AI News: The AI Arms Race is Getting INSANE

Summary

Google and Open AI are making significant strides in the AI industry, with key personnel changes and technological advancements. Ilya Sutskever, a prominent figure in AI, is parting ways with Open AI, while Jan Leike resigns, signaling potential shifts in the field. Open AI appoints Jacob Patok as the new Chief Scientist, who previously led the development of GPT-4 and Open AI Five. Meanwhile, Google introduces Project Astra, an AI assistant with real-world applications like email organization and multi-step task automation. The competition between AI models like Google's Pali and Poly and Open AI's offerings is heating up, with a focus on creating AI agents that can enhance various aspects of life and business, from customer service to personalized assistance.

Takeaways

• 📢 Google and OpenAI are experiencing significant changes with key personnel departures, including Ilya Sutskever from OpenAI and Jan Ley, a machine learning researcher.
• 👕 There is a sense of humor in the script as the narrator asks where to get Ilia's shirt, indicating a lighter moment amidst the news.
• 🤖 OpenAI is undergoing a transition with Ilya Sutskever's departure and Jacob Patok stepping up as the new Chief Scientist, previously involved in the development of GPT-4 and OpenAI Five.
• 🎮 OpenAI Five, an AI developed by OpenAI, was highlighted for its ability to beat human players in Dota 2, showcasing the potential of AI in gaming and teamwork.
• 🔍 The script discusses the broader implications of AI, including the potential for generative AI to replace call centers and improve customer service efficiency.
• 🕶️ Google is hinting at a comeback for AR glasses, which could integrate AI features and provide an overlay of information in the real world.
• 🧩 Project Astra by Google is presented as a new AI assistant with capabilities to interact with objects and provide information, though early demos had mixed results.
• 📈 Google also introduced an email agent that can organize receipts into spreadsheets, automating financial tracking for users.
• 👟 Gemini, Google's AI, showcased its ability to handle tasks like returning orders with minimal user input, demonstrating advances in AI's practical applications.
• 🎨 Google's new AI tool, Vo, is introduced as a competitor to OpenAI's DALL-E, capable of creating tech-to-video animations.
• 🚀 The script emphasizes the rapid development and competition in the AI field, with new models and applications being released by both Google and OpenAI.

Q & A

What is the significant news about Ilia Sutskever and OpenAI?

Ilia Sutskever has decided to part ways with OpenAI, which is a significant event given his role as one of the greatest minds in the field and a co-founder.

Who is Jacob Patokai and what is his new role at OpenAI?

Jacob Patokai is the new Chief Scientist at OpenAI. He previously served as the director of research and spearheaded the development of GPT-4 and OpenAI Five.

What was the initial mission of OpenAI when it was formed by Ilia Sutskever, Elon Musk, and Greg Brockman?

The initial mission of OpenAI was to build AGI (Artificial General Intelligence) and serve as a counterbalance to Google's influence in the AI field.

What is the significance of OpenAI Five's victory in Dota 2?

OpenAI Five's victory in Dota 2 was significant because it was the first time AI beat eSports pros in a live stream, showcasing the advanced capabilities of AI in competitive gaming.

What is the current status of generative AI in customer services as mentioned in the script?

Generative AI in customer services is now capable of doing the job of 700 people, providing higher customer satisfaction, faster response times, and higher quality interactions than human agents.

What is Google's response to the advancements in AI and generative AI models?

Google is developing its own suite of AI products and agents, with features like AR glasses and multi-step researching and reasoning capabilities, to stay competitive in the AI space.

What is Project Astra and how does it work?

Project Astra is Google's answer to OpenAI's assistant. It uses AI to provide quick responses and perform tasks like organizing receipts into spreadsheets or assisting with returns by processing information from images.

What is the significance of the 2 million context window in Google's AI models?

The 2 million context window is a significant step towards Google's goal of an infinite context window, allowing AI models to process and understand vast amounts of information more effectively.

What is the potential impact of AI agents on businesses and industries?

AI agents have the potential to revolutionize businesses by automating tasks, improving efficiency, and providing insights through real-time data analysis, making them a critical skill for future business management.

What are some of the new AI models and tools introduced by Google in the script?

Google introduced models like PolyGamma for image captioning and visual tasks, Gemini 2 for high performance at various developer sizes, and tools like Google's custom AI models and AI agents for various Google products.

