AI for business, explained in under forty minutes — the vocabulary, the actual capabilities, and the limits that get glossed over. Most business AI content is either a sales pitch or a technical deep-dive with nothing in between. This sits in the middle: what the terms and abbreviations actually mean so you can follow a conversation without nodding along, where AI came from and why it suddenly accelerated, a clear-eyed look at what it does well and what it still gets wrong, where the real risks sit, and how businesses have put it to work in practice. Short enough to watch before a meeting, specific enough to be useful in one.
💎 Want the full library? Premium AI courses on every major tool, all in one place. 👉 https://MyAiForThat.com
⏱️ Chapters: 0:00:00 Introduction and Key AI Terms Defined 0:09:18 Where Has AI Come From? 0:15:18 What AI Can and Cannot Do 0:20:37 The Limits of AI 0:26:00 Basic AI Functions in Business 0:29:24 Case Studies 0:34:15 Conclusion
No prerequisites — this assumes no technical background and no prior AI knowledge.
⚠️ On capabilities and limits: this is the fastest-moving part of the subject. Things listed here as limitations may already have been solved, and the case studies reflect the tools available when this was recorded. The framework for thinking about AI adoption holds; the specific capability list is a snapshot.
ℹ️ This course is independent and is not affiliated with, endorsed by, or sponsored by OpenAI or any company mentioned. For educational purposes only, and not business or investment advice.
Hello and welcome to the Expert Academy
and this course on what AI is and how it works.
Artificial intelligence represents
the cutting edge of emerging technologies.
New discoveries and upgrades continue
to appear as innovative businesses
push the limits of what's currently possible.
The launch of GPT-4 is an excellent example of innovation
and how it can completely change
the way ordinary people approach language, work and logic.
You're likely here because you want to both understand the theory of AI
and how to put it into action in a safe, reliable manner.
You might be a business leader
who wants to understand the full potential of AI
to remain ahead of the competition.
Or maybe you're a marketing analyst who wants to learn how AI
can unlock a new dimension of digital design for advertising campaigns.
Whatever the reason, this course will provide you
with the right knowledge,
terminology and real-life examples
to approach AI tools with confidence.
Through a series of six lectures,
you will learn to navigate AI-related terms and abbreviations,
analyze the history of AI and the rise of machine learning,
address current trends and
statistics related to artificial intelligence,
identify AI's current capacities and abilities,
realize the drawbacks and risks of AI,
and use case studies of businesses that have successfully
implemented AI. This course will benefit anyone with an interest in AI,
but it will be especially
valuable to business leaders looking to implement AI in strategic ways,
managers who want to
improve productivity and streamline internal processes,
software engineers hoping to leverage
AI to optimize workflows or improve product features,
compliance professionals who are
keen to understand the ethical and legal issues surrounding AI,
and business owners seeking
to learn more about AI's capabilities and advantages.
Over the coming lectures, we'll discuss the full spectrum of AI,
including terms you may have heard
in passing without receiving a full explanation.
We'll also build your appreciation for AI by tracking
its rapid expansion over the past 80 years.
By understanding how AI has developed over time,
what may seem like disjointed fragments
of information will soon come together to form a much clearer picture.
By the end of this course, you will have a full grasp
of the fundamentals of artificial intelligence
and how it can enhance your everyday operations.
After studying real-world case studies and statistics,
you'll understand how AI can be leveraged in ingenious ways
to keep your business ahead of the competition.
So, are you ready to invest in your future
and take a deep dive into the bustling world of AI?
Then let's get started by defining artificial intelligence
and other key terms you'll need to know
to be successful on your AI journey.
As a complex topic,
artificial intelligence can be difficult to describe in words.
New developments are constantly being made,
which results in even more concepts and terms.
Whilst it can be challenging in the beginning,
learning basic keywords provides you with
a strong foundation to build your personal knowledge base
and technological skills.
So don't worry if you don't initially understand some concepts.
Many will be discussed
and explained in greater detail later in the course.
Number 1.
Artificial Intelligence Artificial intelligence is a branch of computer
science where machines are built to carry out tasks
that could otherwise require human
intelligence. Examples include self-driving cars, chatbots, and robots.
Number two, machine learning.
Machine learning is a branch of AI that involves training a
computer on a dataset.
