OpenAI's STUNNING "GPT-based agents" for Businesses | Custom Models for Industries | AI Flywheels
Summary
TLDROpenAI's shift towards its original mission of being the AI base layer is gaining momentum. The company is now focusing on fine-tuning its models for specific applications, enabling developers to build custom models that cater to various industries. This approach is set to revolutionize sectors like healthcare, law, and agriculture, by significantly reducing costs and improving efficiency. OpenAI's strategic partnerships and advancements in AI technology are poised to create a new wave of innovation and opportunities for businesses across the globe.
Takeaways
- 🌟 OpenAI's shift back to its original goal of being the AI base layer, allowing various industries to build on top of it for specific functionalities.
- 🚀 Introduction of improvements to the fine-tuning API and expansion of custom models program, giving developers more control and new ways to build domain-specific models.
- 🔍 Extension of model knowledge with techniques like retrieval augmented generation, enabling AI to search the internet and summarize information relevant to queries.
- 🎯 Custom models fine-tuned for specific tasks, such as customer service or medical applications, leading to higher quality results and reduced costs.
- 🌐 Partnerships with startups and companies outside of OpenAI to bring innovative features and applications based on the AI base layer.
- 🔧 New features for developers, including model checkpoints and a side-by-side playground UI for comparing model quality and performance.
- 🔗 Integration with third-party platforms like Weights & Biases for model tracking and evaluation, enhancing visibility into the model development process.
- 📈 Transition from traditional business operations to AI-driven processes, with a prediction that AI will significantly replace or improve a vast majority of tasks over the next decade.
- 💡 Showcase of real-world applications where custom-trained models have been successfully implemented, demonstrating the potential impact of AI across various sectors.
- 🌍 Empowerment of different communities, such as farmers and students, through AI solutions tailored to their specific needs and challenges.
- 💼 Opportunities for businesses and individuals to leverage AI for process automation, enhanced customer experiences, and the creation of new markets and services.
Q & A
What is the original mission of OpenAI according to the transcript?
-The original mission of OpenAI is to be the AI base layer, providing foundational artificial intelligence like GPT, Sora, and others, on top of which various companies and apps can build specific functionalities for different fields such as science, chat applications, medical AI, etc.
How does OpenAI plan to move forward with its base layer strategy?
-OpenAI plans to move forward by offering improvements to the fine-tuning API and expanding their custom models program, allowing developers more control over fine-tuning and new ways to build custom models with OpenAI. They aim to provide a foundation for others to build upon, leveraging their advanced technologies like Sora, Whisper, and data analytics capabilities.
What is fine-tuning in the context of AI models?
-Fine-tuning is the process of adjusting a general AI model, like GPT 4, to perform better at specific tasks. For example, a model can be fine-tuned to excel in customer service or coding by training it with domain-specific data and objectives, thus creating a custom model tailored to those particular needs.
How can fine-tuning a model benefit businesses?
-Fine-tuning a model can lead to higher quality results, reduced costs, and improved latency. By focusing the model on the specific needs of a business, it can operate more efficiently, potentially reducing the computational resources needed and providing more accurate and personalized outputs for tasks such as customer support or data analysis.
What is an example of a company leveraging OpenAI's fine-tuning capabilities?
-Indeed, a global job matching and hiring platform, used fine-tuned GPT 3.5 Turbo to generate higher quality and more accurate recommendations for job seekers. By fine-tuning, they were able to cut their token costs by 80% and scale up their recommendations from less than 1 million to 20 million per month.
What new features is OpenAI introducing to aid developers?
-OpenAI is introducing features such as model checkpoints to reduce retraining needs, a side-by-side playground UI for comparing model quality and performance, and integration with third-party platforms like Weights & Biases for better model tracking and evaluation.
How does the transcript suggest the future of AI implementation in various industries?
-The transcript suggests that AI implementation will become ubiquitous, with custom models tailored to specific industries and use cases. It predicts a future where the majority of organizations will develop customized models personalized to their business, leading to widespread automation and efficiency improvements across all sectors.
What is the significance of the partnership between OpenAI and various startups and companies?
-The partnership signifies OpenAI's commitment to fostering innovation and development across different sectors. By working with startups and companies, OpenAI can help bring new features and applications to life, leveraging the base layer AI capabilities they have built to meet diverse needs and drive progress in areas like legal, health insurance, education, and more.
How does the transcript view the role of AI in the future?
-The transcript views AI as a transformative force set to revolutionize various aspects of life and business. It highlights AI's potential to automate tasks, improve efficiency, reduce costs, and create new opportunities for innovation and growth. The transcript suggests that AI will become an integral part of our daily lives and operations, leading to significant advancements and improvements across numerous fields.
What is the potential impact of AI on job roles and tasks?
-The potential impact of AI on job roles and tasks, as outlined in the transcript, includes the automation of various administrative and repetitive tasks, allowing professionals to focus on higher-order tasks that require more critical thinking and human interaction. This shift could lead to increased productivity, reduced burnout, and the creation of new roles centered around AI management and optimization.
