SlideShare a Scribd company logo
1 of 42
Download to read offline
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
Intelligent Systems Lab
The Presence of our Future
HOW AI TRANSFORMS BUSINESSES
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
INTRODUCTION
❑ Lovely tech buzz words and grand promises:
➢ "The Internet Of Things will connect everything"
➢ "Blockchain will democratize everything", and
➢ "AI will solve all of our problems"
❑ AI can potentially find a better way of solving old problems or solve some unsolved problems
❑ To hit the target with AI tech, we first must understand what this technology can do
2
❑ For a machine to be able to determine that an image contains a cat, many thousands of
tagged images had to be processed using Machine Learning
➢ This "self-learning" process is fascinating, useful, and has multiple business use cases
➢ But is it intelligence?
❑ For comparison, a child needs to experience two interactions with a cat to be able to identify
a cat the third time they meet
➢ This is a significant difference in capabilities, and it highlights that how people and machines think, learn,
and understand is very different
❑ Definitely, for the present, but also the foreseeable future, the business opportunities of
leveraging what we call "AI" is more IA (Intelligent Automation) than AI (Artificial Intelligence)
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
AI is a One-Way street
The AI market size is projecting an increase from
20 Billion USD in 2020 to more than 70 Billion
USD by 2024 and 260 Billion USD by 2028.
The market
Companies store a large amount of (proprietary)
data concerning their customers, products,
services and operations.
Data
Data utilization constitutes the only way to enable
company development and survival in the highly
competitive global financial landscape.
Utilization
3
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
HUGE AMOUNT OF DATA
❑ We are creating more than 2.5 Quintillion (i.e. Million Billion) bytes of data every year
➢ 90% of all data ever created, was created in the past 2 years
➢ From now on, the amount of data in the world will double every two years
❑ Data comes from sensors, devices, video/audio, transactional applications, networks, log files,
web, social media, etc.
➢ Data Never Sleeps
❑ Big Data (Volume, Variety, Velocity, Veracity)
“Mining needle in a haystack.
So much hay and so little time”
⮚ Traditional computing techniques are not able to handle such
large datasets
⮚ Adopted solution: Artificial Intelligence
4
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
WHAT IS AI
❑ Artificial Intelligence (AI) is a wide-ranging branch of computer science concerned
with building smart machines capable of performing tasks that typically require
human intelligence
➢ Researchers (and inventors) have long dreamed of creating machines that think
➢ Turing test
❑ Is an interdisciplinary science with multiple approaches, affecting every sector of the
tech industry
❑ Uses statistical methods but is completely a different approach compared to
statistics
5
⮚ Not magic or science fiction - but rather science, engineering, and mathematics
⮚ However, applications… are magic
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
TYPES OF AI
❑ Weak AI or Narrow AI:
➢ Is a simulation of human intelligence
➢ Machines act as if they were intelligent
❑ Strong AI or Artificial General Intelligence (AGI):
➢ Machines are actually thinking (not just simulating thinking)
➢ Machines, much like a human being, can apply that intelligence to solve any problem
➢ Is the kind of Artificial Intelligence we see in the movies
6
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
STRONG AI
❑ Strong or General Artificial Intelligence is very different, and is the type of
adaptable intellect found in humans, a flexible form of intelligence capable of:
➢ Actual thinking - Consciousness
➢ Learning how to carry out vastly different tasks
✓ anything from haircutting to building spreadsheets, or
➢ To reason about a wide variety of topics based on its accumulated experience
7
❑ Is it feasible? Is it desirable?
Interview of humanoid Sophia
In October 2017, Sophia
became the first robot to
receive citizenship
❑ Machines have neither past experiences (if they do, they are limited to narrow fields such as
identifying a pattern of a cat, driving in a lane, or picking an object), and they most certainly do not
have feelings
❑ Missing these components, how can machines think like us?
❑ What is becoming clear is that the holy grail of AI should not be to teach machines to think like
humans, but to teach machines to think like machines
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
WEAK AI
❑ Is what we see all around us in computers today
➢ Intelligent systems that have been taught how to carry out specific tasks without
being explicitly programmed how to do so
❑ This type of machine intelligence is evident in
➢ Medicine diagnosis
➢ Fraud detection
➢ Recommendation systems that suggest products (e.g. Netflix's recommendations)
➢ NLP and speech recognition
✓ e.g. Cortana by Microsoft, Siri virtual assistant and Amazon's Alexa
➢ Vision recognition systems on self-driving cars (and mobiles)
➢ Robotics (Atlas by Boston Dynamics)
➢ Self-driving cars
➢ Game playing (playing chess)
✓ IBM’s DEEP BLUE became the first computer program to defeat the world champion Garry Kasparov in a chess match (1997)
➢ Machine Translation (Google and Microsoft translate)
➢ …etc …etc
❑ These are just a few examples of AI systems that exist today
8
Boston Dynamics' Atlas has now mastered parkour
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
AI AND MACHINE LEARNING
➢ This capability is Machine Learning
9
❑ AI is a new phenomenon, but it is not a new concept
➢ It came into existence in 1956, but it took decades of work to make significant progress toward
developing an AI system and making it a technological reality
❑ Many AI projects have been developed in the past with hard-coding the knowledge of a
problem’s world in formal languages. None has led to major success.
➢ Rule or Knowledge based approach
➢ “frozen software” — software that can only be improved via updates
❑ These difficulties led to AI systems with the ability to obtain
their own knowledge, by extracting patterns from raw data
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
WHAT IS MACHINE LEARNING
❑ Machine learning is the subfield of computer science that gives computers the ability to
learn without being explicitly programmed
Arthur Samuel, 1959
10
❑ Machine learning feeds data into a computer and uses
statistical techniques to help it "learn" how to get
progressively better at a task, without having been specifically
programmed for that task, eliminating the need for millions of
lines of written code.
⮚ Data=Actionable Information
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
MACHINE LEARNING CATEGORIES
11
❑ Supervised learning (using labeled data sets)
➢ Output: Prediction function (model) for Classification or Regression
➢ Applications: Forecasting (e.g. Demand), Predictive maintenance, etc
➢ Many algorithms: Support Vector Machines, Linear regression, KNN,
Decision trees (Random Forest), Naive Bayes, Neural Networks, etc
❑ Unsupervised learning (using unlabeled data sets)
➢ Output: Clusters, Association Rules
➢ Applications: Clustering (e.g. customer segmentation), Market
basket analysis, etc
➢ Algorithms: k-means, Dbscan, Hierarchical, Apriori
❑ Reinforcement learning that allows an agent to learn a
behaviour through trial-and-error interactions with a dynamic
environment
➢ Output: Intelligent agent
➢ Applications: self-driving cars, industry automation, trading, gaming,
robotics, etc
➢ Algorithms: Q-learning, SARSA, Proximal Policy Optimization (PPO),
DQN, Deep Recurrent Q-Networks, …etc
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
ARTIFICIAL NEURAL NETWORKS (ANNS)
❑ They have their origins back to 1943 (artificial neuron as a computational
model)
❑ Their name and structure are inspired by the human brain, mimicking the
way that biological neurons signal to one another
❑ They are comprised of nodes (neuron) layers, containing an input layer,
one or more hidden layers, and an output layer
❑ A neuron (or Perceptron) is a mathematical function that just take inputs
on their “dendrites” and generate output on their “axon branches”
❑ Different architectures: Number of neurons, of layers, activation functions,
dropout rates and optimization techniques are preferable for different
tasks
12
-----------------------------------------
Our brain contains about 86 billion neurons and more than a 100 trillion (or according to some estimates 1000 trillion)
synapses (connections)
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
DEEP LEARNING
❑ Probably one of the hottest topics in Machine Learning today
➢ Achieves remarkable outcomes
❑ Uses large multi-layered Neural Networks
13
❑ Different Kinds
➢ Deep Feedforward Networks or Multilayer Perceptrons (MLPs)
➢ CNN (Convolutional NN) [1995]
✓ TCN (Temporal Convolutional Networks) [2016]
➢ RNN (Recurrent NN) [1986]
✓ LSTMs (Long Short-Term Memory Units) [1997]
✓ Gated RNN (GRU) [2014]
✓ Encoder and Decoder through an attention mechanism [2015]
➢ GAN (Generative Adversarial Networks) [2014] – Generative AI
➢ Transformer [2017]
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
TRANSFORMER NEURAL NETWORK
❑ Like most neural networks, transformer models are basically large encoder/decoder
blocks that process data
➢ Small but strategic additions to these blocks make transformers uniquely powerful
❑ Transformer models apply an evolving set of mathematical techniques, called
attention or self-attention, to detect subtle ways even distant data elements
➢ RNNs favors more recent words at the end of a sentence while earlier words fade away in
volatile neural activations
➢ Attention (and Multi-Head Attention) gives all words equal access to any part of a
sentence in a faster parallel scheme and no longer suffers the wait time of serial
processing