Outlines

00:00

😀 Departures and Changes at Open AI

The script discusses significant changes within Open AI, starting with Ilya Sutskever's departure and Sam Altman's reaction to it. Altman expresses gratitude for Ilya's contributions and introduces Jacob Padok as the new Chief Scientist. The script also touches on the past achievements of Open AI, such as Open AI Five's success in Dota 2, and hints at future projects that Ilya might be involved in, possibly starting a new AI company. It also mentions Jan Ley, another key figure leaving Open AI, and the implications of these departures on the company's future.

05:00

🤖 Advancements in AI and BCI Technology

This paragraph covers a range of topics, from the potential for new AI companies to breakthroughs in brain-computer interface (BCI) technology. It discusses a paralyzed man using BCI to write Chinese characters, marking a significant advancement in BCI applications. Additionally, it covers the use of AI in video games, with AI models controlling characters in Street Fighter, and the potential for generative AI in customer service, suggesting that AI could replace human agents with higher efficiency and customer satisfaction.

10:01

🕶️ Google's Project Astra and AI Developments

The script details Google's Project Astra, an AI assistant that can interact with the environment through a camera and provide information or create stories. It also mentions the mixed results from a demo of the project, with some users experiencing difficulties. Furthermore, it discusses Google's other AI initiatives, including an email agent for organizing receipts and a search agent for multi-step tasks. The paragraph also touches on Google's plans for AI glasses and the potential return of AR technology.

15:03

📈 Google's AI Models and Market Share Insights

The paragraph delves into Google's AI models, such as Gemini 1.5, which is designed for quick, high-frequency tasks at a low cost. It also mentions the introduction of a 2 million context window and the goal of achieving an infinite context window. Additionally, it covers the potential applications of vision language models in security and healthcare, as well as Google's plans for new AI agents and their integration with various Google products. The paragraph concludes with a look at global smartphone sales share, highlighting Android's dominance.

20:04

🛠️ Building Custom AI Agents and Learning Opportunities

This section emphasizes the importance of learning to build custom AI agents, likening the future of business to managing a complex system where AI agents perform specific tasks. It mentions the partnership between Andrew Ng and Crew AI to offer a course on multi-agent systems, suggesting that this is a valuable skill set for the future. The script also promotes a private group focused on learning and building AI autonomous agents, aiming to create a valuable community around AI applications.

25:05

🎓 Encouraging Continued Learning and Community Building

The final paragraph serves as a call to action, encouraging viewers interested in AI to join a community focused on learning and building AI autonomous agents. It highlights the rapid growth of the community and the goal of making it the most informative and valuable resource for those looking to harness AI for various purposes. The script ends with a personal note from the presenter, thanking viewers for watching and inviting them to join the community.

Keywords

💡AGI

AGI stands for Artificial General Intelligence, which refers to a form of artificial intelligence that possesses the ability to understand, learn, and apply knowledge across a wide range of tasks at a level equal to or beyond that of a human. In the video's context, AGI is a central theme as it discusses the developments and implications of AI reaching this level, particularly with the departure of Ilia Sutskever from OpenAI and the future of AI research.

💡OpenAI

OpenAI is a research laboratory that focuses on the development and application of AI technologies. In the script, OpenAI is highlighted as a key player in the field of AI, with significant news about Ilia Sutskever's departure and the appointment of Jacob Patok as the new Chief Scientist, indicating shifts in the organization's direction and leadership.

💡Ilia Sutskever

Ilia Sutskever is a prominent figure in the AI community, known for his work at OpenAI. The script discusses his departure from OpenAI, which is a significant event in the AI field, as he is considered one of the greatest minds of his generation and a guiding light in the field of AI.

💡Jacob Patok

Jacob Patok is introduced in the script as the new Chief Scientist at OpenAI, taking over from Ilia Sutskever. He previously served as the director of research and was instrumental in the development of GPT-4 and OpenAI Five, an AI that defeated human players in the game Dota 2.

💡Anthropic AI

Anthropic AI is mentioned in the script as a company that has attracted high-profile talent from OpenAI, such as Ilia Sutskever and others. The term is relevant as it suggests a shift in the AI landscape, with key figures moving to new ventures that may influence the future of AI development.

💡AI agents

AI agents in the script refer to AI systems that can perform tasks autonomously, such as browsing the web or organizing emails. The video discusses the increasing role of AI agents in various aspects of life and business, highlighting their potential to transform industries and improve efficiency.