The computer often uses this dataset to make predictions and
improve with every new prediction.
Number three, deep learning.
Deep learning is a specific branch of machine learning.
Here the computer is given a very complex data set such
as sound waves or an image that
requires a self-made neural network to process.
Examples of deep learning include facial recognition software
or speech recognition in smart assistants.
Number four, neural networks.
Neural networks allow deep learning to function.
A network mimics the structure of the human brain
with interconnected nodes spread across
multiple layers.
This allows it to process data
that would otherwise be very hard to understand.
Number five, natural language processing.
Natural language processing is often shortened to NLP.
It's an application of deep learning
that allows a computer to understand, interpret
and generate human language based
on the patterns it identifies in data.
ChatGPT is an excellent example of natural language processing.
It can understand your request and provide a reply in return.
Number six, computer vision.
Computer vision is another application
of deep learning that involves a machine's ability
to interpret visual cues such as images and videos.
To give you a better appreciation of the interconnectedness of AI,
computer vision and natural language
processing can be used together.
For example, if you give a trained machine an image of text,
a computer vision model
can interpret each letter, word and phrase.
This information can then be passed
to an NLP model that understands the meaning of
the text.
Number seven, reinforcement learning.
Reinforcement learning is a branch of machine learning
that involves building a self-teaching
system that learns exclusively through punishment and reward.
AI models that can play chess are an
example of this concept.
By understanding the reward is winning the game, the computer can
teach itself to consistently maximize its chances of victory.
Number eight, supervised and unsupervised
learning. Machine learning can also be supervised and unsupervised.
Supervised learning involves
giving a computer labelled data, meaning the desired output is known.
Unsupervised learning has no labels for its data,
so it's up to the computer to decipher
patterns and relationships to form outputs.
For example, if you trained a computer
on 200 images of animals labelled as either a
fish, mammal or reptile,
you would expect it to identify a picture of a lion as a mammal.
That's supervised learning because you know the correct output.
However, if you trained a different computer on the same 200 images
but didn't label them,
the computer would have to make up its own categories.
It might split the images into fish,
mammals or reptiles, or it could completely make a
different list of groups.
Number nine, data mining.
Data mining involves giving massive data sets to a computer
and using AI to identify patterns
humans otherwise wouldn't notice.
This is often used in market analysis
to spot trends in the way customers shop.
Number ten, big data.
Big data is a collective term for datasets
that are too large for humans to comprehend.
The data is often growing at an increasing rate and unlabeled,
meaning only sophisticated
AI algorithms will be able to fully make sense of it.
Number 11.
Algorithms.
Algorithms are simply sets of rules
or steps a computer must follow to complete a task
or problem.
All AI solutions involve some sort of algorithm,
no matter how self-sufficient or unsupervised
they've become.
Number 12.
Automation is the process of replacing a human
with a computer in order to complete a task.
For example, using Grammarly to check the style
of your writing is an example of automation,
as you would otherwise have to proofread the work yourself.
Number 13, Robotics.
Robotics is an extremely sophisticated branch of science
that combines both artificial intelligence and engineering.
The end products are physical robots that complete real-world tasks,
often without external assistance.
And finally, number 14, AI ethics.
AI ethics considers the moral
and societal implications of AI and what can go wrong if
it's implemented incorrectly.
For example, an ongoing ethical debate
in AI surrounds the concept of bias and the potential
for an algorithm to discriminate against marginalized groups.
This demonstrates the importance of considering these elements
and how they will be controlled
before implementation.
To recap, artificial intelligence, or AI,
is a branch of computer science where machines
are coded to complete human tasks.
AI itself has multiple branches, one of which is machine learning.
The biggest subsection of machine learning is deep learning,
which uses neural networks
to function.
Deep learning allows us to interpret very complex datasets,
giving rise to natural language
processing and computer vision.
There are also different machine learning methods such
as unsupervised, supervised and
reinforcement.
All of these help to build powerful models
with applications in automation, prediction
and even robotics.
And that brings us to the end of our introductory lecture.
Your task now is to download and complete the AI terms worksheet.
In the next lecture we will discuss the history
of AI as well as emerging trends that will
give you an idea of where AI currently stands
and where it's going in the future.
See you
there.
Welcome to Lecture 2.
In our introductory lecture we covered
the basic terminology of Artificial Intelligence.