❑ They are in many cases replacing convolutional (CNNs) and recurrent neural networks
(RNNs), the most popular types of deep learning models just five years ago
➢ Today, many AI engineers are working on trillion-parameter transformers
❑ This architecture is now used not only in natural language processing and computer vision,
but also in audio and multi-modal processing
➢ It has also led to the development of pre-trained Large Language Models, such as
Generative Pre-trained Transformers (GPTs) and Bidirectional Encoder Representations
from Transformers (BERT)
14
➢ https://towardsdatascience.com/transformer-neural-network-step-by-step-breakdown-of-the-beast-b3e096dc857f
➢ 15 of the best large language models
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
DEEP LEARNING: APPLICATIONS
15
❑ NLP (e.g. ChatGPT)
❑ Speech recognition (e.g. Voice based search, Siri,
Amazon echo, Amazon Alexa, Microsoft Cortana)
❑ Image recognition (e.g. Tagging faces in photos,
smartphones)
❑ Pattern detection (e.g. Handwriting recognition)
❑ Forecasting (Retail, stocks, ….)
➢ For all the Weak AI applications, also
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
DEEP LEARNING IN THE FOREGROUND
❑ Why DNN (and Deep Learning) is now in the foreground? Because…
❑ We now have fast enough computers (GPUs), software, libraries and huge amount
of data to train large NNs
➢ Programming Languages
✔ Java, Python, R, SCALA, Julia, Spark
➢ Libraries-Frameworks
✔ Keras (Tensorflow, Theano) developed by Google, are python libraries that make writing
deep learning models easy, and give the option of training them on a GPU
✔ PyTorch developed by Facebook
⮚ New architectures/models, like Transformers, used by BERT and GPT3 on which chatGPT
has been based
➢ AI assistant and Developer Tools
✓ Copilot (Microsoft)
⮚ AI companies (like Hugging Face) that offer open source tools and models
16
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
DEEP LEARNING: PROS AND CONS
17
❑ Cons
➢ Neural networks are black boxes
✓ We cannot know how much each independent variable is influencing the dependent variables
➢ Hardware Requirements
✓ It is computationally very expensive and time consuming to train with traditional CPUs (→ GPUs)
➢ Depend a lot on training data
❑ Pros
➢ Powerful tool
➢ Can be used for all categories of ML
➢ Are good to model with nonlinear data with large
number of inputs; for example, images
➢ Once trained, the predictions are pretty fast
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
GENERATIVE AI
❑ The branch of AI, which uses available text, audio files, images, videos to
create a whole new set of the same which seems to be true and perfect in
its own senses
➢ Generation mainly of images, 3D objects, Text (Articles, Blog posts, Product
descriptions), or even code
❑ Gartner predicts that by 2025, the percentage of data generated by
generative AI will amount to 10% of all generated data
18
✓ https://machinelearningmastery.com/what-are-generative-adversarial-networks-gans/
✓ Top 100+ Generative AI Applications / Use Cases in 2023
❑ Generative AI has two famous models for working:
➢ The Generative Adversarial Networks (GANs) combined with AutoRegressive Convolutional Neural
Networks(AR-CNN), and
➢ Transformer-based Models
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
19
CONCLUDING ML
❑ Machine learning is a powerful Artificial Intelligence tool that enables us to crunch petabytes of
data and make sense of a complicated world
➢ …it's transforming a wide variety of industries/organizations
➢ …It's solving previously unsolved problems
❑ AI systems learn and amplify human biases
➢ keep humans in the loop
➢ “People can still play a role – either when validating a fraud or following up
with an action (e.g. a rejected transaction/account)”
❑ Because finding patterns is hard, often not enough training data is available, and also because of the
high expectations it often fails to deliver
❑ Even with machine learning in place, business leaders remain skeptical of error rates
➢ This is a cognitive bias; people magnify the failures of machine learning
➢ Look at the media firestorm on the few self-driving car mishaps vs. the thousands to millions of people dying
due to other causes
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
HOW AI TRANSFORMS BUSINESSES
INDICATIVE APPLICATIONS
20
➢ User Experience (UX)
➢ Chat bots (NLP, Large Language Models)
➢ Retail Applications of AI
➢ AI and Supply Chain
➢ AI in Banking
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
AI IN BIG COMPANIES
❑ Microsoft
➢ Azure
✓ It provides SaaS, PaaS and IaaS
✓ Supports many programming languages, tools and frameworks, Microsoft and third-party
➢ Azure ML
✓ A cloud service for accelerating and managing machine learning projects
➢ Microsoft Cognitive Toolkit (CNTK) (is no longer actively developed)
✓ An open-source toolkit for commercial-grade distributed deep learning
❑ Oracle AI
➢ Oracle Cloud Infrastructure (OCI) AI/Data Science Services
➢ A fully-managed platform for teams of data scientists to build, train, deploy, and manage machine
learning models using Python and open source tools
❑ IBM, Watson and watsonx
➢ IBM's enterprise-ready AI and data platform designed to leverage foundation models and machine
learning
❑ Amazon – AWS Deep Learning AMIs (Amazon Machine Images)
➢ Provides ML practitioners and researchers with a curated and secure set of frameworks, dependencies, and
tools to accelerate deep learning on Amazon
21
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
USER EXPERIENCE (UX)
❑ UX refers to the feeling users experience when using a
product, application, system, or service
❑ It is a broad term that can cover anything from
➢ how well the user can navigate the product
➢ how easy it is to use
➢ how relevant the content displayed is
➢ etc.
22
❑ AI and machine learning (ML) are rapidly changing the way
we design products and services
❑ These technologies are being used to create UX more
➢ personalized
➢ engaging, and
➢ efficient
✓ The Future of UX Design: How AI and Machine Learning Are Changing the Way We Design
✓ https://www.forbes.com/sites/forbesbusinesscouncil/2023/03/20/the-future-of-ai-in-banking/
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
THE IMPACT OF AI AND ML ON UX DESIGN
❑ Personalization
➢ ML algorithms can analyze user data, including demographics, behavior, and preferences, to deliver
personalized content and product recommendations
➢ Personalization can lead to increased engagement and satisfaction for users
➢ AI-powered tools can analyze user data from social media profiles, email, and other sources to create
accurate and detailed personas
23
❑ Engagement
➢ By using natural language processing to generate text that is tailored to the user’s interests in a
conversational way, using:
✓ Virtual assistants (e.g Amazon Alexa, Microsoft’s Cortana, Apple Siri, and Google Assistant)
✓ Chatbots, to navigate complex websites or apps, provide personalized assistance and support
✓ While both are conversational interfaces, a virtual assistant is more general and can perform also some actions
❑ Efficiency
➢ Automating tasks that are typically performed by humans by identifying patterns in user behavior
➢ Predictive analytics: By analyzing patterns in user data, AI can make predictions about future user behavior
and suggest actions or information that might be relevant to the user
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
CHATBOTS – (LARGE) LANGUAGE MODELS
❑ Chatbot. A computer program that simulates and processes human
conversation (either written or spoken), allowing humans to interact with
digital devices as if they were communicating with a real person
➢ Used in customer service and support, such as with various sorts of virtual
assistants
✓ Siri, Amazon echo, Amazon Alexa, Microsoft Cortana
➢ Can take previous conversation into account (context)
➢ Use Language Models (LMs and LLMs)
24
❑ Language Model. A probabilistic model of a natural language that can generate probabilities of a series
of words, based on text corpora in one or multiple languages it was trained on
➢ Uses Machine Learning (Neural Networks) to learn from text and produce original text
➢ Predict the next word in a text, speech recognition, optical character recognition and handwriting recognition
➢ A pre-trained deep-learning model that understands and generates text in a human-like fashion
✓ The "pre-training" in its name refers to the initial training process on a large text corpus, which provides a solid foundation
for the model to perform well on downstream tasks with limited amounts of task-specific data
➢ Utilize mostly the Transformer architecture
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
KNOWN EXAMPLES OF LLMS
25
https://chat.openai.com/chat
❑ GPT (Generative Pre-trained Transformer), [OpenAI 2020] - available in Azure
➢ A Transformer-based language model
➢ The incredible power of GPT-3 comes from the fact that it has read more or less all text that has
appeared on the internet over the past years (like all of Wikipedia or crawls of the web)
❑ ChatGPT (Chat Generative pre-trained transformer)
➢ The dialogue format makes it possible for ChatGPT to answer follow up questions, admit its
mistakes, challenge incorrect premises, and reject inappropriate requests
➢ Is it Strong AI?
❑ BioGPT, focuses on answering biomedical questions [Microsoft 2023] [PubMed]
❑ BERT (Bidirectional Encoder Representations from Transformers: a family of language
models) [Google 2018]
➢ GreekBERT
➢ Greek Longformer (Long-Document Transformer)
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
CHATGPT
26
https://chat.openai.com/chat
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
CHATGPT
27
https://chat.openai.com/chat
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
DALL·E 2
28
❑ DALL·E 2 is a new AI system that can create realistic images and art
from a description in natural language.
https://openai.com/dall-e-2/
The Future of AI
e.g. Students attending a lecture on AI in the future
GPT-3 was trained on sequences that were a combination of
