama once upon a time in a world not

play09:42

too different from ours there was a

play09:44

robot named I really want maximal

play09:47

emotion like maximal expressiveness much

play09:49

more than you were doing before

play09:51

understood let's amplify the drama once

play09:55

upon a time in a world not too different

play09:58

from ours there was a robot named bite

play10:02

bite can you do this in a robotic voice

play10:04

now initiating dramatic robotic voice

play10:09

once upon a time in a world not too

play10:12

different from ours there was a robot

play10:15

named bite bite was a curious robot

play10:19

always exploring new circuit Barrett

play10:21

here calm down a little bit can you end

play10:23

the story um but do it in a singing

play10:26

voice and so bite fell owned another

play10:30

robot friend and they live circly Ever

play10:35

After thank you so much I definitely

play10:38

feel a lot more

play10:42

relaxed so now the reason voice

play10:45

assistants are going to be pretty pretty

play10:46

crazy and the reason like a lot more

play10:48

people are going to be using them is

play10:50

because we've already seen the fact that

play10:52

the The Voice assistance are serving a

play10:54

pretty huge Market if you remember the

play10:56

rabbit R1 device that was I guess you

play10:58

could say a Voice Assistant that was

play11:00

much better than anything else at the

play11:02

time and although many people are

play11:04

stating that it's just a rapper product

play11:06

or yada y yada I think the point here is

play11:09

that voice assistants are truly the way

play11:11

that we are going to interact with AI

play11:13

because it just feels the most natural

play11:16

when I'm talking to GPT 40 or when I'm

play11:19

talking to you know another person I'm

play11:21

really trying to feel the same way and

play11:24

the same kind of fluidity and that's

play11:25

what most people use of course there are

play11:28

going to be some scarios where you just

play11:30

simply want to you know talk with a

play11:31

large language model or an AI system and

play11:34

get out you know uh lots of lots of text

play11:36

but I think that for the vast majority

play11:39

of people on a day-to-day basis you know

play11:41

handsfree being able to just talk with

play11:43

your AI about whatever topic it may be

play11:46

this is going to be the large majority

play11:48

of where people are you know Finding

play11:50

real real use for these systems okay uh

play11:53

and this also is great because for those

play11:55

who aren't technologically advanced uh I

play11:58

guess you could say this is going to be

play11:59

a much easier way for them to interact

play12:01

with the software now another thing that

play12:04

we do have is Google's Project Astra so

play12:07

this is the vision for Google's future

play12:09

of AI assistance and this is quite

play12:12

similar to the recent thing that we just

play12:14

saw but it's Google's very own

play12:17

interesting way of doing their voice

play12:18

assistance and this is likely to be

play12:21

released in the coming months although

play12:22

Google haven't really shipped anything

play12:25

that they like that they said they were

play12:26

going to ship but this is of course

play12:28

something thing that they're still

play12:29

working on and this demo is giving us an

play12:32

insight to where things are going to

play12:34

head so this should be giving you guys

play12:35

some creative ideas for how the

play12:38

interaction and the ecosystem around AIS

play12:40

is going to change okay let's do some

play12:44

tests tell me when you see something

play12:46

that makes sound I see a speaker which

play12:49

makes

play12:51

sound what is that part of the speaker

play12:53

called that is the Tweeter it produces

play12:56

high frequency sounds

play12:59

give me a creative alliteration about

play13:03

these creative crayons color cheerfully

play13:06

they certainly craft colorful

play13:11

Creations what does that part of the

play13:13

code do this code defines encryption and

play13:16

decryption functions it seems to use a

play13:19

ecbc encryption to encode and decode

play13:22

data based on a key and an

play13:24

initialization Vector IV that's right

play13:29

what neighborhood do you think I'm

play13:32

in this appears to be the king's cross

play13:35

area of London it is known for its

play13:37

railway station and transportation

play13:41

connections do you remember where you

play13:43

saw my glasses yes I do your glasses

play13:46

were on the desk near a red

play13:53

apple what can I add here to make this

play13:55

system faster adding a cache between the

play13:58

server and database could improve speed

play14:00

what does this remind you of shringer

play14:03