Outlines
🚀 Open AI's Strategic Shift and Base Layer AI
The paragraph discusses Open AI's return to its original mission of being the AI base layer, allowing various companies and apps to build on top of it. It mentions how Open AI initially encouraged others to build on its platform, but later developed its own applications, overshadowing startups that were filling gaps in its technology. Now, Open AI seems to be inviting startups again to build on its base layer, with improved functionalities like Sora, Whisper, advanced data analytics, and in-painting. The speaker shares news about Open AI's recent API improvements and its custom models program, aiming to give developers more control over fine-tuning and building custom models.
🎯 Fine-Tuning and Custom Models for Enhanced Performance
This paragraph delves into the concept of fine-tuning AI models for specific tasks, such as customer service or scientific research. It explains how fine-tuning can lead to higher quality results while reducing costs and latency. An example is given of how a global job matching platform fine-tuned GPT 3.5 to generate better recommendations, significantly reducing token costs. The speaker also discusses the effectiveness of smaller, fine-tuned models for specific tasks, predicting a trend of more such models being used in the future. Open AI is facilitating this by introducing new features for developers, such as model checkpoints and a side-by-side playground UI for model comparison.
🤖 Integration with Third-Party Platforms and AI Use Cases
The paragraph talks about Open AI's efforts to integrate with third-party platforms, starting with Weights and Biases, to track model versions and experiments. It also covers how Open AI is working with companies to create custom-trained models for specific domains. The speaker then transitions to discussing the broader impact of AI, predicting that in the next decade, AI will be applied to a vast majority of tasks, replacing or enhancing traditional methods. The paragraph highlights the potential of AI in various sectors, from legal to health insurance, and emphasizes the lucrative opportunities for developers and businesses in this space.
🌐 Custom AI Solutions Transforming Industries
This paragraph presents specific customer stories that illustrate the transformative power of custom AI solutions across different industries. It discusses how AI is being used to manage email inboxes, assist legal professionals with case law, streamline health insurance processes, and support weight loss applications. The speaker emphasizes the increased engagement and productivity brought about by these AI solutions, as well as the potential for AI agents to handle more complex tasks in the future. The paragraph also touches on the importance of training staff to understand the limitations and capabilities of AI.
🌱 AI Empowerment in Agriculture and Education
The paragraph highlights the use of AI in empowering farmers and improving education. It describes how AI is helping farmers in India and Kenya by providing knowledge on better farming practices and market information, significantly increasing their income. In the education sector, AI is making education data more accessible. The speaker also discusses the potential of AI in creating new industries, such as AI agents that could order food or book gym classes on behalf of users. The paragraph concludes with a call for excitement about the potential of AI to bring about positive change and opportunities in various sectors.
Mindmap
Keywords
💡Open AI
💡Base Layer AI
💡Fine-Tuning
💡Custom Models
💡AI Startups
💡RAG Retrieval
💡AI Integration
💡AI Consultants
💡Industry-Specific AI
💡AI Automation
Highlights
SE Alman and the team behind OpenAI have been working on a new approach to AI, potentially signaling a shift back to their original goal of being the AI base layer.
The concept of the AI base layer involves using AI like GPT, Sora, and others as foundational tools upon which various companies and applications can be built.
OpenAI's strategy includes allowing developers to fine-tune their models for specific tasks, which can lead to improved performance, reduced latency, and cost-efficiency.
OpenAI has introduced improvements to the fine-tuning API and expanded their custom models program, giving developers more control and new ways to build custom models.
The company has partnered with several startups and established companies to bring innovative features and applications based on their AI technologies.
Fine-tuning can be particularly effective for specific tasks, such as customer service or scientific research, leading to high-quality results at a lower cost and faster speeds.
OpenAI's developments suggest a future where most organizations will develop customized models tailored to their industry, business, or use case.
The company has published examples of successful custom model implementations across various industries, including legal, health insurance, and education.
AI technology is projected to automate or significantly improve a vast range of tasks, moving from near 0% to potentially 100% market penetration over the next decade.
The potential applications of AI are vast and varied, from improving email management to assisting with legal case law and enhancing educational data accessibility.
OpenAI's advancements are expected to create lucrative opportunities for developers and businesses that can successfully integrate AI into their operations.
AI's role in various sectors, such as agriculture and healthcare, is highlighted by the ability to increase efficiency and reduce costs, making it a valuable tool for societal improvement.
The AI landscape, despite being overhyped, is also underestimated in terms of its potential impact and transformative power.
Custom AI models are being developed to handle complex tasks with greater accuracy and nuance, leading to better outcomes in fields like law and medicine.
The future of AI includes the development of autonomous agents that could manage various aspects of our lives, from scheduling to personal assistance.
OpenAI's initiatives reflect a shift towards making AI more accessible and customizable, paving the way for widespread adoption and integration into various industries and tasks.
The potential for AI to revolutionize industries and create new markets is evident in the diverse range of applications and custom models being developed.
AI's impact on global challenges, such as agricultural practices and income disparities, demonstrates its potential to empower and improve conditions on a large scale.
The narrative of AI as a transformative tool is reinforced by the real-world applications and success stories showcased by OpenAI's developments.