In this lecture we are going to delve deep into the past,
present and future of AI, starting
all the way back in the 1950s.
So let's get started.
There have been cultural references
and theories regarding an artificial mind for centuries,
But none had as much impact
as Alan Turing's 1950 Computing Machinery and Intelligence,
where he explained the framework behind building such technology.
However, at the time of writing,
computers were still extremely expensive.
Since they often cost over 200,000 US dollars,
they were reserved for large institutions
such as universities.
Despite Turing's ambitions,
he would have to wait a while before science caught up with
him.
In the same year as Turing's paper,
computer scientist John McCarthy hosted a conference
at Dartmouth University,
coining the term Artificial Intelligence during a discussion
about the future of computing.
Dartmouth University would remain the center point
for early advancements in this field,
establishing AI as an official discipline of study in 1956.
Moving into the 1960s, AI was rudimentary,
but revolutionary for its time.
Their AI models were examples
of rule-based expert systems taking basic inputs to make
a prediction.
For example, medicinal diagnostic machines
were used to assess patient symptoms.
Another great example of early AI pioneering came
from the ELIZA doctor program created by Joseph Weizenbaum.
This NLP system used pattern matching to mimic a psychotherapist,
often prompting the user through questions and queries.
This made it one of the first chatbots ever created.
The breakthroughs and momentum
of the 1960s failed to continue into the 1970s.
A combination of low funding
and unrealistic expectations triggered a plateau in new technologies.
This period would go on to be known as the AI winter.
As data became digital,
the focus shifted to machine learning techniques such as neural
networks and clustering algorithms.
The new predictive systems measured
their own performance through metrics such as accuracy
or mean squared error.
They could learn from their mistakes
to ultimately become highly reliable and powerful tools.
New Wave Machine Learning was developed in 1997
when IBM supercomputer Deep Blue defeated
chess grandmaster Garry Kasparov using sheer brute force computing
and alpha-beta pruning
methods.
The turn of the millennium saw AI begin
to be implemented in a more practical sense.
This is no surprise given the emergence of computer vision,
speech recognition and data
mining.
These techniques were all refined
and enhanced by deep learning algorithms.
Apple's Siri was invented in 2011,
while AlexNet's groundbreaking image classification
abilities were developed in 2012.
Google's DeepMind algorithm
was built using reinforcement learning and could soon beat
humans at Atari video games.
However, compared to the breakneck pace
of development over the last 10 years, AI's
historical growth has been comparatively slow.
Here are some of the biggest AI breakthroughs from the last decade.
In 2015, a team at UC Berkeley used reinforcement learning
to teach a robot fine motor skills,
allowing it to unscrew bottle caps
and remove a nail from a plank of wood.
In 2016, a Texan AI program learned
to diagnose breast cancer from a single mammogram with
99% accuracy.
In 2020, Baidu released LinearFold,
which allowed for the mass analysis of COVID-related
information.
It is believed this algorithm was instrumental
in speeding up the production of vaccines
for the general public. In 2021, Facebook produced SEER,
an unsupervised computer vision model that
could learn and understand unlabeled images of objects.
For example, it could understand what
a chair looks like from viewing
a single image without ever having seen a chair before. And
finally, OpenAI released GPT-4 in March 2023,
representing the most current AI capabilities.
This multimodal system can recreate audio, art,
images, and text based on very little input.
So we've seen how AI has grown.
So what does it look like at this very moment?
Well, according to a 2021 report by consulting firm Gartner,
AI has practical use in 73% of all current businesses.
Of all sectors, medicine is arguably the candidate
with the strongest dependence on AI.
After all, artificial intelligence in healthcare alone
is projected to be worth US$45.2 billion by 2026,
as calculated by Revenue Impact Advisors Markets and Markets.
In 2020, global AI investment had reached US$58.7 billion,
according to the International Data Corporation.
It is believed the top six AI stocks will
be worth a staggering US$30 trillion by 2030.
Given what AI can already do and the rate at which it is growing,
it's quite startling to think
of its capabilities in just a few years' time.
Soon, AI will be used even more widely in the education sector
to offer customized learning programs
so that individuals can understand content at a much faster pace.
It will also further simplify the process
of engineering digital and physical products.
As time goes on, more and more companies will embrace AI
to streamline their operations and increase their safety,
reliability, and efficiency.