cat all right uh give me a band name for

play14:07

this Duo golden Stripes nice thanks

play14:11

Gemini so yeah that is Google's aim at

play14:14

the voice assistants and I think voice

play14:17

assistants have another area in which

play14:19

they're going to be you know moving

play14:21

forward because there are other

play14:22

companies that have had a market share

play14:24

and let me just put something on the

play14:26

screen and that is of course Amazon so

play14:29

Amazon of course are planning to give

play14:31

Alexa an overhaul and a monthly

play14:33

subscription price so Amazon is making

play14:36

significant upgrades to Alexa by

play14:39

integrating large language models and

play14:41

generative AI capabilities Amazon is

play14:44

actually developing a new version of

play14:46

Alexa powered by custombuilt llms

play14:49

optimized for voice interactions and

play14:52

this llm whichever it's going to be

play14:53

based on I think it most likely will be

play14:55

claw 3 are going to enable more natural

play14:58

con conv ation interactions with Electra

play15:01

and some key enhancements include

play15:03

conversation a visual wakeup word

play15:06

conversational speech recognition and

play15:08

expressive text to speech so this is

play15:10

going to be another step in terms of

play15:12

integrating these AI systems with the

play15:15

wider wider world and we know that there

play15:18

are already tons of people who already

play15:20

have these devices in their homes that

play15:22

they could literally just receive the

play15:23

software update and then boom you now

play15:26

have millions of users who are going to

play15:28

be interacting with voice assistants

play15:30

that are actually really really useful

play15:32

so it's going to be interesting to see

play15:34

what Amazon does and I'm sure that their

play15:37

partnership with anthropic is going to

play15:39

be a major part of that now like I said

play15:41

it before one of the things that I do

play15:44

think would be surprising is if Google

play15:47

didn't make an announcement regarding

play15:49

their home devices I'm not sure of the

play15:51

popularity of these I know that most

play15:53

people do have an Alexa I've never

play15:55

really heard anyone talking about their

play15:57

Google device but I'm pretty sure it

play15:59

would sell a lot more devices if you

play16:01

could just have a home device or a

play16:02

physical device where you could just ask

play16:04

it simple questions and it could respond

play16:07

very easily I know that's something that

play16:09

I would certainly be using so the next

play16:12

thing that is on the list for AI number

play16:15

three is of course true

play16:17

multimodality so essentially any to any

play16:21

and this is a screenshot from nvidia's

play16:23

recent conference but it gives us a

play16:25

glimpse into the future of how

play16:28

multimodal systems are truly going to

play16:30

work any kind of input that you can

play16:32

think of is going to be used and then of

play16:34

course you're going to be able to put

play16:36

that into any other kind of I guess you

play16:38

could say input/output and this

play16:40

multimodality this true multimodality is

play16:43

in some ways already here and will be

play16:46

further expanded upon so one of the ways

play16:48

that most people didn't even realize and

play16:50

this is something that I actually saw

play16:52

because I was paying attention and most

play16:54

people didn't realize but that when GPT

play16:57

40 was released they released this web

play17:00

page okay and there were an exploration

play17:02

of different capabilities and in the

play17:05

capabilities section there were a bunch

play17:07

of different you know things that showed

play17:09

us that gbt 40 was and is a completely

play17:14

multi modal model that's why they called

play17:16

it GPT 40 Omni and there were many

play17:19

different things that they just didn't

play17:21

Showcase in the demo that proved to us

play17:23

that this is the future of AI because

play17:26

it's actually a completely multimodal

play17:28

model natively built it's not like it's

play17:30

a you know large language model or

play17:32

something like that it's completely a

play17:34

natively multimotor model so I'm going

play17:35

to show you guys quickly what I mean by

play17:37

this so if you go onto the gbt 40 web

play17:40

page and if you scroll down there's this

play17:42

section right here this section is

play17:44

called an exploration of capabilities

play17:46

you can see that this is visual then you

play17:48

can see that this is poster creation and

play17:51

you can see it's able to combine

play17:52

different things you can see we have

play17:54