words and pixels. It probably contained images and captions
a living room with two white armchairs
and a painting of the collosseum
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
RETAIL APPLICATIONS OF AI
Demand
Forecasting
Inventory
Management
Supply Chain
Management
Optimal Demand Planning
Price
Optimization
Assortment
Planning
Marketing Strategy Optimization
Market
Planning
Chatbot
Assistants
Customer
Segmentation
Customer Relations Intelligence
Churn
Prediction
✔Eliminate Overstocking
and decrease transports
to improve liquidity
✔Make best use of the
marketing budget
✔Provided personalized
experience to customers
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
OPTIMAL DEMAND PLANNING WITH AI
Required data
• Historical Sales
• Historical Promotions
• Upcoming Promotions
• Inventory stock
• Minimum order quantity
• Lead times
• Supplier schedule
• Weather conditions
AI Paradigms
• Supervised Learning
○ Learning from examples
• Regression & Forecasting
• Reinforcement Learning
Demand Forecasting
• Predict future demand for all products
• Benefits:
○ Reduce Overstocking
○ Reduce Stockouts
○ Increase Liquidity
○ Minimize wasted goods
Inventory Management Benefits
• Prepare and submit product delivery orders
• Benefits:
○ Reduce man-months
○ Reduce human errors
○ Improve transparency
Supply Chain Management Benefits
• Optimize the efficiency of the supply chain
• Benefits
○ Reduce unnecessary transports
○ Increase profitability
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
MARKETING STRATEGY OPTIMIZATION WITH AI
Required data
• Historical Sales and Pricing
• Historical Promotions
• Upcoming Promotions
• Store Description
• Space Constraints
• Competitor Pricing
• Competitor Promotions
AI Paradigms
• Supervised Learning
○Learning from examples
• Recommendation System
• Clustering
Market Planning
• Discover effective marketing strategies
• Benefits:
○ Improve Sales and Customer Retention
○ Enhance Brand Awareness
○ Identify Market Trends
Price Optimization
• Determine effective prices based on current demand
• Benefits:
○ Improve revenue and Increase market share
○ React to market dynamics and understand price
elasticity
Assortment planning
• Optimize the allocation of products in store
• Benefits
○ Improved customer experience
○ Improve revenue
○ Achieve optimal product mix
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
CUSTOMER RELATIONS INTELLIGENCE WITH AI
Required data
• Customer Database (anonymised)
• Customer sales history
• Customer promotions history
• Customer conversation history
AI Paradigms
• Supervised Learning
○ Learning from examples
• Data Clustering
• Natural Language Processing
Customer Segmentation
• Group customers with demographics and
purchasing behavior
• Benefits:
○Improve transparency
○Improve strategic decisions
Churn Prediction
• Anticipate the probability of losing customers
• Benefits:
○ Preliminary step for targeted marketing
○ Increase the number of recurring customers
Chatbot assistants
• Automate customer support through helpdesk
• Benefits
○ Decrease man-months and reduce costs
○ Improve customer experience
○ Customer support scalability
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
AI AND SUPPLY CHAIN
❑ McKinsey estimates that logistics companies will generate $1.3-$2 trillion
per year for the next 20 years in economic value by adopting AI into their
processes
❑ Top AI use cases in the logistics industry
➢ Logistics Planning
✓ Requires coordinating suppliers, customers, and different units within the
company
✓ Demand forecasting
✓ Dynamic supply planning to optimize supply chain flow
➢ Automated Warehousing
✓ Warehouse robots, Damage detection, Predictive maintenance
➢ Autonomous Things
✓ Self-driving vehicles, Delivery drones
➢ Analytics
✓ Dynamic Pricing, Route optimization, Sales and marketing analytics, Lead
scoring (propensity modeling)
33
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
AI IN BANKING
34
❑ AI for corporate banking
➢ automates (time-consuming, repetitive) tasks
➢ boosts customer services through chatbots
➢ detects fraud
➢ optimizes investment, and
➢ predicts market trends
❑ Results
➢ increased productivity
➢ lower costs, and
➢ more individualized services
https://appinventiv.com/blog/ai-in-banking/
❑ Why Must the Banking Sector Embrace AI?
➢ For many years, the banking industry has been transforming
from a people-centric business to a customer-centric one
➢ Customers now expect a bank to be there for them
whenever they need it – which means being available 24
hours a day, 7 days a week – and they expect their bank to
do it at scale
➢ The way banks can do this is with AI
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
VERIFICATION OF AI SYSTEMS
❑ How can you ensure that your AI system behaves as expected, meets its specifications, and
does not cause harm or violate ethical principles?
❑ One way is to use formal methods
➢ A set of techniques that use mathematical logic, models, and proofs to specify, design, analyze, and
verify software and hardware systems, increasing trust and transparency
➢ They can also help you to find and fix errors, bugs, or inconsistencies in your AI system, and to provide
evidence of its correctness, safety, and reliability
❑ Runtime verification
➢ Is a computing system based on extracting information from a running system and using it to detect and
possibly react to observed behaviors satisfying or violating certain properties
➢ Avoids the complexity of traditional formal verification techniques, such as model checking and theorem
proving, by analyzing only one or a few execution traces and by working directly with the actual system
❑ Machine Learning applications for Formal Verification
➢ Specification Mining, Automate Troubleshooting
35
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
ETHICS OF AI
36
❑ Questions:
➢ Why do we concern only with AI based decision making software?
➢ Do humans make fair decisions?
❑ Proposals to include a human-in-the-loop (HITL) is crucial in catching and fixing a decision-aid system’s outputs
❑ The use of Artificial Intelligence in the EU will be regulated by the AI Act, the world’s first
comprehensive AI law (Proposed April 2021 and Updated: 14-06-2023)
➢ Once approved, these will be the world’s first rules on AI. It says that
✓ AI can create many benefits, such as better healthcare; safer and cleaner transport; more
efficient manufacturing; and cheaper and more sustainable energy
✓ AI systems used in different applications are classified according to the risk they pose to users
✓ The different risk levels will mean more or less regulation
❑ Recommendation on the Ethics of Artificial Intelligence, by UNESCO
➢ UNESCO produced the first-ever global standard on AI in November 2021
➢ This framework was adopted by all 193 Member States
✓ The protection of human rights and dignity is the cornerstone of the Recommendation, with
respect from policymakers to data governance, environment and ecosystems, gender,
education and research, and health and social wellbeing, among many other spheres
❑ Responsible AI: an approach to developing and deploying artificial intelligence (AI) from both an ethical
and legal point of view
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
Concluding
37
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
HOW AI PRODUCTS SHOULD BE
38
❑ Human-centric
➢ AI products should amplify and augment rather than
displace human abilities
❑ Customizable
➢ AI products should provide a clean interface to
configure them to the needs of each client
❑ Affordable and non-complex
➢ The AI industry should provide low-cost and easy-to-
use applications for everyone, reducing deployment
time and minimizing the need for specialized IT
support for clients
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
INCORPORATING AI INTO YOUR COMPANY
❑ There is no second opinion that AI is transforming businesses
➢ It offers convenience, accessibility, automation and efficiency—all
directly related to achieving more productivity and enhancing user
experience
39
❑ AI allows businesses to reach a larger audience and establish long-term customer relationships
❑ Implementing AI-powered tools in business operations requires a solid plan. Here are a few tips:
➢ Remember that AI can’t fix every issue or perform every task
➢ So, you must examine specific use cases of AI that go well with your company’s overall feasibility and ROI
➢ Set realistic expectations and measure the possible outcomes of this strategy (KPIs)
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
40
AI – THE PRESENCE OF OUR FUTURE
❑ With sixty years of accumulated research, after several false starts and
broken promises, AI is today a reality resulting in a potential for socio-
economic impact
➢ With time, AI is getting more sophisticated and powerful
❑ AI can potentially find a better way of solving old problems or solve some unsolved problems
❑ Customer experience (CX and UX) is key to business success in the digital age
➢ The impact of AI and ML on UX design is significant and will continue to grow in the future
➢ These technologies are already being used to create more personalized, engaging, and efficient user experiences
❑ To make the best use of what tech has to offer today, it’s high time you ditch your legacy systems and
focus on which processes, behaviors, environments, and supply chains can be made more intelligently
automated by integrating AI into your business operations
Data is all you need
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
Intelligent Systems Lab
…THE END
Questions? Thank you!
Prof. Ioannis Vlahavas - School Of Informatics -
AUTH
GENERATIVE ADVERSARIAL NETWORKS (GANS)
❑ In a GAN, two neural networks contest with
each other in the form of a zero-sum game,
where one agent's gain is another agent's
loss
❑ One model (“generator”) learns to generate
new plausible samples
❑ The other model (“discriminator”) learns to
differentiate generated examples from real
examples
❑ Given a training set, this technique learns to
generate new data with the same statistics
as the training set
42