💡Google

Google is a major technology company that is also deeply involved in AI research and development. The script mentions Google's response to OpenAI's advancements, with Google developing its own suite of AI products and services, including Project Astra, which is an AI assistant with augmented reality capabilities.

💡Project Astra

Project Astra is Google's initiative to create an AI assistant with augmented reality features. The script describes a demo where Astra is shown to quickly and intelligently respond to various prompts, indicating the potential for AI to provide immediate and contextual assistance in everyday tasks.

💡AI and customer service

The script discusses the impact of AI on customer service, with generative AI being able to perform the job of hundreds of human agents. It highlights how AI can improve customer satisfaction, speed, and quality of service, suggesting a shift towards AI-driven customer interactions.

💡AI in gaming

AI in gaming is showcased in the script through examples like OpenAI Five and the use of AI models in fighting games. It demonstrates the versatility and capability of AI to learn and excel in complex tasks, such as competitive gaming, which can have implications for AI's broader applications.

💡Brain-computer interface

A brain-computer interface (BCI) is a technology that enables direct communication between a human brain and a computer. The script mentions a paralyzed man using a BCI to write Chinese characters, showcasing the potential of BCI to restore communication and mobility for individuals with disabilities.

Highlights

Sam Altman announced feeling AGI today and shared a shot with the OpenAI core team.

Ilia Sutskever, co-founder of OpenAI, is parting ways with the company.

Jan Leike, machine learning researcher co-leading Superalignment at OpenAI, resigned.

Multiple people leaking information from OpenAI have been terminated.

Jacob Patoci is the new Chief Scientist at OpenAI, previously director of research for GPT-4 and OpenAI Five.

OpenAI Five successfully defeated human players in Dota 2, showcasing advanced AI capabilities.

Ilia Sutskever thanked the team and hinted at starting a new project.

Anthropic AI's CTO Mike Krieger joined the team as Chief Product Officer.

A robot with a brain-computer interface enabled a paralyzed man to write Chinese characters using thoughts.

Project Astra by Google aims to bring back AR glasses with AI integration for various applications.

Generative AI in customer service is performing tasks equivalent to 700 people, enhancing efficiency and customer satisfaction.

Google's Project Astra demonstrated advanced AI capabilities, though mixed reviews were received on its performance.

Gemini 1.5 Flash by Google offers fast, multimodal model with low-cost token processing.

Gemini 2 aims to provide industry-leading performance with efficient operation on GPUs or TPUs.

PolyGemma, a vision-language model by Google, is optimized for image captioning, visual Q&A, and other tasks.

Google announced plans for AI agents capable of organizing emails and managing receipts automatically.