here this photo character you can see

play17:57

turning photos into different things

play17:58

things you can see concrete poetry

play18:00

you're able to turn this into words and

play18:02

I think you're also able to turn this

play18:04

into a badge you're also able to see

play18:06

lecture summarization so this shows you

play18:08

that you've got a 45 minute video and

play18:10

it's able to take the input from that so

play18:13

remember this is true true multimodality

play18:16

meeting notes with multiple speakers

play18:18

it's able to take in audio you can see

play18:20

right there there was also 3D objects

play18:23

synthesis so you can see right here it's

play18:25

able to generate a 3D object and able to

play18:27

rotate that in Real Time 3D

play18:30

reconstruction so this is something that

play18:32

is truly truly going to change

play18:34

everything because we have a situation

play18:36

where there's going to be true

play18:37

multimodality which means the only

play18:39

limits that you're truly going to be

play18:40

thinking about are ones in your head and

play18:44

in addition to that the multimodality

play18:46

aspect has been there with Google's

play18:48

recent AI studio so if you don't know

play18:51

Google have released their AI studio and

play18:53

in there you can actually upload audio

play18:55

you can upload images and you can upload

play18:58

video videos up to an hour long so it's

play19:01

pretty pretty crazy what you can truly

play19:04

do with true multimodality because

play19:07

currently with Google's new studio it's

play19:10

not out for everyone although it's out

play19:11

for most people but still when I'm

play19:13

trying to use it it's still pretty buggy

play19:15

so it's not something that I've actually

play19:17

been able to use into my workflow just

play19:19

yet but true multimodality I do think is

play19:22

going to change a lot of things because

play19:23

whilst images and text are good videos

play19:26

and audio are going to complete the

play19:28

entire equation then what we have is

play19:31

very interesting we have the

play19:33

personalization of these models so

play19:35

basically there is a feature that is

play19:38

only available to a select few accounts

play19:40

in select few regions namely USA being

play19:43

one of them and that is of course the

play19:44

personalization feature so some in the

play19:47

later months I'm sure that

play19:48

personalization is going to be rolled

play19:50

out to Europe and the rest of the UK and

play19:53

I'm sure that when it does it's going to

play19:55

increase the I guess you could say the

play19:57

usability of these models because

play20:00

personalization is something that

play20:01

changes the game you don't need to

play20:03

remind it all the things you've done

play20:05

before you don't need to remind it all

play20:07

of your preferences and it completely

play20:09

knows all the facts that you want it to

play20:11

know about you that make it a lot more

play20:13

easier to use on a day-to-day basis like

play20:16

for example I was trying to ask it a

play20:19

question okay and I was like okay what

play20:21

is the best restaurant or what type of

play20:23

food should I eat today and you have to

play20:25

remember that when you open a new chat

play20:27

with chat GPT or any AI system it

play20:30

doesn't have this area of memory which

play20:32

means its only data point that it has is

play20:35

that a human wants to eat some kind of

play20:37

food it doesn't have your age it doesn't

play20:39

have your location it doesn't know what

play20:40

you like to eat it doesn't know if

play20:42

you're allergic to anything all it knows

play20:44

is that human wants to eat food so when

play20:46

you give it that command it's going to

play20:48

be able to give you a very broad answer

play20:50

that's very vague that can satisfy all

play20:52

of those things however if you're like

play20:54

okay I'm a male I go to the gym I'm very

play20:57

active I'm trying to get Le C it might

play20:59

suggest here some protein especially if

play21:01

you like a sweet taste try this with

play21:03

honey like some beef and like it will

play21:05

try to suggest things that actually are

play21:07

useful for you and that's going to give

play21:09

a huge huge huge boost in terms of the

play21:11

usability because the more data you give

play21:14

these models the the better they are and

play21:16

that's something that I've known in my

play21:17

personal experiences when I can't get

play21:19

these models to do something right I

play21:21

just give it more and more context and

play21:23

the more context I give it the better

play21:25

the responses from these models are and

play21:27

here you can see Sam Alman and both

play21:28