More Related Content

What's hot

The Future of Wearables... EXPLAINED!
The Future of Wearables... EXPLAINED!The Future of Wearables... EXPLAINED!
The Future of Wearables... EXPLAINED!Paul Brown
 
Artificial Intelligence
Artificial Intelligence Artificial Intelligence
Artificial Intelligence XashAxel
 
Power Point Presentation on Artificial Intelligence and Cool Current Projects...
Power Point Presentation on Artificial Intelligence and Cool Current Projects...Power Point Presentation on Artificial Intelligence and Cool Current Projects...
Power Point Presentation on Artificial Intelligence and Cool Current Projects...PuneetGautam6
 
Artificial Intelligence for Product Managers by former Yahoo! PM
Artificial Intelligence for Product Managers by former Yahoo! PMArtificial Intelligence for Product Managers by former Yahoo! PM
Artificial Intelligence for Product Managers by former Yahoo! PMProduct School
 
Creative AI & multimodality: looking ahead
Creative AI & multimodality: looking aheadCreative AI & multimodality: looking ahead
Creative AI & multimodality: looking aheadRoelof Pieters
 
Artificial intelligence & Machine learning role in financial services
Artificial intelligence & Machine learning role in financial servicesArtificial intelligence & Machine learning role in financial services
Artificial intelligence & Machine learning role in financial servicesPrudhvi Parne
 
Artificial intelligence - An Overview
Artificial intelligence - An OverviewArtificial intelligence - An Overview
Artificial intelligence - An OverviewGiri Dharan
 
The 7 Biggest Artificial Intelligence (AI) Trends In 2022
The 7 Biggest Artificial Intelligence (AI) Trends In 2022The 7 Biggest Artificial Intelligence (AI) Trends In 2022
The 7 Biggest Artificial Intelligence (AI) Trends In 2022Bernard Marr
 
ARTIFICIAL INTELLIGENCE
ARTIFICIAL INTELLIGENCEARTIFICIAL INTELLIGENCE
ARTIFICIAL INTELLIGENCEMidhuti
 
Artificial intelligence
Artificial intelligence Artificial intelligence
Artificial intelligence pramiidhaaavula
 
Simplified Introduction to AI
Simplified Introduction to AISimplified Introduction to AI
Simplified Introduction to AIDeepu S Nath
 
Artificial Intelligence explained simplistically
Artificial Intelligence explained simplisticallyArtificial Intelligence explained simplistically
Artificial Intelligence explained simplisticallyNBC Bearings
 
Artificial Intelligence in Business
Artificial Intelligence in BusinessArtificial Intelligence in Business
Artificial Intelligence in BusinessAjay Kumar
 
Machine Learning
Machine LearningMachine Learning
Machine LearningKumar P
 
Artifical Intelligence
Artifical IntelligenceArtifical Intelligence
Artifical IntelligenceHarsha Varyani
 

What's hot (20)

Introduction to data science
Introduction to data scienceIntroduction to data science
Introduction to data science
 
The Future of Wearables... EXPLAINED!
The Future of Wearables... EXPLAINED!The Future of Wearables... EXPLAINED!
The Future of Wearables... EXPLAINED!
 
Artificial Intelligence
Artificial Intelligence Artificial Intelligence
Artificial Intelligence
 
Power Point Presentation on Artificial Intelligence and Cool Current Projects...
Power Point Presentation on Artificial Intelligence and Cool Current Projects...Power Point Presentation on Artificial Intelligence and Cool Current Projects...
Power Point Presentation on Artificial Intelligence and Cool Current Projects...
 
Artificial Intelligence for Product Managers by former Yahoo! PM
Artificial Intelligence for Product Managers by former Yahoo! PMArtificial Intelligence for Product Managers by former Yahoo! PM
Artificial Intelligence for Product Managers by former Yahoo! PM
 
Evolution of AI in workplace.pptx
Evolution of AI in workplace.pptxEvolution of AI in workplace.pptx
Evolution of AI in workplace.pptx
 
Creative AI & multimodality: looking ahead
Creative AI & multimodality: looking aheadCreative AI & multimodality: looking ahead
Creative AI & multimodality: looking ahead
 
Artificial intelligence & Machine learning role in financial services
Artificial intelligence & Machine learning role in financial servicesArtificial intelligence & Machine learning role in financial services
Artificial intelligence & Machine learning role in financial services
 
Artificial intelligence - An Overview
Artificial intelligence - An OverviewArtificial intelligence - An Overview
Artificial intelligence - An Overview
 
The 7 Biggest Artificial Intelligence (AI) Trends In 2022
The 7 Biggest Artificial Intelligence (AI) Trends In 2022The 7 Biggest Artificial Intelligence (AI) Trends In 2022
The 7 Biggest Artificial Intelligence (AI) Trends In 2022
 
Artificial Intelligence
Artificial IntelligenceArtificial Intelligence
Artificial Intelligence
 
ARTIFICIAL INTELLIGENCE
ARTIFICIAL INTELLIGENCEARTIFICIAL INTELLIGENCE
ARTIFICIAL INTELLIGENCE
 
Artificial intelligence
Artificial intelligence Artificial intelligence
Artificial intelligence
 
Simplified Introduction to AI
Simplified Introduction to AISimplified Introduction to AI
Simplified Introduction to AI
 
Artificial Intelligence explained simplistically
Artificial Intelligence explained simplisticallyArtificial Intelligence explained simplistically
Artificial Intelligence explained simplistically
 
Artificial Intelligence in Business
Artificial Intelligence in BusinessArtificial Intelligence in Business
Artificial Intelligence in Business
 
Introduction to Artificial Intelligence and few examples
Introduction to Artificial Intelligence and few examplesIntroduction to Artificial Intelligence and few examples
Introduction to Artificial Intelligence and few examples
 
Future of conversational AI
Future of conversational AIFuture of conversational AI
Future of conversational AI
 
Machine Learning
Machine LearningMachine Learning
Machine Learning
 
Artifical Intelligence
Artifical IntelligenceArtifical Intelligence
Artifical Intelligence
 

Similar to How AI Transforms Businesses

Lesson 1 intro to ai
Lesson 1   intro to aiLesson 1   intro to ai
Lesson 1 intro to aiankit_ppt
 
Machine Learning, AI and the Brain
Machine Learning, AI and the Brain Machine Learning, AI and the Brain
Machine Learning, AI and the Brain TechExeter
 
ARTIFICIAL INTELLIGENCE.pptx
ARTIFICIAL INTELLIGENCE.pptxARTIFICIAL INTELLIGENCE.pptx
ARTIFICIAL INTELLIGENCE.pptxBryCunal
 
Advancement in artificial intelligence: Should Humans be Worried?
Advancement in artificial intelligence: Should Humans be Worried?Advancement in artificial intelligence: Should Humans be Worried?
Advancement in artificial intelligence: Should Humans be Worried?Raymond Owusu
 
IT Presentation .pptx
IT Presentation .pptxIT Presentation .pptx
IT Presentation .pptxbhola22
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligenceMayank Saxena
 
LUISS - Deep Learning and data analyses - 09/01/19
LUISS - Deep Learning and data analyses - 09/01/19LUISS - Deep Learning and data analyses - 09/01/19
LUISS - Deep Learning and data analyses - 09/01/19Alberto Paro
 