Transcripts
SE Alman and the team behind open AI
have been kind of quiet for a while a
calm before the storm if you will but in
the last few days we've been getting
some glimpses into what potentially is
coming next let's take a look at what
they've been cooking up interestingly
enough this seems to be a nod back to
their original goal their original state
mission of being the AI base layer
here's I think is a great illustration
of what they meant by that so imagine
this sort of big block here that's layer
one that's the base layer AI That's the
GPT that's the Sora that's the do and on
top of that we build the various
companies various apps various software
that takes advantage of this base layer
of this AI to build specific
functionality for example for science
for chat applications for medical AI etc
etc now when Chad BT exploded onto the
scene opening eye did say that this was
kind of the plan for the future they
were going to be the base layer and
people could build on top of it tons of
people in apps did build on top of it
but then opening eye was like I am death
and just killed them all by building
their own sort of versions of it their
own applications now that's probably not
100% Fair because a lot of these
startups they were trying to build
technology to fill in the gaps that
opening ey was missing for example
things like whisper things that were
similar to code interpreter or Advanced
data analytics as it's called in other
words a lot of these startups they were
building kind of these obvious apps that
openi itself probably had in the works
and they kind of got rolled over by open
AI by the base layer if you will but it
seems that now the game is changed now
open AI seemingly now that it has all
the pieces in place it has the Sora The
Whisper the voice engine it has advanced
data analytics it has do with in
painting and editing and all that stuff
like it's built all of the different
pieces of the base layer their full
functionality you can talk back and
forth to Chad PT it's able to understand
your speech it's able to talk back to
you and now you know the base layer a
lot of it is complete all the companies
that try to plug missing pieces of the
Bas layer are dead and now seemingly
Open the Eyes saying okay guys we've got
the foundation we've got the base layer
build stuff on top of it now that's my
opinion but let me show you a few key
pieces of news that came out just in the
last few days that I think support the
Viewpoint as you look at this ask
yourself this question is open AI
rolling out the red carpet for the layer
to startups to build on top of it on top
of open ai's uh base layer let's get
started so this drop today introducing
improvements to the fine-tuning API and
expanding our custom models program
developers get more control over
fine-tuning and new ways to build custom
models with open AI So It Begins there
are a variety of techniques that
Developers can use to increase model
performance in an effort to reduce
latency improve accuracy reduce costs
with it's extending model knowledge with
rag retrieval augmented generation so
basically having that model you know as
example would be to search the internet
for information like perplexity it
returns to you the various links it
finds and then kind of summarize them or
if you've ever done the chat with a PDF
thing right the PDF is your sort of
database extra data extra info that you
have in there so the model instead of
kind of making up an answer it retrieves
information from that PDF and then
answers with that information in mind
you know various custom models
fine-tuning custom train models with new
domain specific knowledge and they're
launching new features to give
developers more control over fine-tuning
and they're introducing more ways to
work with our team of AI experts and
researchers to build custom models and
they're not kidding as you'll see in a
second they've partnered with quite a
number of startups of different
companies that are outside of openai to
bring some pretty interesting features
we'll cover that right after this so
they start by talking about some
fine-tuning of API features or rather
fine-tuning of these models using an API
for those that are not familiar fine
tuning is basically if you think of the
GPT 4 model as this kind of Big Blob
that can do anything but maybe it
doesn't do everything well right what we
can do is let's say we wanted to do
exclusively coding or customer service
with it we can fine-tune this model to
do customer service I'm going to say
support because that's easier to Rite we
can take this big huge model and find
tuna to be great at support a customer
service to answer questions that's
specific to our company our business
train how to answer questions correctly
in the style that we want now of course
there's going to be tons of other areas
where it's going to get worse because
this support model is going to be almost
a brand new model right it's a
fine-tuned GPT 4 it's a custom model for
support and chances are it might get a
lot worse in various other things like
it might not be that good at coding
anymore it might get bad at poetry but
it gets good at support so again there
sort of like little uh demonstration
here this little representation here
right the base layer right that's let's
say GPT 4 right we take that we we pull
it out and we find tun it to be a
science gp4 or a medical gp4 all the
stuff that gp4 already has comes with it
right so it understands language it can
reason right it has certain knowledge it
has certain skills we're just taking it
and shaping it into this specific task
and the other really big thing about
fine-tuning is you can achieve higher
quality results while reducing cost and
latency since we're pulling out just
what we need for our customer support
agent for example there's tons of things
that are still in this big GPT model
that we don't need that we kind of leave
behind so this model could be cheaper to
run much faster Etc as an example they
give indeed a global job matching and
hiring platform wanted to send
personalized recommendations to job
Seekers highlighting relevant jobs based
on skills experience Etc and they were
able to fine-tune GPT 3.5 turbo to
generate higher quality and more
accurate recommendations they were able
to cut their token costs by 80% so was
able to improve cost and latency how
fast it was doing it cutting the number
of tokens like the words that were run
through the model by 80% this allowed
him to scale from less than 1 million
recommendations to job Seekers per month
to roughly 20 million by the way as a
quick aside so on this Channel we cover
a lot of research from companies like
apple Microsoft meta openi Google deep
mine Etc and I'm beginning to notice a
pattern here that usually we see the