And that brings us to the end of Lecture 2.
Now make sure to download and complete the History of AI worksheet.
In Lecture 3 we are going to give you an idea of what AI can
and can't accomplish, so that
you can understand the full scale of current AI capabilities.
See you
there.
Welcome to Lecture 3.
In our previous lecture we delved into the past,
present and future of AI.
Now we're going to look at what AI can and can't do,
with a supply chain case study to
back it up.
So let's get started.
With the growth artificial intelligence has
recently experienced, it wouldn't be outlandish
to assume that it can at least attempt any challenge.
However, whilst AI systems are often lightning fast in computation
and highly accurate in
the right circumstances, there are still many things AI cannot achieve.
These shortcomings may be possible
in the next 25 to 50 years, but it's important
to understand AI's limitations in the present
to have a realistic view of this technology.
So what can AI currently do?
Well, it can process large amounts of complex data,
both labeled and unlabeled, very quickly.
Perform repetitive tasks with high levels of accuracy and reliability.
Automate workflows for greater efficiency.
Make predictions and forecasts to promote data-driven decision making.
Translate written or verbal text.
Create text in response to a user input,
such as a question or statement.
Produce text-based information from a prompt.
make use of deep learning to recognize and categorize images,
videos and sound, utilize
reinforcement learning to make a sequence
of good decisions based on a perceived reward,
aid with medical diagnosis by analyzing scans,
symptoms and other recorded data, and facilitate
advanced engineering such as self-driving cars
and fine motor prosthetics.
The benefits of AI are often given more attention than
its limitations, but those drawbacks
are equally important.
AI cannot mimic a human consciousness
and have subjective experiences or self-generated opinions,
understand, recognize and interpret complex human behavior such
as emotions and social cues with
high accuracy, act as a complete replacement for human creativity,
make moral decisions that
consider both the consequence and the action itself,
completely eliminate bias that's inherent
in underlying data sets used to train AI models,
understand deep linguistic nuances such as
cultural references, sarcasm,
and context with total accuracy, carry out motor tasks
that require powerful physical dexterity and depth perception,
and finally, an AI model
can't learn new things outside the scope of its input data.
However, many of these concerns and restrictions aren't permanent.
For example, Google's AI chatbot Bard recently
began teaching itself Bengali with very little
prompting.
This indicates that AI models may eventually
be able to recognize their own limitations
and seek out new data sources to learn from.
One of the best ways to learn about what AI can accomplish
is by seeing how it functions
in a real-life setting.
AI has become especially valuable in fields
with repetitive tasks and relatively simple
workflows.
For example, logistics and supply chain mechanics rely on predictions
and efficiency, both of
which AI can help optimize.
An AI tool in a logistics company can use sales data
and seasonal consumer trends to
to predict how much stock is required in the near future.
Apply distance optimization algorithms
to work out which supply routes
are the quickest and most profitable.
Identify the closest warehouse or storage location
that has an item in stock
after a customer places an order.
Identify which items are most popular in certain regions
and anticipate problems with delivery and shipping vehicles
based on maintenance records.
So what about the current companies
that help with AI in the supply chain?
Number 1. Symbiotics.
Symbiotics produces robotics solutions such as sensors and handling
mechanisms to improve warehouse efficiency.
The use of robotics to find and move hard-to-reach
stock also reduces the number of workplace accidents too,
making their implementation
a win-win for the purchasing company.
Number 2. Coupa. Coupa is a multimodal financial AI tool
that allows you to track and analyze
your company's spend,
from procurement of materials to final delivery. The product also
automates accounting and protects against fraud,
ensuring you save money in a safe and secure
manner. And number three, Data Art.
Data Art is a customized machine learning and AI solution that
provides tailored technology based on the needs of the client.
Examples include architecture design,
recommendation systems, and intensive text mining.
Whilst your business will have its own unique
needs and tasks, there are some key factors
that should prompt you to consider an AI system.
If a task is repetitive and saps you of your time and money,
it's probably worth investigating whether you can automate the process.
Furthermore, if a task requires something only a human is capable of,
such as compassion or empathy,
it's best to leave those responsibilities to human teams.
And that brings us to the end of Lecture 3.
Your next task is to download
and complete the What AI Can and Can't Do worksheet.