A quick guide to artificial intelligence working - Techahead
A quick guide to artificial intelligence working - TechaheadA quick guide to artificial intelligence working - Techahead
A quick guide to artificial intelligence working - TechaheadJatin Sapra
 
Introduction to artificial intelligence
Introduction to artificial intelligenceIntroduction to artificial intelligence
Introduction to artificial intelligenceSindhuVelmukull
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligenceosman ansari
 
Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine Learning Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine Learning Aditya Singh
 
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKS
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKSARTIFICIAL INTELLIGENCE & NEURAL NETWORKS
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKSEr Kaushal
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligencesaloni sharma
 

Similar to How AI Transforms Businesses (20)

Machine learing
Machine learingMachine learing
Machine learing
 
Lesson 1 intro to ai
Lesson 1   intro to aiLesson 1   intro to ai
Lesson 1 intro to ai
 
Machine Learning, AI and the Brain
Machine Learning, AI and the Brain Machine Learning, AI and the Brain
Machine Learning, AI and the Brain
 
ARTIFICIAL INTELLIGENCE.pptx
ARTIFICIAL INTELLIGENCE.pptxARTIFICIAL INTELLIGENCE.pptx
ARTIFICIAL INTELLIGENCE.pptx
 
Ai introduction
Ai introductionAi introduction
Ai introduction
 
Ai introduction ppt
Ai introduction pptAi introduction ppt
Ai introduction ppt
 
Advancement in artificial intelligence: Should Humans be Worried?
Advancement in artificial intelligence: Should Humans be Worried?Advancement in artificial intelligence: Should Humans be Worried?
Advancement in artificial intelligence: Should Humans be Worried?
 
IT Presentation .pptx
IT Presentation .pptxIT Presentation .pptx
IT Presentation .pptx
 
Deep learning
Deep learningDeep learning
Deep learning
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
AI KIMSRAD.pptx
AI KIMSRAD.pptxAI KIMSRAD.pptx
AI KIMSRAD.pptx
 
MIS
MISMIS
MIS
 
LUISS - Deep Learning and data analyses - 09/01/19
LUISS - Deep Learning and data analyses - 09/01/19LUISS - Deep Learning and data analyses - 09/01/19
LUISS - Deep Learning and data analyses - 09/01/19
 
A quick guide to artificial intelligence working - Techahead
A quick guide to artificial intelligence working - TechaheadA quick guide to artificial intelligence working - Techahead
A quick guide to artificial intelligence working - Techahead
 
Introduction to artificial intelligence
Introduction to artificial intelligenceIntroduction to artificial intelligence
Introduction to artificial intelligence
 
When AI becomes a data-driven machine, and digital is everywhere!
When AI becomes a data-driven machine, and digital is everywhere!When AI becomes a data-driven machine, and digital is everywhere!
When AI becomes a data-driven machine, and digital is everywhere!
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine Learning Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine Learning
 
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKS
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKSARTIFICIAL INTELLIGENCE & NEURAL NETWORKS
ARTIFICIAL INTELLIGENCE & NEURAL NETWORKS
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 

More from Uni Systems S.M.S.A.

Uni Systems for Power Platform.pptx
Uni Systems for Power Platform.pptxUni Systems for Power Platform.pptx
Uni Systems for Power Platform.pptxUni Systems S.M.S.A.
 
Microsoft Fabric Intro D Koutsanastasis
Microsoft Fabric Intro D KoutsanastasisMicrosoft Fabric Intro D Koutsanastasis
Microsoft Fabric Intro D KoutsanastasisUni Systems S.M.S.A.
 
Create Unique Experiences through a CRM Approach St Kontos
Create Unique Experiences through a CRM Approach St Kontos Create Unique Experiences through a CRM Approach St Kontos
Create Unique Experiences through a CRM Approach St Kontos Uni Systems S.M.S.A.
 
Bring your data to the era of AI D. Agagiotis.pdf
Bring your data to the era of AI D. Agagiotis.pdfBring your data to the era of AI D. Agagiotis.pdf
Bring your data to the era of AI D. Agagiotis.pdfUni Systems S.M.S.A.
 
Innovative approaches with AI, Data Analytics & CRM
Innovative approaches with AI, Data Analytics & CRMInnovative approaches with AI, Data Analytics & CRM
Innovative approaches with AI, Data Analytics & CRMUni Systems S.M.S.A.
 
Change the Business Landscape with Voice of Customer.pdf
Change the Business Landscape with Voice of Customer.pdfChange the Business Landscape with Voice of Customer.pdf
Change the Business Landscape with Voice of Customer.pdfUni Systems S.M.S.A.
 
Change the Business Landscape with Voice of Customer
Change the Business Landscape with Voice of CustomerChange the Business Landscape with Voice of Customer
Change the Business Landscape with Voice of CustomerUni Systems S.M.S.A.
 
The Evolution in Customer Experience: Migration to Cloud for Contact Center
The Evolution in Customer Experience: Migration to Cloud for Contact CenterThe Evolution in Customer Experience: Migration to Cloud for Contact Center
The Evolution in Customer Experience: Migration to Cloud for Contact CenterUni Systems S.M.S.A.
 
The Evolution of Customer Experience
The Evolution of Customer Experience The Evolution of Customer Experience
The Evolution of Customer Experience Uni Systems S.M.S.A.
 
Using cloud native development to achieve digital transformation
Using cloud native development to achieve digital transformationUsing cloud native development to achieve digital transformation
Using cloud native development to achieve digital transformationUni Systems S.M.S.A.
 
Cloud migration Through Automation
Cloud migration Through AutomationCloud migration Through Automation
Cloud migration Through AutomationUni Systems S.M.S.A.
 

More from Uni Systems S.M.S.A. (20)

Microsoft Power Platform.pptx
Microsoft Power Platform.pptxMicrosoft Power Platform.pptx
Microsoft Power Platform.pptx
 
Uni Systems for Power Platform.pptx
Uni Systems for Power Platform.pptxUni Systems for Power Platform.pptx
Uni Systems for Power Platform.pptx
 
D365 Demonstration CRM G Aspiotis
D365 Demonstration CRM G AspiotisD365 Demonstration CRM G Aspiotis
D365 Demonstration CRM G Aspiotis
 
AI pitch SSideri
 AI pitch SSideri  AI pitch SSideri
AI pitch SSideri
 
Microsoft Fabric Intro D Koutsanastasis
Microsoft Fabric Intro D KoutsanastasisMicrosoft Fabric Intro D Koutsanastasis
Microsoft Fabric Intro D Koutsanastasis
 
Create Unique Experiences through a CRM Approach St Kontos
Create Unique Experiences through a CRM Approach St Kontos Create Unique Experiences through a CRM Approach St Kontos
Create Unique Experiences through a CRM Approach St Kontos
 
Bridging The Gap D Karystinos
 Bridging The Gap D Karystinos Bridging The Gap D Karystinos
Bridging The Gap D Karystinos
 
Bring your data to the era of AI D. Agagiotis.pdf
Bring your data to the era of AI D. Agagiotis.pdfBring your data to the era of AI D. Agagiotis.pdf
Bring your data to the era of AI D. Agagiotis.pdf
 
Innovative approaches with AI, Data Analytics & CRM
Innovative approaches with AI, Data Analytics & CRMInnovative approaches with AI, Data Analytics & CRM
Innovative approaches with AI, Data Analytics & CRM
 
Medallia VoC in action .pdf
Medallia VoC in action .pdfMedallia VoC in action .pdf
Medallia VoC in action .pdf
 
CX Powered by Uni Systems.pdf
CX Powered by Uni Systems.pdfCX Powered by Uni Systems.pdf
CX Powered by Uni Systems.pdf
 
Change the Business Landscape with Voice of Customer.pdf
Change the Business Landscape with Voice of Customer.pdfChange the Business Landscape with Voice of Customer.pdf
Change the Business Landscape with Voice of Customer.pdf
 
CX Powered by Uni Systems
CX Powered by Uni SystemsCX Powered by Uni Systems
CX Powered by Uni Systems
 
Change the Business Landscape with Voice of Customer
Change the Business Landscape with Voice of CustomerChange the Business Landscape with Voice of Customer
Change the Business Landscape with Voice of Customer
 
Medallia VoC in action
Medallia VoC in actionMedallia VoC in action
Medallia VoC in action
 