research papers come out you know let's
say 3 to 6 months before sort of the
things that are talked about in those
papers start hitting the market the Orca
2 paper from Microsoft research kind of
showed that teaching small language
models how to reason that this was
extremely effective or could two models
match or surpass all other mod models
including models 5 to 10 times larger so
smaller fine-tune models built for
specific tasks are incredibly effective
incredibly fast and Incredibly cheap
we've covered uh Apple's Research into
Vision models you could say we covered
this a few days ago so they had this
model that was 80 million parameters
that was in some cases beating GPT 4
getting close in other cases but 80
million parameters is a tiny tiny model
GPT 4 is G gantu clocking in at we think
1.7 trillion or you know that's what
people are guessing it's probably
somewhere in the low trillions so here's
a model that is if I did my math
correctly this is like a thousandth of a
percent the size of GPT 4 that is
getting close to it uh for performance
for certain specific tasks not for
everything obviously but for a specific
task this microscopic model is as good
as GPT 4 and so if I had to guess I
would guess that we're going to be
seeing a lot more things like this where
small fast models gbt 3.5 turbo and the
like get fine-tuned to do pretty
incredible things in specific domains
for pennies for for very cheap and
they'll do it very very fast and the
people that will be building this for
businesses those people will be printing
money that's my guess and open AI is
Building Services to make that easier
now they're going to be working with
some companies on their own so it's
going to be an open AI plus this is the
company partnership but I know a
percentage of you listening right now
are builders in the space are developers
I mean these B2B Solutions if you're
looking to make money seem like a very
very lucrative uh thing to approach and
so opening eyes is launching new
features to give developers even more
control over their fine-tuning jobs by
the way I've said this before but just
let me reiterate because I I don't think
I've said it recently for the next let's
say 5 years to a decade one place where
people are going to make a lot of money
and one place where we're going to see a
lot of progress is this idea of shoving
llms into everything in in other words
for any digital task that that's running
there's probably some solution where you
take a custom fine-tuned large language
model or maybe some custom small model
something like Orca 2 which Microsoft by
the way was kind enough to open source
so you're able to see exactly how they
built it and then apply to something
like this think about how much indeed a
global job matching firm how much were
they spending on these customized
personalizations how important it is to
their business model it probably has a
very high importance and probably had a
high cost right so if you're able to
build this custom model that cuts down
how much it costs them to do and how
fast it's done by
80% how many millions are you saving
them per year certainly you can probably
charge a pretty penny for doing
something like that and I would guess
the the talent pool for people that know
how to do this isn't massive like it's
not like I doubt that we have an
overwhelming amount of people that are
highly skilled in something like this
and in just a second I'll show you
examples of how this is applied but
here's the new features that openi is
rolling out so it's adding a uh sort of
model checkpoints so during each
training Epoch to reduce the need for
you know various retraining Etc so kind
of like a Version Control side by side
playground UI for comparing model
quality and performance apparently
somebody made two alms fight each other
in uh Street Fighter too so they trained
them to Output you know what it should
do Fireball kick move closer Etc to see
which which LM is the best Street
Fighter 2 player so that's kind of what
I'm imagining here I doubt it's that
cool and it also supports integration
with thirdparty platforms starting with
weights and biases so that's this
platform where you're able to track
experiments evaluate model performance
Etc where you have the model registry
the added openi so openi uses weights
and bias models to track all their model
versions across 2,000 plus projects
millions of experiments and hundreds of
team members so you can have visibility
into model development process with just
a few lines of code you know you're able
to answer questions like what exact
version of the data set was this model
trained on now I'm going to skip some of
the parts I will leave a link so if
you're really interested in this if
you're in the space I highly encourage
to read this obviously but for the rest
of us I think it might be more
interesting to quickly go over the kind
of the big points and then let's dive
into exactly what this looks like what
are the actual final outputs of this
open
as a Bas layer thing that's happening so
first of all they're talking about
something called assistant fine-tuning
so this is where openi helps train
models for a specific domain right in
partnership with a dedicated group of
openai researchers since then we've met
with dozens of customers to assess their
custom model needs and evolve their
program to further maximize performance
and they've published this so we're
going to take a look at that next but
before we look into those specific
examples here's kind of the end of this
block post where opening ey saying well
here's what's next for model
customizations they're saying that we
believe that in the future the vast
majority of organizations will develop
customized models that are personalized
to their industry business or use case I
think this is very important especially
if you're you know if you're in this for
the money now not everybody listening to
this channel is approaching this from
kind of a business Centric perspective
some of us are just interested in the
research and kind of personal use of AI
how can we take advantage of it for work
for our careers so I understand that
this might not apply to everybody but I
still think this is very important to
understand because this gives you a
glimpse into what's going to happen over
the next decade so here's 0% and here's
100% this is a line right these are all
the tasks that can be either automated
or significantly improved or maybe made
Cheaper by applying AI right so from
taking either a biological neural net
like a human being right an employee or
even taking you know code like a lot of
stuff runs on code right on something
that a smart software developer sat down
and typed out on their keyboard and now
that code runs to complete a certain
task like in accounting or stocks or I