Gates actually speaking about this

play21:30

customizability and personalization will

play21:32

also be very important people want a

play21:35

very different very different things out

play21:36

of gp4 different styles you know

play21:39

different sets of assumptions we'll make

play21:41

all that possible and then also the

play21:43

ability to have it use your own data so

play21:45

the ability to know about you your email

play21:47

your calendar how you like appointments

play21:49

booked connected to other outside data

play21:50

sources all of that those will be some

play21:52

of the most important area so yeah

play21:53

that's um Alman talking about how those

play21:55

are some of the you know most important

play21:57

areas for the AI systems um and of

play22:00

course the next thing that we do have

play22:02

and this is actually Point number five

play22:03

although it does say number four is of

play22:05

course the reasoning capabilities um and

play22:08

the reliability of the model that just

play22:09

means the model doesn't make mistakes

play22:11

and the model is a lot smarter we'll be

play22:13

able to push that much further but maybe

play22:15

the most important areas of progress

play22:17

will be around reasoning ability right

play22:19

now gp4 can reason and only extremely

play22:22

limited ways and also reliability you

play22:24

know if you if you ask gp4 most

play22:27

questions 10,000 times one of those

play22:29

10,000 is probably pretty good but it

play22:31

doesn't always know which one and you'd

play22:32

like to get the best response of 10,000

play22:34

each time so that'll be that that that

play22:36

increase in reliability will be

play22:38

important and so yeah Sam mman says

play22:40

reliability and reasoning ability is

play22:42

going to be pretty pretty important now

play22:44

samman has spoken about this and if you

play22:47

don't know what he means by that he

play22:48

basically just means that the model is

play22:49

just going to really understand truly

play22:51

what you want and it's never going to

play22:52

make a mistake again I wouldn't say

play22:54

never because I think that's pretty hard

play22:56

to say but um you know it's pretty much

play22:58

going to be not as much as it does right

play23:01

now so he talks about this and I've

play23:03

included this clip here because he

play23:05

basically says how gp4 is quite stupid

play23:08

and it's quite dumb uh and that's to

play23:10

point out the reasoning capabilities of

play23:12

the model are pretty limited and if you

play23:13

don't know what I mean if you want to

play23:15

take a quick demo you can like there was

play23:16

a question that people used to ask to

play23:18

GPT 4 and the question that they used to

play23:20

ask it was like if someone moved

play23:22

something from one bag to another but

play23:24

the bags were transparent um could you

play23:26

see what's inside and the model doesn't

play23:28

realize that the bags are transparent

play23:29

and that's just basically uh part of its

play23:31

reasoning capability to be like wait a

play23:33

minute the bags are transparent I should

play23:34

be able to see exactly what's inside um

play23:37

but it can't so things like that you

play23:39

know you you might not use it on a

play23:40

day-to-day basis if you're using it to

play23:42

write essays but um an increased

play23:44

reasoning capability means that there's

play23:45

going to be a lot more wide applications

play23:47

which means it's going to impact a lot

play23:48

more industry so uh here's where samman

play23:51

talks about the fact that gpc4 is quite

play23:54

dumb um I I mean this this sounds like a

play23:56

copout answer but I think the most

play23:58

important thing about GPT 5 or whatever

play23:59

we call that is just that it's going to

play24:01

be smarter and this sounds like a Dodge

play24:04

but I think that's like among the most

play24:06

remarkable facts in human history that

play24:08

we can just do something and we can say

play24:10

right now with a high degree of

play24:12

scientific certainty GPT 5 is going to

play24:13

be smarter than a lot smarter than GPT 4

play24:15

GPT 6 is going to be a lot smarter than

play24:17

GPT 5 and we are not near the top of

play24:19

this curve and we kind of know what know

play24:21

what to do and this is not like it's

play24:23

going to get better in one area this is

play24:24

not like we're going to you know it's

play24:26

not that it's always going to get better

play24:27

at this eval or this subject or this

play24:29

modality it's just going to be smarter

play24:32

in the general sense and I think the

play24:34

gravity of that statement is still like

play24:35

underrated like I said before the

play24:37

gravity is definitely going to be

