AI News: The AI Arms Race is Getting INSANE
Summary
TLDRGoogle 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
😀 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.
🤖 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.
🕶️ 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.
📈 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.
🛠️ 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.
🎓 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.
Mindmap
Keywords
💡AGI
💡OpenAI
💡Ilia Sutskever
💡Jacob Patok
💡Anthropic AI
💡AI agents
💡Project Astra
💡AI and customer service
💡AI in gaming
💡Brain-computer interface
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
Просмотреть больше связанных видео
![](https://i.ytimg.com/vi/3u8HndJlA0c/hq720.jpg)
Will "Claude Investor" DOMINATE the Future of Investment Research?" - AI Agent Proliferation Begins
![](https://i.ytimg.com/vi/c2It5EvBfXA/hq720.jpg)
GitHub's Devin Competitor, Sam Altman Talks GPT-5 and AGI, Amazon Q, Rabbit R1 Hacked (AI News)
![](https://i.ytimg.com/vi/yinHx5UnYs0/hq720.jpg)
Open sourcing the AI ecosystem ft. Arthur Mensch of Mistral AI and Matt Miller
![](https://i.ytimg.com/vi/ID5tc61iksY/hq720.jpg)
ChatGPT Can Now Talk Like a Human [Latest Updates]
![](https://i.ytimg.com/vi/32OsnJqrPB4/hq720.jpg)
BATALHA de INTELIGÊNCIA ARTIFICIAL! - Gemini | ChatGPT-4o
![](https://i.ytimg.com/vi/kx9yRgnb3rk/hqdefault.jpg?sqp=-oaymwEXCJADEOABSFryq4qpAwkIARUAAIhCGAE=&rs=AOn4CLCPc8xg2S1UNf-xE8tybI8F7sxlaA)
GOOGLE entra a gamba tesa per spazzare via ChatGPT (ma inciampa su una demo Fake)
5.0 / 5 (0 votes)