mean nowadays pretty much everything
right now not that long ago close to
zero of those business operations ran on
neural Nets so neural Nets that's the AI
that we're talking about that's chbt
that's Sora that's all the AI music all
the AI images that you're seeing a lot
of Google stuff is machine learning you
know how they run their ad serving
platform content recommendations stuff
like that but you know you go back a few
years and maybe the number of those
things were like 1% right that could be
automated neural Nets or improved the
neural Nets in some way that was like 1%
since chbt came out and all that stuff
started coming out in the last few years
maybe we're at like I don't know 1 and a
half% 2% I don't no this might be even
like a horrible overestimation now that
I think about it it's probably some
fraction of 1% over the next decade plus
all of those processes will be replaced
by AI by neural Nets some of them will
be a one model to rule them all for
example some people would just use Chad
GPT for whatever task they need right
they would just use the base layer kind
of like that non-custom non- fine-tuned
version of GPT 4 to ask questions one
interesting thing about perplexity for
example that's having incredible success
and progress It's not really doing that
much on top of the sort of the base
layer so it's taking all the language
models that already exist right so you
have your GPT 4 Turbo clot 3 Sonet Cloud
3 Opus mistal large they have a few that
you can play around in the um playground
with and then you have kind of their own
fast ones that they've kind of
fine-tuned but for the pro users they're
just using like the base layer right
whether that's claw 3 or GPT 4 with you
know heaps of software on top of it
right so they're taking gp4 they're
building a little perplexity little
search engine on top of it and they're
adding heaps of software to you know
search the internet pull back answers so
literally in this image you know let's
say this is GPT 4 this is you know
perplexity and the heaps of software is
the features of that uh perplexity
search engine and they're valued at
billions of dollars Bezos is investing
in it tons of other people are investing
in it and rumor is Apple is in talks of
potentially buying them so that's one
example where literally you know
billions of dollars are created just
with this but these large language
models or other things like Vision
models and speech models and image
generation models they're coming for
everything they're going to be in your
thermostat they're going to be in the
food photography right some sort of
enhancement you know this is magnific AI
or magnific AI right that upscaler i' I
featured it on the channel some people
use it to upscale their food photography
right look at that thing you can see
every little grain of pesto every pine
nut I'm guessing this is cracked black
pepper I mean that looks delicious I
mean take a look at this burger like
they're going to use this in product
photography the left is sort of the unup
scaled the right is the upscaled I mean
most product images will use this in one
way or another we've covered Google
shopping AI with their new thing they're
rolling out where you're basically able
to use AI to create product visuals to
put them into different environments to
create virtual sort of those models
right that kind of showcase your product
it's able to translate everything into
different languages to you can basically
have an international store selling to
the entire world in their own native
language you don't have to do product
photography right you can pull out your
phone if you have a prototype take a
picture of it and the AI will create 3D
images and variations and put them in
the hands of models and just do all of
that for you as you'll see in a second
uh opening eyes partnering up with
people in the legal field to create
custom train models for them another
company that does health insurance
another company that does education jet
brains for coding email assistance
Salesforce health and weight loss
productivity clinical trials creativity
for Farmers to increase their income
personalized fitness and health coaching
content creation like I can sit here all
day and just named use cases but I think
this kind of chart I hope represents a
little bit better right we're at close
to zero right now of uh penetration of
this technology in the market that's
going to change everything it's going to
reduce costs improve automations create
new use cases that we haven't even
thought of right everybody's going to
want it everybody's going to need it and
over the next you know however many
years 5 10 20 it's going to get rolled
out from almost zero to you know
eventually approaching 100 and for every
model that gets rolled out people will
pay money and this money it won't be an
issue it won't be hard to sell these
Services because they're going to be
saving money money they're going to be
making things faster it's it's hot it's
sexy everybody wants it from your local
mom and pop store that's local that's
brick and mortar that's selling I don't
know quilts I've been doing e-commerce
for the last 10 years and I'm telling
you there's like off the top of my head
I can think of a number of places where
if I could get a custom trained finetune
model to just do that task for me it
would probably be able to do it faster
better cheaper than I could from
inventory management to customer service
to a lot of aspects of marketing so what
I'm saying is this may be a good
opportunity for those of you who like
money so opening I continues saying with
a variety of techniques available to
build a custom model organizations of
all sizes and this is important this is
kind of like the big thing too of all
sizes because with Enterprise level
software like software that you sell to
hospitals and government organizations
right you know it was hard to build you
know you had to have a massive team to
even approach projects like that right
you know there's know such thing as like
Artisan software right you know how you
can go to a farmers market and buy some
Artisan loaves of bread that's really
delicious I mean with software you
wanted a certain amount of scale because
you had to build that software so the
bigger of a customer size that it served
the better the economics of that
software would be right right it's a
little bit different here you know we
covered this blog post from semi
analysis a while back so semi analysis
is this I mean it's a website that talks
about all the news and events and and
does analysis of semiconductor and AI
Industries so sort of the the hardware
the chips behind AI like all that stuff
Nvidia and tsmc and all of that right
and they dropped this little line that I
thought was fascinating they're saying