Transcripts

some big news today out of Google and

open AI first and foremost Sam Alman

says he's feeling the AGI today here's a

shot with the opening eye core team and

to everyone's great relief Ilia satk is

okay he's alive seems happy and

everyone's looking pretty cheerful my

only question is where do I get this

shirt Ilia where did I get this shirt

unfortunately this isn't because of

anything happy and pleasant happening it

looks like Ilia and open AI are going to

part ways not only that but Jan Ley

machine learning researcher co-leading

super alignment at openi well he says I

resigned which of course people are

saying is not a good sign now of course

we have multiple people that have been

leaking information out of open AI have

been terminated Ilia is leaving John Ley

is leaving what is happening well let's

start here s posted this Ilia and open

AI are going to part ways this is very

sad for me Ilia is easily one of the

greatest minds of Our Generation a

Guiding Light in our field and a dear

friend now I'm not going to read all of

this because you know in times like this

you generally post the things that

you're supposed to say the things that

everyone expects you can even notice

that Sam Alman drops his usual casual un

capitalized letters and this is more

formal I'm sure he means what he says

but this obviously is polished because

it's going to be going to a lot of

people as you can see here 4.2 million

views there are a couple things that

caught my eye he's saying I am forever

grateful for what he did here and

committed to finishing the mission we

started together which was building open

AI building AGI being a counterbalance

to Google that was the initial role that

him and Elon and Greg formed back in

2016 2017 whenever that was or

potentially even started talking about

in 2015 if I recall correctly and Jacob

is going to be our new Chief scientist

so that's Jacob patoki he's right there

on the left here in this shot and he

previously served as the director of

research spearheading the development of

GPT 4 and open I5 you might not remember

open5 or not be familiar with it it was

from a very very long time ago very

ancient time if we're talking about AI

timelines from 2019 where the open ai5

complet

wiped the floor with the human players

in Dota 2 so of course both open5 and

deep mindes Alpha star were the two kind

of competing AI that both the companies

open Ai and Deep Mind tried to make them

competitive at playing these blizzard

games or at least you know game ideas

that at some point were part of Blizzard

games cuz DOTA did start out as a custom

map in Warcraft 3 I believe but here's

that shot I see Greg right there in the

background he's right there there is is

Jacob and here's kind of a replay of

that match I believe so I don't know the

exact specific details but as you can

see here the open AI 5 are called uh the

names of the players are open AI 3 open

AI 2 and then it says bot in quotations

there they are and yeah I mean kind of

looks like they're playing in a similar

fashion that a real life human player

would play moving back and forth

retreating Etc one interesting thing

that I heard during the interview with

the human players that were playing

against these Bots afterwards they

mentioned that the one thing that they

found kind of disturbing about playing

with the Bots is that the opening ey

Bots had a very good understanding of

when to engage in a fight and when not

to meaning that if they were retreating

it was probably because they knew that

they would lose if they attacked there

was a very high chance that they would

indeed win so as the human players

picked up on that they realized that if

the Bots engaged in a fight that would

probably not be good for them and that

is GG game over open AI taking game two

taking the series two to zero this was

the first time in AI beating Esports

Pros on live stream but the other big

finding was while the training process

was exclusively focusing on beating

other Bots the opening I5 discovered the

rudimentary ability to be a teammate

with humans so kind of cooperative play

playing alongside humans to defeat the

other team and so now Jacob is stepping

up as the new Chief scientist at open AI

ilas husk also posted recently saying

after almost a decade I made the

decision to leave open AI the company's

trajectory has been nothing short of

miraculous and I'm confident that openi

will build AGI that is both safe and

beneficial under the leadership of Sam

Greg Mir Madi and now under the

excellent research leadership of I'm

guessing that's Jacob Jacob Pachi at

open AI Jacob thanked Ilia for all of

his wisdom and mentorship and Ilia is

saying I'm excited for what comes next

the project that is very personally

meaningful to me about which I will

share details in due time could it be

that Andre and ilas start a truly open

AI company is it possible that a lot of

these people are leaving to strike off

on their own certainly that's seems to

be what happened with anthropic

anthropic AI by the way also is not

sitting this one out as the co-founder

and CTO the chief technical officer of

insta has joined anthropic AI as their

Chief product officer Mike creger he's

going to be diving in and helping shape

the future of AI powered products at

anthropic and on the other side of the

world there's this a robot is able to

draw Chinese characters as a man using a

computer brain interface somewhat

similar to the neuralink is right those

characters out on a screen the post says

robots with the help of neuro implants

helped a paralyzed man a 76-year-old

paralized man has made history by using

his thoughts to write eight Chinese

characters this incredible feat marks

the first successful use Shang

University's brain implants to enable

writing Chinese characters the patient

controls a robotic hands with his

thoughts to write the Chinese characters

offering a hope for those who have lost

speech and Mobility to communicate again

and they're noting here that unlike like

alphabetic writing implementing Chinese

script through a brain computer