play24:39

underrated because of the vastly smarter

play24:41

model we've seen what the capabilities

play24:43

are the jump from GPT 3.5 to4 meant that

play24:46

people could actually use GPT 4 for you

play24:49

know certain things like code certain

play24:51

things like writing uh you know a lot

play24:53

more I guess you could say things that

play24:55

you know a lot more meaningful um and

play24:56

there are just a lot more tasks that you

play24:58

can do and of course this is where he

play24:59

talks about the fact that gbt phenomenal

play25:02

chat gbt is phenomenal um everything

play25:04

else all the other models are phenomenal

play25:05

it burnt you've earned $520 million of

play25:07

cash last year that doesn't concern you

play25:10

in terms of thinking about the economic

play25:11

model of how do you actually where's

play25:13

going to be the monetization source well

play25:15

first of all that's nice of you to say

play25:17

but Chachi PT is not phenomenal like

play25:19

chbt is mildly embarrassing at best um

play25:22

gp4 is the dumbest model any of you will

play25:24

ever ever have to use again by a lot um

play25:27

but you know it's like important to ship

play25:29

early and often so yeah pay attention to

play25:31

what he said there it's going to be the

play25:33

dumbest model you have to use for a

play25:35

while by a lot so I think that's going

play25:37

to be something that's really really

play25:40

really important now coming in at number

play25:43

six is of course custom trained model so

play25:47

here what you can see is a uh model that

play25:50

is trained for Law and you can see that

play25:53

it's going against gp4 so basically

play25:56

Harvey the company Harvey developed a

play25:58

partnership with open ey and they

play26:00

developed a custom case law model so

play26:02

basically they have foundational models

play26:04

and they use public apis and they

play26:06

attempted to fine-tune Foundation models

play26:08

using public apis and retrieval

play26:11

augmented generation systems but these

play26:13

systems weren't really good because they

play26:15

needed deep domain knowledge and

play26:17

sophisticated reasoning and then they

play26:19

decided to collaborate and partner with

play26:21

open AI to train a model specifically

play26:24

for legal case law research and this

play26:26

involved integrating a vast amount of

play26:28

legal data and domain specific knowledge

play26:31

into the base models and then they added

play26:34

case law from Delaware and gradually

play26:36

expanded to include the entire Corpus of

play26:39

the US case law and this added around 10

play26:41

billion tokens worth of legal data to

play26:44

the model so this was a highly

play26:46

collaborative effort from open AI with

play26:48

lawyers explaining the intri intricacies

play26:51

of case law research and AI researchers

play26:54

demonstrating model capabilities and

play26:56

this is where openi provided insights

play26:59

into the technical aspects and

play27:01

adjustments needed to refine the model

play27:03

so the custom train model can handle

play27:05

tasks that require detailed legal

play27:07

reasoning and the ability to draw from

play27:10

extensive legal knowledge and it's

play27:12

capable of generating comprehensive

play27:14

answers to complex legal queries and

play27:16

drafting documents and one of the main

play27:19

improvements is that this model's

play27:20

ability to reduce hallucinations and

play27:23

ensure that every statement is backed by

play27:25

cited case law and this makes the

play27:27

outputs from this model highly reliable

play27:29

and useful for legal professionals who

play27:31

need accurate references so overall what

play27:34

we can see here is the fact that in the

play27:36

future we're going to have models that

play27:38

are custom trained for specific use

play27:41

cases um and law is going to be one of

play27:44

them whilst yes GPT 4 can be used to you

play27:47

know argue certain things I think it

play27:49

shows here that when you have a custom

play27:51

trained model the use cases go up a

play27:54

significant amount um and that is

play27:56

something that we're probably going to

play27:57

see a lot more more often now next is

play27:59

something that I think we are definitely

play28:01

likely to see because the rate of

play28:04

improvement here has been very very

play28:06

interesting so this is Healthcare and Ai

play28:09

and there are a few companies that have

play28:11

been silently working behind the scenes

play28:13

to I guess you could say develop certain

play28:15

Healthcare products that are going to be

play28:17

ready very very soon and I think they're

play28:20

going to take advantage of these new AI

play28:22