In the next lecture, we'll discuss the limits of AI in greater detail.
See you there.
Welcome to lecture 4 of our course
on AI. In lecture 3 we looked
at AI's capacity
to understand what it can and can't do.
In this lecture we are going to delve
deeper into its limits
and see where AI can sometimes fall short in a business
context. So let's dive in.
Number one, creativity. Unless an AI model is
extremely complex and has the capacity
to teach itself new things, it will be
restricted by the data it's given.
The data you feed it at the very beginning
is what the AI will use to notice trends,
spot patterns and draw logical decisions.
The system utilizes these patterns to react
to new data inputs, but that often
leads to a lack of originality
or flexibility in function. For example, a
machine learning model
that spots aggressive cancer in medical scans won't
suddenly be able to spot other conditions.
Number two, ethical biases.
As we briefly mentioned before, AI is often used for
making predictions and projections.
However, high accuracy often results in unintentional bias
against certain groups,
especially if the underlying data is already biased. An excellent
example of this dates back to 2020,
when the UK government introduced an algorithm to predict the
grades of A-level students who weren't capable
of doing exams due to COVID lockdowns. The algorithm
considered a student's local area,
as well as the type of school they attended. This meant students
who studied at lower performing schools were given
a lower prediction despite being capable
of a much higher grade.
Number three, low domain expertise.
AI systems are rarely multi-purpose in nature,
meaning each new algorithm has a very specific
purpose.
Whether it's a facial recognition tool,
chatbot, or medical scanner, each model will
have its own boundaries.
When new data falls outside those boundaries,
the model often struggles.
For example, facial recognition software may not work as well
when presented with people
who have unique facial structures, shapes or physical attributes.
It is critical you account for all possibilities
when building your own algorithm to be as
reliable as possible.
Number 4.
Data limits.
AI is heavily dependent on the data that you give it,
so the information must be high quality,
well structured and without bias.
When possible, use high amounts of data
that are fully representative of the population
account for rare events.
For example, a facial recognition model built on a data
set that solely consists of white male faces
is going to struggle if presented
with a female face or faces from different ethnicities. Number five,
requirement for supervision.
Whilst a lot of AI is capable of operating by itself
to perform functions such as spotting patterns
and building categories from
unlabeled data, it still requires some form
of human supervision. This is
especially the case if your model has moral
and ethical implications. If your algorithm is
involved in making hiring decisions it would
be unethical to allow the AI system to continue on
indefinitely without checking for bias.
Many algorithms are constantly evolving models that
may develop undesirable behaviors as they encounter new data.
And number six, regulations and legal
requirements. AI is an evolving field so new laws
and regulations will continue to emerge.
Part of staying up to date on new AI technologies
and their applications
is maintaining compliance with legal standards.
These laws cover many tenets of AI ethics such as bias,
liability and security.
There are also laws that focus specifically on data and privacy
including the EU's General Data Protection Regulation.
It can be wise to consult specialist advisors
if you don't understand how
these laws will affect your business after implementing AI tools.
There are also limitations that impact AI
in a more general context that isn't necessarily
limited to business applications.
Given an AI model any basic logical task such as solving
a mathematical equation
or finding a hidden pattern in a set of images and it will likely
complete it with relative ease.
However day-to-day tasks that seem commonplace for humans such as
building an object or throwing a football may
be impossibly hard for an AI. Now this is because
physical dexterity, innovative problem solving
and kinesthetic challenges are skills humans
have built for millions of years.
Evolutionary traits have allowed us to dominate the animal kingdom,
whereas things like differential
calculus are acquired through targeted focus.
Whilst we can still understand abstract concepts,
we can't compete with computers that perform
calculations in seconds.
This disparity was notably displayed
in a robot built by Singaporean engineers.
It took 20 minutes to assemble an IKEA chair,
something that would have taken humans a fraction
of the time to do.
This is known as Moravec's paradox.
It shows that AI still has a long way to go
to match humans in terms of innovation and
all-round intelligence.
And that brings us to the end of Lecture 4.
Your task now is to download and complete the Limits of AI worksheet.
In the next lecture, we will discuss how you can leverage AI
for business applications.
See you
there.
Welcome to Lecture 5.
In Lecture 4 we looked at the limits of AI
and how to minimise its impact.