interface is much more complex the

volunteer imagines the writing process

which is reflected in the motor area's

neuronal activity then the system

analyzes the signals to capture the

intended writing path and control the

robotic arm to right Gro is rebuilding

the experiment we had two llm models

fight each other in Street Fighter this

time of course using the gro chip to

make it very very fast as you can see

here I mean the characters are moving

very quickly not a lot of lag not a lot

of downtime they're constantly at each

other blocking and countering moves it

seems like very impressive so this is

Gemma 7B versus Chad bt4 turbo

controlling two Ken characters in a

retro game they decided to leave out the

trademark name of the video game you can

see how many more moves Gemma produces

in the text screen and the red text so

yeah certainly there's a lot more red

text uh than they are green text I just

think it's very interesting when they

use video games to show various

abilities of AI I think it's a very good

combination and and where possible we

should use video games to demonstrate AI

I think it should be one of the rules

like we have the three laws of robotics

if we ever have some laws for AI one of

them should be thou shalt use video

games to demonstrate the abilities of AI

where possible who's with me call

centers might be on our way out I want

to focus in on the generative AI side

because you caught a lot of Headlines by

the fact that you said look now

generative AI within our customer

services is doing the job of 700 people

is it still 700 is it more than that

it's probably slightly more by now I

think the reason we share we knew that

number would kind of catch the attention

but it's partially because I feel

there's a lot of Bzz around Ai and there

is you know a lot of things demos being

announced and you were talking about you

know Google today announcing something

but there are a few things that are

practically there and a lot of Business

Leaders are asking themselves like where

is the actual implications on my

business and this was the first time we

at least felt like look here's a

practical application that isn't like

just a small it's real life consumers

prefer it because it's actually higher

customer satisfaction than the human

agent it's faster to use it's higher

quality and it's having profound impact

why pay a human when AI can do it

cheaper and better and faster more

scalable and now it seems like consumers

prefer interacting of AI over humans and

Google is hitting back strong with their

own Suite of products their own AI

agents and you might have noticed in one

of the demos they use some sort of AI

glasses if you're wondering what those

glasses were this offers a hint Google

Executives hint that AR glasses are

poised for a comeback if you recall the

Google Glass that was 10 years ago back

in 2014 the product didn't really work a

lot of people were not too happy about

it stories are of people getting

attacked for wearing them in public you

know specifically in in San Francisco

the point is it didn't land well it it

didn't work at that time but now

executives are signaling that new AI

features could set the stage for AI

glasses to return so ARS augment the

reality so you're able to see certain

projections through the glasses that

appear in the real world some sort of an

overlay combined with something like an

AI assistant and this whole thing kind

of falls under the project Astra so

that's kind of the demo that we saw

where the ladies walking around the

office asking various questions while

pointing a camera at various objects and

the assistant responds seemingly very

quickly very intelligently and as you

may have noticed I mean some people are

a little bit cautious about trusting

Google with their new demos because

there's some against a while back

where seemingly one of the demos didn't

work as advertised it was stitched

together to make it seem like something

else it was of course the vision demo it

seemed like as you were drawing

something on a page the assistant

immediately recognized it and was able

to rapidly make suggestions later we

found out that it was edited and that's

not quite how it worked but based on the

feedback of people that actually went to

the developer conference it seems like

the latency how quickly those answers

are produced that is real so first of

all take a look at this quick demo that

Demi how he's posted about project Astra

what famous piece of literature does

this remind you

of that image of a skull reminds me of

Hamlet do you enjoy Shakespeare's work

yes I do um can you tell me what this

famous thought experiment might be and

explain it to me in one

sentence that looks like the ship of

Theus thought experiment it questions

whether an object that has had all of

its components replaced

remains fundamentally the same

object can you give me a creative poem

about these

objects a vibrant orb a mind laid bear

time silent passage fills the air the

fruit of thought the Sands that fall

what Memories will it

recall cool can you explain this diagram

to me in a couple sentences

this appears to be a diagram depicting

spectral lines from different

elements it is likely illustrating the

characteristic frequencies of light

emitted by those elements in a

spectroscopic

analysis cool thanks Gemini but the

results were mixed here Santiago saying

that there were some issues with the

Google demo take a

look okay so I was able to test project

Astra which is uh Google's answer to

open a I assistant now it was a little

bit weird and I only had about 2 minutes

to test um we went in this booth four

different people and again they gave us

only 2 minutes and they had a bunch of

objects there on a looking down camera

big screen uh somebody had to wear

headphones on a microphone to talk to

the assistant and we could use the

objects uh just to put them in front of

Astra and ask questions and tell it to

do stories and whatnot now I I want to

be fair because it was loud uh there was

a lot of noise and whatnot but the demo

did not go great uh the assistant had