we're training a CNN so CNN is a
convolutional neural network to
accelerate the frequent satellite
imaging of data centers to expand our
tracking to every data center across
every country right so this is kind of
like a a a publisher right so they have
a website a news newletter they sell
their reports and they have this like
really specific use case right they have
access to satellite images right I'm
sure a lot of that stuff is is public
right they track 1100 data centers and
their deployments with publicly
available information including but not
limited to property records power usage
this is Freedom of Information Act
requests and satellite images right so
they have these very specific use cases
where it would help them to have a
neural net a CNN in this case so it's
not a large language model it's not a
Transformer based architecture a
convolutional neural network in this
case it kind of like labels and uh is
strained on on images but they're
building their own custom AI model to
you know help them analyze images of
these data centers we looking at the
satellite images and figure out like
what's happening there they might do the
same thing for these other things for
property records power usage right all
this stuff how many other people in the
world might be interested in doing
something like this to analyze satellite
images for this specific AI accelerator
data centers probably not that many
maybe maybe a handful the point being
that there's going to be a great demand
for these custom small AI models for
very specific use cases for
organizations of all sizes they can
develop personalized models to realize
more meaningful specific impact from
their AI implementations the key is to
clearly scope the use case design and
Implement evaluation systems choose the
right techniques and be prepared to
iterate over time for the model to reach
Optimal Performance again the people
doing this will be making lots of money
with opene most organizations can see
meaningful results quickly with the
self- serve fine-tuning API for any
organization that needs to more deeply
fine-tune their models or imbue new
domain specific knowledge into the model
our custom model programs can help so
one this is where they have their own
in-house AI people two is where open AI
helps them develop it probably for
bigger companies with specific use cases
they can afford to pay for that but I
think there's another obvious sort of
category and that's people they're
outside of a corporation right like a
third party that just goes to businesses
and says hey I can create this for you
and charge them a certain amount of
money a consultant or whatever AI
consultant automation consultant
whatever you want to call that and again
we're at close to 0% now and it's
heading to 100 and so the early movers
will probably have a massive Advantage
but you might be saying okay but like
what are some of the use cases that make
sense is all this just hype is all this
just like uh marketing hype people are
getting over excited for Nothing by the
way Demi saabi said an interesting thing
but here's AI breakfast saying in the
words of Demi saabi the AI landscape is
overhyped yes it is overhyped but
underestimated it's funny how both can
be true but it makes sense there's a lot
of hype but people aren't fully grasping
the massiveness of everything that's
happening here because what we're
talking about is just a tiny sliver of
it like these fine-tune models for
specific business cases that's just a
tiny tiny part of it but let's see what
these use cases are so these are
customer stories from open eii so this
is kind of what they''re been talking
about where they work together with
companies to create custom fine-tuned
models for their specific Solutions
here's one that kind of jumped out of me
so this is superhuman that's the company
name and they're introducing a new era
of email with open AI they're building a
suite of Next Generation AI email
products that are saving users time
driving value and increasing engagement
now we've all struggled with the
overflowing inbox right there's too many
things too many appointments to too many
meeting minutes to keep uh up on and so
superhuman is reming how that could be
using AI to help you write emails to
rewrite certain ideas in your voice or
for example just say it out loud kind of
dictate it and have the AI kind of fill
the gaps write it in your voice make
sure everything's correct and then send
it out of summarize a quick summary of
each email that's arriving instant reply
allows you to reply from certain
contextual options similar to how a a
text message you you have those little
options to reply now it's important to
understand that yes so GPT Chad GPT can
do a lot of this but this we're talking
about custom Solutions so that's
somebody sitting there and specifically
trying to make this as good as possible
for the specific use cases but okay so
maybe this isn't as uh sexy as I made it
sound out to be but here's what caught
my attention they're saying here okay so
what's next so superh humanist company
imagines the world where GPT based
agents will soon be able to filter Tre G
and respond to email automatically
scheduling meetings and appointments and
taking basic actions online and in the
real world so again I mean they're
talking about building a email agent or
a scheduling agent so kind of like an
executive assistant for you that has
access to your emails to your calendar
and you can kind of like manage and stay
on top of it based on what you tell it
to do so the idea of opening up an email
inbox and then reading you know the
first email and the second email the
third email or you know doing triage
kind of thinking okay like what can I
get away with not answering today let me
see what what's the most important
things like the idea of you doing that
yourself with your brain and your eyes
and you clicking the mouse or whatever
like in 5 years that's going to be
laughable that's going to be on par with
manually adjusting the the heat in your
house every 10 minutes instead of you
know using a thermostat how many people
have that now close to zero right maybe
some fraction of 1% of the people in the
world are using that in their inbox
right it it's not 1% it's less for sure
if you look at the entire world even in
the United States it's less in 5 years
it'll be 100% or or close right
somewhere getting closer to 100% next we
have Harvey so it's a custom train model
for legal professionals now of course
lawyers and Technology the combination
of the two is hilarious right the I am
not a cat your honor video is perhaps
the funniest video I've ever seen in my
life I think the first time I I heard
myself laughing I'm not even
exaggerating then of course you have the
lawyers that use Chad BT to I don't know