Now we are going to discuss how AI can
be used for your business to maximise profit
and increase productivity.
We've seen what AI can and can't do,
but how does this apply in a business context?
AI can help your business secure a competitive advantage by
automating repetitive tasks streamlining customer service through the
use of chatbots or virtual assistants identifying customer trends
and buying patterns
optimizing supply chain and logistics,
managing inventory and forecasting future needs,
detecting internal and external cybersecurity threats,
providing personalized improvement plans to employees,
providing smart recruitment capabilities
such as screening applications,
scheduling interviews and creating job descriptions,
analyzing social media accounts
for heightened brand awareness,
tracking customer sentiment and the customer experience,
conducting customer segmentation
for personalized email marketing campaigns,
and creating and maintaining virtual training programs.
You might even start to think of your own ways
to utilize AI for your own business
once you implement AI in other areas or departments.
More and more companies are beginning to see the everyday benefits
that artificial intelligence provides,
so it will soon become a critical element
of staying competitive in your industry.
The 2021 Global Human Capital Trends report by consulting firm Deloitte
confirmed that 53% of executives increased
their AI investments over the past year.
In a similar survey, Deloitte
also found that 63% of organizations achieved a competitive advantage
from the use of specialized AI technologies such
as natural language processing and machine learning models.
These findings highlight the fact that companies
are willing to invest in their future
by implementing AI tools and embracing cutting-edge technologies.
Even though natural language processing
and other technologies are still developing,
modern businesses are still eager to have
the most sophisticated models available.
The consulting firm Accenture reported in a study this year
that labor productivity could increase up
to 40% by 2035 as a result of implementing AI.
These findings are especially powerful
considering other shifts in the workforce,
such as the rise of remote and hybrid working arrangements.
They demonstrate that AI has the ability to drive productivity
and efficiency even during periods of social and economic change.
All in all, AI is more than a simple business trend.
It's also a powerful addition to many work-based processes
that improve productivity, enhance team-centric results,
and give you a competitive edge.
Resisting the AI wave may leave your business behind the competition,
so it's important you begin
to consider how you can implement it correctly as soon as possible.
We are now at the end of Lecture 5.
Your next task is to download and complete the Applications
of AI in Business worksheet.
In Lecture 6, we are going to look at four case studies of companies
implementing AI tools
and the lessons that can be learned from each. See you there.
Welcome to Lecture 6. In Lecture 5,
we delved into the potential applications of AI in a business
context. Now we're going to look at real-life case studies
of AI in global companies to show
how AI has been used in different sectors and industries.
By now you should be aware of what AI can and can't do
and understand many of the key technical terms surrounding it.
However, you may still be unsure about how exactly AI
is used in the real world.
But let's take a look at some examples of companies that integrated AI
into their internal processes and how it worked out for the better.
Number one, Wildtrak. Wildtrak is a conservation company that looks to
protect endangered species, such as the cheetah,
by monitoring their movements and behaviours
as a species.
Their crowdsourced computer vision-based footprint
identification technology gives Wildtrak information
about an animal, such as its age, sex, direction and species.
A single footprint may not give a lot away,
but in large quantities Wildtrak hopes to
give animals a fighting chance against poachers,
hunters and extinction.
The takeaway from Wildtrak's story is that crowdsourced data,
which is data that's provided
by the general public,
can really help a business develop powerful solutions at a faster pace.
Number two, Epic.
Epic provides AI solutions for law firms,
improving their workflows and ensuring projects
focus on the right things.
By utilizing cutting-edge AI platforms such as NextLP
and Brainspace, Epic can reduce
document review time and eliminate data irrelevant to the case at hand.
Each new case is extra training data for the software,
meaning companies partnered with
Epic can further tailor its solution to their needs
and field of expertise with ever-increasing
accuracy.
The main takeaway from Epic is to consider relying
on external solution providers.
There's no need to try and build an algorithm yourself.
In fact, the cost may be prohibitive
for many organizations depending on the size, scale
and complexity of the AI tool.
According to a 2023 article
by the medical records management company Scribe Technology Solutions,
it can cost up to 500,000 US dollars to develop a custom AI system.
Pre-built solutions by comparison might only cost 40,000 US dollars,
but the model won't be tailored to your exact needs.
Number three, fintech OS.