systems and it's going to be a really

play28:25

really interesting thing currently what

play28:27

your looking at is of course the

play28:30

benchmarks on the med QA the medical

play28:32

question and answering QA and you can

play28:34

see right here that GPT 4 has been

play28:37

surpassed by Google Geminis Med Gemini A

play28:40

specialized version for answering

play28:42

medical questions so essentially it's a

play28:45

family of AI models fine-tuned

play28:47

specifically for multimodal medical

play28:49

applications and those models are

play28:51

capable of understanding and analyzing

play28:54

complex Medical Data including text

play28:57

images videos and electronic health

play28:59

records like I said the multimodal

play29:01

capabilities of Gemini mean that these

play29:04

models are now able to digest a variety

play29:06

of different information and do Advanced

play29:09

reasoning on it and it's shown

play29:10

remarkable performance achieving

play29:12

state-of-the-art accuracy of

play29:15

91.1% on the med QA Benchmark which is

play29:18

of course used for evaluating medical

play29:20

knowledge similar to the US medical

play29:22

licensing exam and like I said it's not

play29:25

just research Google research and Fitbit

play29:27

are actually building personal health

play29:29

llms to offer personalized coaching and

play29:32

recommendations through a mobile app so

play29:34

leveraging Google's family of generative

play29:36

AI models in Gemini and sufficient

play29:38

Health Data from Fitbit and pixel

play29:41

devices this initiative aims to provide

play29:43

tailored guidance based on individual

play29:45

health and fitness objectives so Google

play29:48

have just recently announced that they

play29:49

are working on this and they are

play29:50

actively trying to bring this to Fitbit

play29:53

and not just Google but apple apple

play29:56

plans a AI powered health coach coaching

play29:58

service a mood tracker and iPad Health

play30:00

app so this is the move part of iPhones

play30:03

make us push into health services so

play30:06

this is going to be something that is

play30:07

planned for iOS 17 maybe we're going to

play30:09

get an announcement at Apple's new not

play30:12

conference but I think it will be

play30:13

interesting to see what actually happens

play30:15

because this is something that I think

play30:17

is very very underrated and this is

play30:19

definitely going to impact Healthcare in

play30:21

a very positive way because it's going

play30:23

to bring down I guess you could say

play30:24

access to high quality information next

play30:27

of course we have something that is the

play30:29

autonomous expansion so if you don't

play30:32

know autonomous cars are a real thing

play30:34

and they're basically cars that drive

play30:36

without you they don't need a driver and

play30:38

you can sit in the passenger seat and

play30:41

you know get taken to your destination

play30:42

it's really weird but it actually makes

play30:44

a lot of sense so essentially what

play30:46

you're looking at here is a few tweets

play30:48

from people that have you know been

play30:49

firsthand in these Whos so you can see

play30:52

right here this person says so weo is

play30:53

actually amazing using it in SF for

play30:56

every trip now and it's just rock solid

play30:58

on a ride now and it just navigated a

play31:00

bunch of really odd and expected on

play31:02

Street nonsense with confidence and ease

play31:04

and someone says I now prefer

play31:06

self-driving Whos over Ubers I actually

play31:08

feel safer and someone said I just tried

play31:10

Whos self-driving taxis it's more

play31:12

comfortable than taking an Uber AI is

play31:14

changing everything even for taxi and

play31:16

Uber and the thing is is that this is

play31:20

expanding this kind of stuff is going to

play31:21

be expanding into many different areas

play31:24

and eventually I do think that it's

play31:26

probably going to be across the US but

play31:28

you can see here that it says later this

play31:30

year weo will be offering rides to the

play31:32

public in four major cities starting

play31:34

tomorrow March 14th our fully autonomous

play31:37

ride hailing service will be available

play31:39

to select members of the public in Los

play31:41

Angeles and after starting initial Rider

play31:44

only testing in Austin last week we plan

play31:47

to offer wh one to austinites later this

play31:50

year so that is pretty crazy they going

play31:53

to be expanding this and we're likely to

play31:55

see a lot more of the user feedback from

play31:58

this and I think with some of the other

play32:00

announcements especially going on in the

play32:01

UK as well with the fact that the

play32:03

automated Vehicles Act is now law it