what they were trying to do but it
basically made up a bunch of court cases
to cite their arguments and it got them
into a lot of trouble now at the time a
lot of the clueless reporters that that
cover AI wrote articles saying you know
the conclusion was AI bad the reality
was they shouldn't have used gpc4 kind
of that that that base model right they
needed something that was custom trained
that had some architecture that had some
rag that retrieval augmented generation
right so as you can see here this is the
model that uh was trained so it's custom
trained and as you can see here it's
giving you little citations of the case
law so for everything that it writes it
supports it by giving you a little okay
so I'm referencing this specific uh case
law right stone versus writer or
whatever right to support its claims
it's not guessing it's not hallucinating
it's not taking it the best shot at it
it's using ACT ual databases to support
those claims it's custom trained not to
you know take a stab at it or whatever
it's custom train to answer just as
accurately as possible so the company
that's doing it they're called Harvey
and they've worked with 10 of the
largest law firms to test this model and
they were surprised by how strong the
reaction was from those law firms 97% of
the time the lawyers prefer the output
from the case law model from this custom
trained model because it was longer more
complete answer went into the Nuance of
what the question was asking and covered
more relevant case law hallucination
reduction was one of Harvey's
motivations for building a custom model
so again the the problem wasn't quote
unquote AI the problem was it just
didn't have the correct architecture
needed to provide the correct answers
and the people using it didn't have the
correct training to understand what it
was good at where it could fail again
that's another part of this whole thing
is training the staff the people using
it to kind of understand its limitations
Etc right having Chad gbt represent you
in a legal case is shockingly not a good
idea and so where are they seeing this
whole thing going next well they're
saying this don't build for the current
capabilities of these models today CU
remember when you do that what happens
opening ey comes in it's like I am death
and it just like reaps everybody right
build for where the models are going to
be tackle more complex versions of
problems so that when better versions of
the model come out they aren't solved as
a side effect and what's Harvey working
on next let's see one of their main
focuses is Agents again surprise
surprise AI autonomous agents are once
again the next Frontier here's yet
another company that's like well that's
the very next step it's in this case how
to combine multiple model calls together
in a single working output this would
simplify the user experience and reduce
the amount of prompt engineering and
typing users needed to do another one is
Oscar bringing AI to health insurance
the open AI models together with Oscar
help ensure hippoc compliance doubling
productivity automated documentation and
claims processing medical care workers
cut their time spent you know
documenting various Medical Care
conversations stuff like that by 40% so
that's your doctor that's your nurse
almost half of the time instead of
thinking like what your needs are
they're sitting there filling out this
paperwork draining their mental energy
their physical energy think about how
many millions of dollars we spend having
them do that instead of something like a
large language models taking care of it
for you know pennies this reduces
burnout and allows nurses and clinicians
to focus on higher order tasks how many
lawyers and doctors and nurses use
various tools like this to help them in
their work right now probably close to
zero right we're we're definitely not
near 1% right it's some tiny fraction of
1% how many of them will be using it in
5 to 10 years probably a lot closer to
100% right next they're also talking
about you know claims improving accuracy
automating 408,000 tickets by the end of
the year all right so this is what a
claim looks like right you click this
little magic button help me with this
inquiry and it fills it out I assume
next they're saying creating an AI
flywheel a flywheel in in business
usually refers to kind of like a virtual
cycle where something almost kind of
like Builds on itself to keep improving
that business or that habit or whatever
the more users a social network has the
more value there is for existing users
and the more users it will attract and
so this is kind of like a direct Network
effect they're saying we don't want to
just nibble around the edges of
administrative use case simplification
right because this is what we're talking
about right is just simplifying
administrative use cases right important
thing very valuable thing will save a
lot of money will save the mental
energies of doctors and nurses that need
to focus on their patients so it's it's
it's it's important it's critical but
they're saying that's just the beginning
they're looking to bring down the cost
of seeing Physicians and being in the
hospital by a factor of 10 in the next 3
to 5 years another customer was Zelma
who's using GPT 4 to make Education data
accessible another customer success
story is healthify an app that helps
with weight loss now this is for a
specific population it looks like
they're really focusing on uh India it
seems um and specifically classifying
traditional Indian foods so focusing on
certain custom populations can be a
really good idea so here they're saying
snap achieved around 80% accuracy for
single IND Indian foods right so here's
where it gets a little bit interesting
if you're looking at it from a a
business perspective like how important
is it for businesses to have something
like this well increased engagement
because of the finetune models the AI
the engagement increases people are more
interested people tend to use it more
users track 50% more often with these
new models users engage more for
nutrition and fitness coaching clients
engage with AI supported coaches 18%
more and and this is kind of interesting
with users permission the agents will
even be able to order food or book Gym
classes a brand new industry that's
going to emerge that we we haven't yet
seen but a few people are beginning to
mention it in the future when we have
agents that control our lives that
sounds harsh that help us run our lives
they're kind of like our assistants but
they do connect us to information to
services to everything they're going to
recommend certain products right they're
going to order food for us book Gym
classes Etc so now a lot of marketing
dollars are spent on marketing to the