The Evolution in Customer Experience: Migration to Cloud for Contact Center
The Evolution in Customer Experience: Migration to Cloud for Contact CenterThe Evolution in Customer Experience: Migration to Cloud for Contact Center
The Evolution in Customer Experience: Migration to Cloud for Contact Center
 
The Evolution of Customer Experience
The Evolution of Customer Experience The Evolution of Customer Experience
The Evolution of Customer Experience
 
Using cloud native development to achieve digital transformation
Using cloud native development to achieve digital transformationUsing cloud native development to achieve digital transformation
Using cloud native development to achieve digital transformation
 
Microsoft: Invent with Purpose
Microsoft: Invent with PurposeMicrosoft: Invent with Purpose
Microsoft: Invent with Purpose
 
Cloud migration Through Automation
Cloud migration Through AutomationCloud migration Through Automation
Cloud migration Through Automation
 

Recently uploaded

Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...dajasot375
 
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Jack DiGiovanna
 
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...Sapana Sha
 
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改yuu sss
 
04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationshipsccctableauusergroup
 
B2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxB2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxStephen266013
 
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一办理学位证纽约大学毕业证(NYU毕业证书)原版一比一
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一fhwihughh
 
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...soniya singh
 
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一F sss
 
Call Girls In Dwarka 9654467111 Escorts Service
Call Girls In Dwarka 9654467111 Escorts ServiceCall Girls In Dwarka 9654467111 Escorts Service
Call Girls In Dwarka 9654467111 Escorts ServiceSapana Sha
 
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhi
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝DelhiRS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhi
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhijennyeacort
 
ASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel CanterASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel Cantervoginip
 
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...Florian Roscheck
 
20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdfHuman37
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queensdataanalyticsqueen03
 
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一F La
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)jennyeacort
 
Customer Service Analytics - Make Sense of All Your Data.pptx
Customer Service Analytics - Make Sense of All Your Data.pptxCustomer Service Analytics - Make Sense of All Your Data.pptx
Customer Service Analytics - Make Sense of All Your Data.pptxEmmanuel Dauda
 
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxNLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxBoston Institute of Analytics
 
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档208367051
 

Recently uploaded (20)

Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
 
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
 
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...
Saket, (-DELHI )+91-9654467111-(=)CHEAP Call Girls in Escorts Service Saket C...
 
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
 
04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships
 
B2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxB2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docx
 
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一办理学位证纽约大学毕业证(NYU毕业证书)原版一比一
办理学位证纽约大学毕业证(NYU毕业证书)原版一比一
 
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
 
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
办理学位证中佛罗里达大学毕业证,UCF成绩单原版一比一
 
Call Girls In Dwarka 9654467111 Escorts Service
Call Girls In Dwarka 9654467111 Escorts ServiceCall Girls In Dwarka 9654467111 Escorts Service
Call Girls In Dwarka 9654467111 Escorts Service
 
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhi
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝DelhiRS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhi
RS 9000 Call In girls Dwarka Mor (DELHI)⇛9711147426🔝Delhi
 
ASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel CanterASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel Canter
 
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
 
20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queens
 
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
 
Customer Service Analytics - Make Sense of All Your Data.pptx
Customer Service Analytics - Make Sense of All Your Data.pptxCustomer Service Analytics - Make Sense of All Your Data.pptx
Customer Service Analytics - Make Sense of All Your Data.pptx
 
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxNLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
 