Banking companies are often built on legacy systems,
which are inflexible but take years to replace.
FintechOS provides a fluid solution to this problem,
providing a platform that supports
innovation and speed without affecting
the reliability of the original system.
Solutions like FintechOS prove that even
the most complex systems can be upgraded and
brought into alignment with modern technologies.
Even highly regulated industries like finance can benefit
from artificial intelligence.
It may require additional implementation planning
and closer regulatory oversight, but the outcome
will be a leaner, more reliable platform.
And number four, 20th Century Studios.
The movies and visual entertainment business is a colossal industry,
worth around 94 billion
US dollars in 2022, with that figure only set to rise.
It's no wonder then that multinational studios
are leveraging AI to maximize profits.
That's exactly what 20th Century Studios has done.
The company built a machine learning model
that utilizes computer vision for movie trailers
natural language processing to analyze plot synopsis.
It can also analyze
ticketing data. When these techniques
are combined the model is able to get a
better understanding of different audience segments and which movies
people will find most appealing.
This deep learning program known as Merlin
Flow has been used to curate trailers
for upcoming films based on the content
that target audiences have enjoyed before.
It may also be used to determine
which scripts are worth investing in.
If none of the large audience segments
likely to be interested in a movie then it may not be worth making.
The main takeaway from this
is the importance of AI in marketing.
Deep learning models can be used to understand your
business's target audience
and see exactly what content your customers will respond to. An AI
model can learn from adverts designed by your competitors
or even through hyper personalized
marketing campaigns. That brings us to the end of lecture six.
Congratulations on making it to the
the end of this course on artificial intelligence and how it works.
It's time now to download and
complete the case studies worksheet.
In our concluding lecture we'll give a brief overview
of everything we've covered in the course.
See you there.
Congratulations on completing this course
on artificial intelligence and its potential applications.
In this concluding lecture we will
be reviewing everything you learned in previous parts of the course.
By now we hope you understand
what AI is, including the different branches
that make up the ever-changing field of computer
science.
Now that you recognise the capabilities and limitations of AI,
you'll be better equipped
to implement it in an effective and ethical way.
We started off the course by looking
at all the technical terms surrounding AI.
You learnt about key topics such as unsupervised learning,
reinforcement learning, computer
vision and natural language processing.
In Lecture 2 we reviewed the history of AI,
giving context to how artificial intelligence
has become such an important
and revolutionary addition to both technology and industry.
Then in Lecture 3 we discussed what AI can and can't do.
This included a segment on how AI can solve repetitive data-based tasks
while often struggling
with emotional challenges that require social nuance and context.
This transitioned nicely into Lecture 4 where we delved deeper
into AI's limits and how
you can work around these inherent limitations.
In Lecture 5 we looked into the ways that you
can implement AI correctly in business
to maximize its potential and streamline your
internal operations. We
also evaluated where AI needs improvement such as handling fine motor
skills or tasks that require physical dexterity.
Finally in Lecture 6 we looked at different case
studies of how AI has been used by real businesses.
We saw how computer vision and crowdsourcing
helped wildlife to aid endangered species
and protect animals from hunters.
We highlighted the importance of AI in marketing to drive profits
and better understand your
target audience through data analytics.
We also bolstered the idea that you don't need
to rely on your own coding abilities
to build AI solutions for your business.
There are plenty of consultancy companies,
SaaS solutions, and bespoke platforms designed
to overcome the problems your organization is facing.
We also assess the costs of these options
and why one may be more appropriate for your
business than another.
After completing this course,
we are confident you will have a rich understanding of artificial
intelligence.
From its layered history,
to its vast growth in recent years, to its multiple branches
and fields of expertise,
you should now be aware of the importance of AI in modern life.
You should also understand how AI can be utilized
in a valuable manner to streamline workflows,
automate mundane tasks, and improve business revenue.
By now, you're undoubtedly aware
of AI's current capabilities
and what it may be able to accomplish in the future.
You know how to appreciate its moral and legal implications
and recognize the steps you can take to monitor them.
You should also know exactly how to integrate AI
in a business context and draw lessons
from real-life case studies to leverage AI
in the most beneficial way possible.
We have a whole catalog of courses available,
So be sure to check out the categories
and topics relevant to you and your organization.
Best wishes from the Expert Academy and good luck.