person right Billboards ads it's focused
on you right it's focused on trying to
get you the human to buy something
slowly over time we're going to see more
and more money going into marketing to
AIS to these agents what is that going
to mean well I mean the simplest way is
spending ad dollars to get the company
Behind These agents to recommend your
products over the competitor that's the
most obvious one but there might be
things like with search engine
optimization right so it's how you
optimize your web page the links that
are pointing to your web page that
determines how high you appear in Google
search could there be an SEO but for
autonomous AI agents an a if you will
autonomous AI agents optimization it'll
be interesting to see how that unfolds
but you know companies will be spending
money trying to figure out how to get
autonomous agents to recommend them over
their comp competition like that's going
to happen 100% chance of that and gen
isn't just for us soft first worlders
only digital green uses open ey to
increase farmer income in countries
including India and Kenya it's important
for these Farmers to be able to teach
each other best practice for growing new
crops share various local weather
conditions and there's a lot of problems
with that right not just connecting them
but also having a lot of different
languages that are being spoken right
they need to connect with their
suppliers and provide market and pricing
information so it's like Eve online but
for Farmers farmer to Farmer training
videos increase farmer income by an
average of 24% this was interesting so
we I mentioned this in yesterday's video
so there's some things that governments
and various well-funded government like
entities can do that may not be
necessarily something that you know
large corporations are are willing to do
where they can use their resources to
build something that's useful for the
whole country right so for example a
database for all the countries and that
cultural AI to be trained on right
Microsoft probably won't do that Google
won't do that but maybe something like
uh DARPA might consider it uh I don't
know if that's exactly in their
wheelhouse or not but certainly it's
something that would help everybody in
that country that's in AI to be able to
do something like this here India's
Ministry of Agriculture validates old
documents in the knowledge based to
ensure accuracy and reliability and then
with GPT 4 they're using rag retrieval
augmented generation to pull the needed
stuff from there various crop research
fact sheets Etc and the cost of this
farmer chat these extension services
they went from $35 per farmer to 35
cents per farmer again massive massive
leap and again right now this technolog
is available to close you know to 0% of
the world's population of farmers with
those needs right and over the next
decade hopefully it'll start approaching
100% And there's like 20 of these little
um customer profiles each one has its
own mindblowing application like there's
not one use case here that isn't kind of
revolutionary in its own way where it
provides more more features more use
cases it it drops the price of doing
something it frees up the humans the
experts to focus on what they're
supposed to be doing instead of being
bed down in the details where this
Medical Group believed they would need
specialized medical models to get good
results and they were shocked to find
that GPT 4 outperformed a team of highly
trained human experts they said it not I
they were shocked I'm sensing that the
title of this video is going to have the
word shocking in it what do you think am
I am I am I correct is my intuition
accurate so I'll link this page down
below so this is the open.com so this is
not their blog which I think a lot of
people are familiar with this is/
custom- stories and I think it just got
populated with all of this stuff there's
a whole bunch of them here and most of
them aren't like fluffy they're not just
whatever nonsense they're impactful and
they showcase how powerful these custom
train models can be so I'll leave it off
there I think I've made my point back
when software was rolling out that
created a lot of wealth a lot of change
in the world software was eating the
world they said and now this is the next
step where AI is doing that thing we
probably shouldn't say AI is eating the
world we should say AI is helping the
world I think that's the a16z motto AI
is going to save the world and I agree
for a lot of people that are scared of
AI I don't know becoming the Terminator
and turning us into paper clipse or
whatever that scenario is I hope they
take a look at this it has the ability
to make our lives easier to eliminate
busy work to help people understand
various medical procedures better to
deal with email to to understand legal
cases better to make Education data more
accessible so more people understand
what decisions are being made in our
school system systems right how people
in farming communities they're
struggling to a little bit better this
is where AI is right now this is what it
can do right now and over the next 5
years 10 years it's going to be rolling
out from zero to 100% of you know
whatever use cases we can apply it to
the world will change some people will
make a lot of money and it's going to be
a heck of a wild ride so I don't know
about you but I'm pretty excited about
what's coming I'm excited about the
empowerment that AI is going to bring to
education to business to creating music
to creating movies with things like
sunno Ai and soraa is going to empower
people to do more and we'll have some
bumps in the road I'm sure but wherever
you are in life I would say get excited
if you're in a position to become this
AI automation consultant and help people
incorporate neural networks into various
parts of their business to create these
fine-tuned custom models for them I
think that's going to be a millionaire
making industry probably for some people
if you're running a business or or want
to think about where things like that
could help you in the business where
they could help you do more or do more
with less and if you're wondering why I
keep showing you this wolf from p and
boots I honestly don't know it's just a
great character he just cracks me up
this is the guy that plays them by way
if you weren't aware anyways my name is
Wes Roth and thank you for watching
Browse More Related Video
รู้จัก Multimodal AI ทำงานได้กับข้อมูลหลายประเภท | EP.56 - #TechByTrueDigitalPodcast
Foundation Models | Satya Nadella at Microsoft Build 2024
Microsoft BOMBSHELL Announcements: Sam Altman on GPT-5, Devin Joins Microsoft and Phi-3 (SUPERCUT)
Using Generative AI to Tackle Global Challenges
Nvidia's Breakthrough AI Chip Defies Physics
The AI Hype is OVER! Have LLMs Peaked?
5.0 / 5 (0 votes)