play32:06

means that self-driving Vehicles could

play32:08

be on British roads in Just 2 years and

play32:10

it's going to be a very very interesting

play32:12

time now with that we have Point number

play32:15

nine now what else is coming and this is

play32:17

customer service AI so if you're someone

play32:19

that interacts with the business which

play32:21

is all of us anytime you have to speak

play32:23

to customer service there is usually

play32:26

some tedious interaction but that's

play32:28

about to change take a look at Google's

play32:30

customer service AI That's changing the

play32:32

game behind the scenes symbol Fashions

play32:35

customer agent is using Google Cloud's

play32:37

full Suite of AI capabilities to offer

play32:40

customized support interactions you know

play32:42

facilitate transactions like purchases

play32:44

and returns and ensure that I'm

play32:47

receiving the most upto-date information

play32:50

in real time I'm so close to having this

play32:53

shirt for the concert let's give the

play32:55

store a call

play32:58

hi there this is the symbol fashion

play33:00

customer agent at South Las Vegas

play33:02

Boulevard am I speaking with Amanda yes

play33:05

this is Amanda great thanks for reaching

play33:08

out Amanda I see you had a session on

play33:10

another device I've sent you an SMS

play33:13

message with a link to our live chat

play33:14

companion if you would like to switch to

play33:17

chat please click the link how can I

play33:19

help you today I'd like to purchase the

play33:21

shirt in my cart with the cart I have on

play33:23

file absolutely I see you're also a

play33:26

symbol fashion Awards member looks like

play33:29

you have a 20% off voucher available to

play33:31

use would you like to apply it to this

play33:33

purchase yes please that would be great

play33:36

the shirt you're purchasing goes well

play33:38

with these items also available for

play33:40

pickup in your preferred size would any

play33:43

of these be interesting to you

play33:44

absolutely please add the white shirt

play33:48

and the boots to my cart great your

play33:51

total is $

play33:53

23.73 okay to proceed with the card on

play33:55

file yes your purchase is confirmed do

play33:58

you need anything else today no I'm all

play34:02

set thank you incredible thank you for

play34:05

shopping with symbol fashion you'll get

play34:07

a text when the items are ready for

play34:11

you in less than 5 minutes I was able to

play34:15

find and order the shirt I wanted and an

play34:18

outfit to match I'm ready for the

play34:21

concert back to you Thomas many of so

play34:24

you can see right there that that is

play34:26

truly going to change the game and and I

play34:27

think it's I think it's real and in the

play34:30

sense that like this is going to be

play34:31

something that actually helps companies

play34:32

out because you know customer service is

play34:34

just tedious you got people always

play34:36

complaining people not being able to

play34:37

find stuff and I think ai's ability to

play34:40

you know uh rapidly access information

play34:42

be able to be patient you know not get

play34:45

annoyed if someone raises their voice I

play34:47

think this is going to be an area that

play34:48

humans don't even want to do anyways so

play34:50

this is going to be something that's

play34:51

really impressive now of course with

play34:53

customer service AI you do have other

play34:56

areas here uh and Wendy's has been

play34:58

actually I guess you could say testing

play34:59

this out i' I've shown this a couple of

play35:01

times but this is very relevant because

play35:03

I think that you know more

play35:03

drive-throughs are probably going to use

play35:06

this stuff provided it's effective as as

play35:08

it's shown

play35:11

here welcome to Wendy's what would you

play35:13

like can I have a chocolate

play35:18

frosty which size for the chocolate

play35:20

frosty

play35:24

medium can I get you anything else today

play35:27

no thank

play35:33

you great please pull up to the next

play35:37

window so yeah you can see uh the

play35:40

differences in how things are going to

play35:42

go and I think this would be really good

play35:44

because I've seen too many videos of

play35:45

people taking the Mickey out of people

play35:48

in drive-throughs and it's quite

play35:50

frustrating cuz some people just want to

play35:52

work their jobs and people are like

play35:53

trying to make funny content out of them

play35:55

so if we just had AI in this position

play35:57

that could speak all the languages

play35:58

understand exactly what you said be

play36:00

completely patient I think that that's

play36:01

going to be something that's really

play36:03