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
 

How AI Transforms Businesses

  • 1. Prof. Ioannis Vlahavas - School Of Informatics - AUTH Intelligent Systems Lab The Presence of our Future HOW AI TRANSFORMS BUSINESSES
  • 2. Prof. Ioannis Vlahavas - School Of Informatics - AUTH INTRODUCTION ❑ Lovely tech buzz words and grand promises: ➢ "The Internet Of Things will connect everything" ➢ "Blockchain will democratize everything", and ➢ "AI will solve all of our problems" ❑ AI can potentially find a better way of solving old problems or solve some unsolved problems ❑ To hit the target with AI tech, we first must understand what this technology can do 2 ❑ For a machine to be able to determine that an image contains a cat, many thousands of tagged images had to be processed using Machine Learning ➢ This "self-learning" process is fascinating, useful, and has multiple business use cases ➢ But is it intelligence? ❑ For comparison, a child needs to experience two interactions with a cat to be able to identify a cat the third time they meet ➢ This is a significant difference in capabilities, and it highlights that how people and machines think, learn, and understand is very different ❑ Definitely, for the present, but also the foreseeable future, the business opportunities of leveraging what we call "AI" is more IA (Intelligent Automation) than AI (Artificial Intelligence)
  • 3. Prof. Ioannis Vlahavas - School Of Informatics - AUTH AI is a One-Way street The AI market size is projecting an increase from 20 Billion USD in 2020 to more than 70 Billion USD by 2024 and 260 Billion USD by 2028. The market Companies store a large amount of (proprietary) data concerning their customers, products, services and operations. Data Data utilization constitutes the only way to enable company development and survival in the highly competitive global financial landscape. Utilization 3
  • 4. Prof. Ioannis Vlahavas - School Of Informatics - AUTH HUGE AMOUNT OF DATA ❑ We are creating more than 2.5 Quintillion (i.e. Million Billion) bytes of data every year ➢ 90% of all data ever created, was created in the past 2 years ➢ From now on, the amount of data in the world will double every two years ❑ Data comes from sensors, devices, video/audio, transactional applications, networks, log files, web, social media, etc. ➢ Data Never Sleeps ❑ Big Data (Volume, Variety, Velocity, Veracity) “Mining needle in a haystack. So much hay and so little time” ⮚ Traditional computing techniques are not able to handle such large datasets ⮚ Adopted solution: Artificial Intelligence 4
  • 5. Prof. Ioannis Vlahavas - School Of Informatics - AUTH WHAT IS AI ❑ Artificial Intelligence (AI) is a wide-ranging branch of computer science concerned with building smart machines capable of performing tasks that typically require human intelligence ➢ Researchers (and inventors) have long dreamed of creating machines that think ➢ Turing test ❑ Is an interdisciplinary science with multiple approaches, affecting every sector of the tech industry ❑ Uses statistical methods but is completely a different approach compared to statistics 5 ⮚ Not magic or science fiction - but rather science, engineering, and mathematics ⮚ However, applications… are magic
  • 6. Prof. Ioannis Vlahavas - School Of Informatics - AUTH TYPES OF AI ❑ Weak AI or Narrow AI: ➢ Is a simulation of human intelligence ➢ Machines act as if they were intelligent ❑ Strong AI or Artificial General Intelligence (AGI): ➢ Machines are actually thinking (not just simulating thinking) ➢ Machines, much like a human being, can apply that intelligence to solve any problem ➢ Is the kind of Artificial Intelligence we see in the movies 6
  • 7. Prof. Ioannis Vlahavas - School Of Informatics - AUTH STRONG AI ❑ Strong or General Artificial Intelligence is very different, and is the type of adaptable intellect found in humans, a flexible form of intelligence capable of: ➢ Actual thinking - Consciousness ➢ Learning how to carry out vastly different tasks ✓ anything from haircutting to building spreadsheets, or ➢ To reason about a wide variety of topics based on its accumulated experience 7 ❑ Is it feasible? Is it desirable? Interview of humanoid Sophia In October 2017, Sophia became the first robot to receive citizenship ❑ Machines have neither past experiences (if they do, they are limited to narrow fields such as identifying a pattern of a cat, driving in a lane, or picking an object), and they most certainly do not have feelings ❑ Missing these components, how can machines think like us? ❑ What is becoming clear is that the holy grail of AI should not be to teach machines to think like humans, but to teach machines to think like machines
  • 8. Prof. Ioannis Vlahavas - School Of Informatics - AUTH WEAK AI ❑ Is what we see all around us in computers today ➢ Intelligent systems that have been taught how to carry out specific tasks without being explicitly programmed how to do so ❑ This type of machine intelligence is evident in ➢ Medicine diagnosis ➢ Fraud detection ➢ Recommendation systems that suggest products (e.g. Netflix's recommendations) ➢ NLP and speech recognition ✓ e.g. Cortana by Microsoft, Siri virtual assistant and Amazon's Alexa ➢ Vision recognition systems on self-driving cars (and mobiles) ➢ Robotics (Atlas by Boston Dynamics) ➢ Self-driving cars ➢ Game playing (playing chess) ✓ IBM’s DEEP BLUE became the first computer program to defeat the world champion Garry Kasparov in a chess match (1997) ➢ Machine Translation (Google and Microsoft translate) ➢ …etc …etc ❑ These are just a few examples of AI systems that exist today 8 Boston Dynamics' Atlas has now mastered parkour
  • 9. Prof. Ioannis Vlahavas - School Of Informatics - AUTH AI AND MACHINE LEARNING ➢ This capability is Machine Learning 9 ❑ AI is a new phenomenon, but it is not a new concept ➢ It came into existence in 1956, but it took decades of work to make significant progress toward developing an AI system and making it a technological reality ❑ Many AI projects have been developed in the past with hard-coding the knowledge of a problem’s world in formal languages. None has led to major success. ➢ Rule or Knowledge based approach ➢ “frozen software” — software that can only be improved via updates ❑ These difficulties led to AI systems with the ability to obtain their own knowledge, by extracting patterns from raw data
  • 10. Prof. Ioannis Vlahavas - School Of Informatics - AUTH WHAT IS MACHINE LEARNING ❑ Machine learning is the subfield of computer science that gives computers the ability to learn without being explicitly programmed Arthur Samuel, 1959 10 ❑ Machine learning feeds data into a computer and uses statistical techniques to help it "learn" how to get progressively better at a task, without having been specifically programmed for that task, eliminating the need for millions of lines of written code. ⮚ Data=Actionable Information
  • 11. Prof. Ioannis Vlahavas - School Of Informatics - AUTH MACHINE LEARNING CATEGORIES 11 ❑ Supervised learning (using labeled data sets) ➢ Output: Prediction function (model) for Classification or Regression ➢ Applications: Forecasting (e.g. Demand), Predictive maintenance, etc ➢ Many algorithms: Support Vector Machines, Linear regression, KNN, Decision trees (Random Forest), Naive Bayes, Neural Networks, etc ❑ Unsupervised learning (using unlabeled data sets) ➢ Output: Clusters, Association Rules ➢ Applications: Clustering (e.g. customer segmentation), Market basket analysis, etc ➢ Algorithms: k-means, Dbscan, Hierarchical, Apriori ❑ Reinforcement learning that allows an agent to learn a behaviour through trial-and-error interactions with a dynamic environment ➢ Output: Intelligent agent ➢ Applications: self-driving cars, industry automation, trading, gaming, robotics, etc ➢ Algorithms: Q-learning, SARSA, Proximal Policy Optimization (PPO), DQN, Deep Recurrent Q-Networks, …etc
  • 12. Prof. Ioannis Vlahavas - School Of Informatics - AUTH ARTIFICIAL NEURAL NETWORKS (ANNS) ❑ They have their origins back to 1943 (artificial neuron as a computational model) ❑ Their name and structure are inspired by the human brain, mimicking the way that biological neurons signal to one another ❑ They are comprised of nodes (neuron) layers, containing an input layer, one or more hidden layers, and an output layer ❑ A neuron (or Perceptron) is a mathematical function that just take inputs on their “dendrites” and generate output on their “axon branches” ❑ Different architectures: Number of neurons, of layers, activation functions, dropout rates and optimization techniques are preferable for different tasks 12 ----------------------------------------- Our brain contains about 86 billion neurons and more than a 100 trillion (or according to some estimates 1000 trillion) synapses (connections)
  • 13. Prof. Ioannis Vlahavas - School Of Informatics - AUTH DEEP LEARNING ❑ Probably one of the hottest topics in Machine Learning today ➢ Achieves remarkable outcomes ❑ Uses large multi-layered Neural Networks 13 ❑ Different Kinds ➢ Deep Feedforward Networks or Multilayer Perceptrons (MLPs) ➢ CNN (Convolutional NN) [1995] ✓ TCN (Temporal Convolutional Networks) [2016] ➢ RNN (Recurrent NN) [1986] ✓ LSTMs (Long Short-Term Memory Units) [1997] ✓ Gated RNN (GRU) [2014] ✓ Encoder and Decoder through an attention mechanism [2015] ➢ GAN (Generative Adversarial Networks) [2014] – Generative AI ➢ Transformer [2017]
  • 14. Prof. Ioannis Vlahavas - School Of Informatics - AUTH TRANSFORMER NEURAL NETWORK ❑ Like most neural networks, transformer models are basically large encoder/decoder blocks that process data ➢ Small but strategic additions to these blocks make transformers uniquely powerful ❑ Transformer models apply an evolving set of mathematical techniques, called attention or self-attention, to detect subtle ways even distant data elements ➢ RNNs favors more recent words at the end of a sentence while earlier words fade away in volatile neural activations ➢ Attention (and Multi-Head Attention) gives all words equal access to any part of a sentence in a faster parallel scheme and no longer suffers the wait time of serial processing ❑ They are in many cases replacing convolutional (CNNs) and recurrent neural networks (RNNs), the most popular types of deep learning models just five years ago ➢ Today, many AI engineers are working on trillion-parameter transformers ❑ This architecture is now used not only in natural language processing and computer vision, but also in audio and multi-modal processing ➢ It has also led to the development of pre-trained Large Language Models, such as Generative Pre-trained Transformers (GPTs) and Bidirectional Encoder Representations from Transformers (BERT) 14 ➢ https://towardsdatascience.com/transformer-neural-network-step-by-step-breakdown-of-the-beast-b3e096dc857f ➢ 15 of the best large language models
  • 15. Prof. Ioannis Vlahavas - School Of Informatics - AUTH DEEP LEARNING: APPLICATIONS 15 ❑ NLP (e.g. ChatGPT) ❑ Speech recognition (e.g. Voice based search, Siri, Amazon echo, Amazon Alexa, Microsoft Cortana) ❑ Image recognition (e.g. Tagging faces in photos, smartphones) ❑ Pattern detection (e.g. Handwriting recognition) ❑ Forecasting (Retail, stocks, ….) ➢ For all the Weak AI applications, also