trouble following instructions uh we had

to constantly say uh please go on please

complete the request sometimes uh the

guy in front of me asked the assistant

to to write a story about the objects

that that the assistant was looking at

and the assistant said uh yeah sure I

can do that and then it just stopped and

we had to say okay go on and we

constantly had to sort of like guide the

assistant to tell the story um I went up

to the screen and I drew something and

on the screen like it was a toou screen

I drew like a little sailboat on a

smiley face and it worked okay the

assistant was able to recognize the

sailboat and it was able to recognize

the smiling face but it felt choppy it

felt like it's not ready yet well and it

isn't this is just a demo is nobody else

can use it uh so we'll have to see when

it comes out how this is but my

impression my first impression again

only two minutes using this uh it was

not great now obviously I have in mind

the open AI presentation the

presentation they gave yesterday that it

went super smooth and it was it was

amazing to see that demo but it was a

demo as well so I don't want to make to

compare them because I didn't get to use

the open AI demo before using asra uh so

that's that's how I felt uh we'll have

to to

see what happens in the next few weeks

all right bye here's an image of

somebody who got Early Access to check

out project Astra here's kind of what

that looked like no there's something

through the Apple what do you think that

is worm yes is that all you're adding

yes that was it okay we're going to

delete this now that looks like it work

and uh yeah you guys

turn that was amaz so it definitely

looks like there's people that have

really good results with it some people

didn't necessarily like everything so I

think we'll have to make our decisions

when we get our hands on it but

certainly it looks like the latency is

good there's not too much delay and uh

we'll see when this comes out they also

announced having an email agent to

continuously organize all the receipts

in your inbox into a spreadsheet this is

the thing that I was kind of excited

about because it looks like you can

organize all your receipts and have the

AI constantly update your financials

right by having the intake of your email

or whatever else pulling that into

Google Sheets updating those so

basically instead of you going line by

line sell by sell and inputting those

this will automatically take care of

that for you which I got to say seems

very cool if we can do that accurately

effectively Etc another one was

returning an order say you have some

shoes that you didn't like or whatever

what if you wanted to return something

and you you didn't have time to take

care of yourself it looks like they took

a picture of that shoe and said return

the shoe and Gemini took care of

everything for you did the email

processed that whole thing and then just

told you what to expect without your

input without you babysitting at a

search agent with multi-step researching

and reasoning to find a pilates class

one person pointed this out and I can't

get this out of my head anymore because

on one hand we're approaching AGI this

technology that will fundamentally shift

the world but when we're demonstrating

how we're moving towards it what are the

use cases we're demonstrating by talking

about things like here's how to find a

yoga class here's how to plan a trip to

Miami you didn't like your pair of shoes

you poor baby you're too lazy to return

them well AGI can help and one other

very interesting thing is that they've

introduced a vo so vo is their answer to

opening eyes Sora the ability to create

various Tech to video animations and I

mean it looks pretty good certainly I

wouldn't say it's as good as what Sora

seems to be but it's certainly better

than or maybe on the level with Runway

ml Pika Labs Etc we'll find out for sure

again when we get our hands on it and we

actually get to play around with it

ourselves but once again we're seeing

that no one really truly has a mo no

matter what you release your competitors

will not be that far behind and a few

other really big and interesting

developments around the various llm

models that they're developing we have

Gemini 1.5 flash this is the fast

multimodal model with exceptional speed

and efficiency for quick high frequency

tasks very inexpensive 35 cents per

input so looks like 1 million tokens are

priced extremely well 35 cents for input

53 cents per output so that's for 1

million tokens they've also announced

the 2 million context window and have

said that their goal really is to have

an infinite context window they're

saying that this is yet another step

towards that infinite context window

right now however we do have the 2

million token context window in private

preview by the way a lot of this there's

a wait list for so definitely join if

you wanted to test some of the stuff out

I'm on the weight list for that new Sor

competitor vo I'm going to sign up for

this Gemini 1.5 Pro 2 million context

window right now and then there's this

additions to the Gemma family we have

Pali Gemma our first Vision language

open model that's optimized for image

captioning visual Q&A and other image

labeling tasks now of course the idea of

image labeling it's a big deal if you

think about where that technology is

going think about for example security

cameras that are you know placed all

over the world recording 24-hour footage

a lot of that is useless since it's not

easy to kind of parse that information

right you have to have somebody sitting

there looking at it or going over it but

it's expensive to extract insights from

that footage if you're able to have a

vision language model that is able to

caption those images accurately the use

cases are massive for example you can

have one over every patient in every

patient's room at the hospital

captioning maybe not every frame but

every few seconds right It Go goes in

and says the nurse came in at 10:15 and

administered this medicine can be used

as alarms for if people are you know