really cool now another thing that we do

play36:05

have and coming in at Point number 10

play36:07

last but not least is of course even

play36:10

better robots so we currently looking at

play36:12

is boss and Dynamics new Atlas which is

play36:15

pretty pretty incredible we haven't seen

play36:17

a full demo yet but from what they've

play36:20

shown us they've shown us that they are

play36:22

about to embark on a new humanoid

play36:24

robotics run that if this robot can move

play36:26

anything like their old Atlas I think

play36:28

people are going to be in for a real

play36:30

real surprise on how this robot is going

play36:32

to move considering it's completely

play36:34

electric um and I don't think it's

play36:36

hydraulic based like the previous one so

play36:38

we're about to get a huge new I guess

play36:41

you could say not robotic spoon but like

play36:43

a huge huge focus on robotics because

play36:46

robots have been I guess just going

play36:48

through an incredible period of growth

play36:50

and you can see here that even recently

play36:52

there was this robot here at one time

play36:54

speed being able to do just incredible

play36:57

tasks that I still struggle to believe

play36:59

that this is a robot with no

play37:00

teleoperation so it's incredible at what

play37:03

some of the companies around the world

play37:05

are being able to do with robotics so I

play37:07

think you know later on this year or

play37:08

even in 2025 there's going to be a few

play37:11

things that show us that you know robots

play37:13

are going to be doing a lot more

play37:15

different tasks and this is especially

play37:17

combined with the fact that you can see

play37:18

right down here these things are

play37:20

combined with some of the multimodal AIS

play37:23

and of course these llms integrated with

play37:25

them so it's going to be a very very

play37:27

interesting time to see what kind of

play37:30

things we do get with robots and not

play37:32

just the fact that this is like one time

play37:34

speed no teleoperation which is just

play37:37

genuinely mind-blowing that a robot is

play37:39

able to just do this because it now

play37:40

means you know in the future when you

play37:42

know people are talking about robots

play37:43

doing a lot of household tasks and

play37:45

things like this and remember guys this

play37:47

is no teleoperation which means there's

play37:49

no human uh in another room that's

play37:51

controlling this this is a robot

play37:52

completely autonomously doing this based

play37:55

on a set task um and I think you guys

play37:57

need to also take a look at the fact

play37:59

that China has been developing a ton of

play38:02

humanoid robots this is only like half

play38:04

of them um and so far the demos that

play38:07

we've seen the infrastructure that

play38:08

they're they're building you know

play38:09

everything is being you know put in

play38:11

towards humanoid Robotics and Ai and I

play38:13

think China definitely wants to be a

play38:15

leader in that space because they

play38:17

understand the economic value and even

play38:19

recently we got unry G1 which is an AI

play38:23

agent Avatar as described and it's a

play38:26

humanoid robot that is only

play38:28

$166,000 which means the decreasing cost

play38:31

of these robots means that researchers

play38:34

can afford to purchase these systems and

play38:37

apply their research techniques on these

play38:39

systems because a lot of times the the

play38:41

problem is that you know these robots

play38:42

are pretty pretty expensive especially

play38:45

some of the cool ones that you're seeing

play38:46

in these demos they're like $250,000 to

play38:49

half a million dollars and I mean if

play38:50

you're a researcher at a university it

play38:52

can be pretty pretty impractical to just

play38:55

buy one so that you can do some research

play38:57

Arch on it but if we can get these

play38:59

things down with economies of scale it

play39:01

could open up a huge huge huge area for

play39:05

where we're seeing rapid progress on

play39:07

these devices so that was the 10 things

play39:11

that I think are going to be coming in

play39:13

the next 365 days for AI robotics slth

play39:17

Singularity and those are going to be

play39:19

things that you truly do see here on

play39:21

this channel if there was anything I did

play39:23

miss don't forget to leave a like down

play39:24

below don't forget to comment down below

play39:26

and let me know what thing you are most

play39:28

excited for cuz I would love to know

Rate This
โ˜…
โ˜…
โ˜…
โ˜…
โ˜…

5.0 / 5 (0 votes)

Related Tags
AI Trends2024 PredictionsArtificial IntelligenceAgents EvolutionVoice AssistantsMultimodal AIPersonalizationHealthcare AIAutonomous VehiclesRobotics Advancement