going to Cardiac Arrest or something

like that I mean there's obviously other

sensors for that but it could be an

extra layer of security if someone tries

to escape or anything that other sensors

wouldn't catch this would be able to in

near instant real time caption what's

happening and then another language

model for example kind of parse that

data and if something's happening that's

not supposed to happen it can alert the

staff or can log when the patient last

received the medicine all automated all

in real time as we've seen with open

eyes technology it's pretty good at

detecting people's emotional state based

on their faces a lot of businesses would

certainly enjoy being able to kind of

get a feel for what their customers like

and don't like right just automatically

keep track of it obviously could be a

little bit distopian right but you know

people are going to want to use this

again that's where this is going that's

not necessarily would poly Gemma is

capable of right now but we know that's

where it's progressing we've already

seen these companies publishing details

specifically focusing on on medicine on

the kind of the medical establishments

having something like this be able to

provide insight into what's happening at

hospitals for example would be a big big

deal also for example for announcements

for streaming Games sports Esports Etc

and then Gemma 2 which is going to be

looks like launches in June and is built

for industry leading performance at the

most useful devel ER sizes it

outperforms some models that are more

than twice the size and will run

efficiently on gpus or a single TPU so

that's the tensor Processing Unit the

Google's own sort of Chip infrastructure

in vertex AI again Google's sort of AI

Hub where a lot of the stuff is right so

polya can be opened in vertex AI if you

have that account but you have to have

access first also the briefly mentioned

Google gems which to me kind of sounds

like the custom gpts as in the GPT store

by open AI this ability to create these

little functional pieces that do one

specific task some custom thing that you

want them to do and more and more agents

that are coming soon that will be

dropping soon and getting integrated

with various Google products now keep in

mind notice that Google and pretty much

everybody else are talking about AI

agents for example here Gemini can

natively browse and use Chrome as a tool

looks like openi is partnering with

apple they'll be be on Apple phones and

the Mac OS Google of course has their

massive massive Android fan base a lot

of things run in the Android sort of

ecosystem people probably slightly

underestimate how much stuff Google and

the Android ecosystem has here's the

global smartphone sales share by

operating system Apple's presence of

course is very strong very loyal but

globally and especially at kind of like

the the cheaper prices the cheaper range

Android is massive here's another way to

display that so this is the mobile

operating system market share worldwide

as of 2022 Android is at around 71% iOS

is 27 and really not much else exists

also getting back to what I was saying

earlier about everyone moving into the

AI agent space it's important to

understand that these companies yeah

they'll produce a very useful agents for

the Everyday People they'll help them

with emails and this and that and that's

going to be excellent but the real big

Power of AI agents will be the skill set

you need the ability to build them for

yourself your own custom ones maybe ones

that run locally in your machine ones

that power your business in fact if

you've ever played factorio where you

build this little Factory and optimize

it to produce various things

meticulously track its efficiency and

outputs I wish I remembered who said

this first it wasn't

my idea but I really like it they said

that in the future business will be like

managing factorio only the various

little pieces of it will be your very

own AI autonomous agents earlier today I

believe this gentleman on the right is

Andrew a he's one of the top tier

machine learning researchers ex Google

he's up there amongst people like

Jeffrey Hinton Yan theun Etc he's

amongst the greats if you will and he

did a very Noble thing I think by

actually starting

deeplearning.ai and I think he also

partners with corera to build out free

training for various AI applications now

a lot of this is focused on developer

side but there's a lot that he does for

free and one of the recent things that

he introduced is this course again it's

free it's pretty short maybe 15 modules

and this gentleman this excited

Gentleman on the left he is the person

that built crew Ai and so they've

partnered to produce this multi- a agent

systems with crew AI again I'm not

affiliated with them this is not me

pitching you anything I'm going through

it right now I think it just came out

today if I understand correctly and I'll

kind of update everyone on how good I

thought it was I've tried crew AI I've

messed around with it it's good I

thought autogen was better for some

things but I will take another look at

it after going through this course by

the way for people that are building

stuff like this with this guys is doing

partnering with Andrew a and doing

something like this that's smart

whatever you're building if you're

teaching people how to use it on a

platform where it will be seen that's

definitely a smart idea it's a good idea

and if this is right up your alley if

this is something that you're interested

in I do want to encourage you to check

out my natural 20 private group where we

all learn to use AI learn to build these

AI autonomous AI agents we're a thousand

strong and growing rapidly and my goal

is to build this into the best online

community the most valuable and

informative Community let built around

how to wield AI for whatever purposes

you want so if that's something that

appeals to you join us won't you it

would be great to have you along but

whatever you decide to do thank you for

watching this video my name is Wes rth

and I'll see you next time