  • 16. Prof. Ioannis Vlahavas - School Of Informatics - AUTH DEEP LEARNING IN THE FOREGROUND ❑ Why DNN (and Deep Learning) is now in the foreground? Because… ❑ We now have fast enough computers (GPUs), software, libraries and huge amount of data to train large NNs ➢ Programming Languages ✔ Java, Python, R, SCALA, Julia, Spark ➢ Libraries-Frameworks ✔ Keras (Tensorflow, Theano) developed by Google, are python libraries that make writing deep learning models easy, and give the option of training them on a GPU ✔ PyTorch developed by Facebook ⮚ New architectures/models, like Transformers, used by BERT and GPT3 on which chatGPT has been based ➢ AI assistant and Developer Tools ✓ Copilot (Microsoft) ⮚ AI companies (like Hugging Face) that offer open source tools and models 16
  • 17. Prof. Ioannis Vlahavas - School Of Informatics - AUTH DEEP LEARNING: PROS AND CONS 17 ❑ Cons ➢ Neural networks are black boxes ✓ We cannot know how much each independent variable is influencing the dependent variables ➢ Hardware Requirements ✓ It is computationally very expensive and time consuming to train with traditional CPUs (→ GPUs) ➢ Depend a lot on training data ❑ Pros ➢ Powerful tool ➢ Can be used for all categories of ML ➢ Are good to model with nonlinear data with large number of inputs; for example, images ➢ Once trained, the predictions are pretty fast
  • 18. Prof. Ioannis Vlahavas - School Of Informatics - AUTH GENERATIVE AI ❑ The branch of AI, which uses available text, audio files, images, videos to create a whole new set of the same which seems to be true and perfect in its own senses ➢ Generation mainly of images, 3D objects, Text (Articles, Blog posts, Product descriptions), or even code ❑ Gartner predicts that by 2025, the percentage of data generated by generative AI will amount to 10% of all generated data 18 ✓ https://machinelearningmastery.com/what-are-generative-adversarial-networks-gans/ ✓ Top 100+ Generative AI Applications / Use Cases in 2023 ❑ Generative AI has two famous models for working: ➢ The Generative Adversarial Networks (GANs) combined with AutoRegressive Convolutional Neural Networks(AR-CNN), and ➢ Transformer-based Models
  • 19. Prof. Ioannis Vlahavas - School Of Informatics - AUTH 19 CONCLUDING ML ❑ Machine learning is a powerful Artificial Intelligence tool that enables us to crunch petabytes of data and make sense of a complicated world ➢ …it's transforming a wide variety of industries/organizations ➢ …It's solving previously unsolved problems ❑ AI systems learn and amplify human biases ➢ keep humans in the loop ➢ “People can still play a role – either when validating a fraud or following up with an action (e.g. a rejected transaction/account)” ❑ Because finding patterns is hard, often not enough training data is available, and also because of the high expectations it often fails to deliver ❑ Even with machine learning in place, business leaders remain skeptical of error rates ➢ This is a cognitive bias; people magnify the failures of machine learning ➢ Look at the media firestorm on the few self-driving car mishaps vs. the thousands to millions of people dying due to other causes
  • 20. Prof. Ioannis Vlahavas - School Of Informatics - AUTH HOW AI TRANSFORMS BUSINESSES INDICATIVE APPLICATIONS 20 ➢ User Experience (UX) ➢ Chat bots (NLP, Large Language Models) ➢ Retail Applications of AI ➢ AI and Supply Chain ➢ AI in Banking
  • 21. Prof. Ioannis Vlahavas - School Of Informatics - AUTH AI IN BIG COMPANIES ❑ Microsoft ➢ Azure ✓ It provides SaaS, PaaS and IaaS ✓ Supports many programming languages, tools and frameworks, Microsoft and third-party ➢ Azure ML ✓ A cloud service for accelerating and managing machine learning projects ➢ Microsoft Cognitive Toolkit (CNTK) (is no longer actively developed) ✓ An open-source toolkit for commercial-grade distributed deep learning ❑ Oracle AI ➢ Oracle Cloud Infrastructure (OCI) AI/Data Science Services ➢ A fully-managed platform for teams of data scientists to build, train, deploy, and manage machine learning models using Python and open source tools ❑ IBM, Watson and watsonx ➢ IBM's enterprise-ready AI and data platform designed to leverage foundation models and machine learning ❑ Amazon – AWS Deep Learning AMIs (Amazon Machine Images) ➢ Provides ML practitioners and researchers with a curated and secure set of frameworks, dependencies, and tools to accelerate deep learning on Amazon 21
  • 22. Prof. Ioannis Vlahavas - School Of Informatics - AUTH USER EXPERIENCE (UX) ❑ UX refers to the feeling users experience when using a product, application, system, or service ❑ It is a broad term that can cover anything from ➢ how well the user can navigate the product ➢ how easy it is to use ➢ how relevant the content displayed is ➢ etc. 22 ❑ AI and machine learning (ML) are rapidly changing the way we design products and services ❑ These technologies are being used to create UX more ➢ personalized ➢ engaging, and ➢ efficient ✓ The Future of UX Design: How AI and Machine Learning Are Changing the Way We Design ✓ https://www.forbes.com/sites/forbesbusinesscouncil/2023/03/20/the-future-of-ai-in-banking/
  • 23. Prof. Ioannis Vlahavas - School Of Informatics - AUTH THE IMPACT OF AI AND ML ON UX DESIGN ❑ Personalization ➢ ML algorithms can analyze user data, including demographics, behavior, and preferences, to deliver personalized content and product recommendations ➢ Personalization can lead to increased engagement and satisfaction for users ➢ AI-powered tools can analyze user data from social media profiles, email, and other sources to create accurate and detailed personas 23 ❑ Engagement ➢ By using natural language processing to generate text that is tailored to the user’s interests in a conversational way, using: ✓ Virtual assistants (e.g Amazon Alexa, Microsoft’s Cortana, Apple Siri, and Google Assistant) ✓ Chatbots, to navigate complex websites or apps, provide personalized assistance and support ✓ While both are conversational interfaces, a virtual assistant is more general and can perform also some actions ❑ Efficiency ➢ Automating tasks that are typically performed by humans by identifying patterns in user behavior ➢ Predictive analytics: By analyzing patterns in user data, AI can make predictions about future user behavior and suggest actions or information that might be relevant to the user
  • 24. Prof. Ioannis Vlahavas - School Of Informatics - AUTH CHATBOTS – (LARGE) LANGUAGE MODELS ❑ Chatbot. A computer program that simulates and processes human conversation (either written or spoken), allowing humans to interact with digital devices as if they were communicating with a real person ➢ Used in customer service and support, such as with various sorts of virtual assistants ✓ Siri, Amazon echo, Amazon Alexa, Microsoft Cortana ➢ Can take previous conversation into account (context) ➢ Use Language Models (LMs and LLMs) 24 ❑ Language Model. A probabilistic model of a natural language that can generate probabilities of a series of words, based on text corpora in one or multiple languages it was trained on ➢ Uses Machine Learning (Neural Networks) to learn from text and produce original text ➢ Predict the next word in a text, speech recognition, optical character recognition and handwriting recognition ➢ A pre-trained deep-learning model that understands and generates text in a human-like fashion ✓ The "pre-training" in its name refers to the initial training process on a large text corpus, which provides a solid foundation for the model to perform well on downstream tasks with limited amounts of task-specific data ➢ Utilize mostly the Transformer architecture
  • 25. Prof. Ioannis Vlahavas - School Of Informatics - AUTH KNOWN EXAMPLES OF LLMS 25 https://chat.openai.com/chat ❑ GPT (Generative Pre-trained Transformer), [OpenAI 2020] - available in Azure ➢ A Transformer-based language model ➢ The incredible power of GPT-3 comes from the fact that it has read more or less all text that has appeared on the internet over the past years (like all of Wikipedia or crawls of the web) ❑ ChatGPT (Chat Generative pre-trained transformer) ➢ The dialogue format makes it possible for ChatGPT to answer follow up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests ➢ Is it Strong AI? ❑ BioGPT, focuses on answering biomedical questions [Microsoft 2023] [PubMed] ❑ BERT (Bidirectional Encoder Representations from Transformers: a family of language models) [Google 2018] ➢ GreekBERT ➢ Greek Longformer (Long-Document Transformer)
  • 26. Prof. Ioannis Vlahavas - School Of Informatics - AUTH CHATGPT 26 https://chat.openai.com/chat
  • 27. Prof. Ioannis Vlahavas - School Of Informatics - AUTH CHATGPT 27 https://chat.openai.com/chat
  • 28. Prof. Ioannis Vlahavas - School Of Informatics - AUTH DALL·E 2 28 ❑ DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language. https://openai.com/dall-e-2/ The Future of AI e.g. Students attending a lecture on AI in the future GPT-3 was trained on sequences that were a combination of words and pixels. It probably contained images and captions a living room with two white armchairs and a painting of the collosseum
  • 29. Prof. Ioannis Vlahavas - School Of Informatics - AUTH RETAIL APPLICATIONS OF AI Demand Forecasting Inventory Management Supply Chain Management Optimal Demand Planning Price Optimization Assortment Planning Marketing Strategy Optimization Market Planning Chatbot Assistants Customer Segmentation Customer Relations Intelligence Churn Prediction ✔Eliminate Overstocking and decrease transports to improve liquidity ✔Make best use of the marketing budget ✔Provided personalized experience to customers
  • 30. Prof. Ioannis Vlahavas - School Of Informatics - AUTH OPTIMAL DEMAND PLANNING WITH AI Required data • Historical Sales • Historical Promotions • Upcoming Promotions • Inventory stock • Minimum order quantity • Lead times • Supplier schedule • Weather conditions AI Paradigms • Supervised Learning ○ Learning from examples • Regression & Forecasting • Reinforcement Learning Demand Forecasting • Predict future demand for all products • Benefits: ○ Reduce Overstocking ○ Reduce Stockouts ○ Increase Liquidity ○ Minimize wasted goods Inventory Management Benefits • Prepare and submit product delivery orders • Benefits: ○ Reduce man-months ○ Reduce human errors ○ Improve transparency Supply Chain Management Benefits • Optimize the efficiency of the supply chain • Benefits ○ Reduce unnecessary transports ○ Increase profitability
  • 31. Prof. Ioannis Vlahavas - School Of Informatics - AUTH MARKETING STRATEGY OPTIMIZATION WITH AI Required data • Historical Sales and Pricing • Historical Promotions • Upcoming Promotions • Store Description • Space Constraints • Competitor Pricing • Competitor Promotions AI Paradigms • Supervised Learning ○Learning from examples • Recommendation System • Clustering Market Planning • Discover effective marketing strategies • Benefits: ○ Improve Sales and Customer Retention ○ Enhance Brand Awareness ○ Identify Market Trends Price Optimization • Determine effective prices based on current demand • Benefits: ○ Improve revenue and Increase market share ○ React to market dynamics and understand price elasticity Assortment planning • Optimize the allocation of products in store • Benefits ○ Improved customer experience ○ Improve revenue ○ Achieve optimal product mix
  • 32. Prof. Ioannis Vlahavas - School Of Informatics - AUTH CUSTOMER RELATIONS INTELLIGENCE WITH AI Required data • Customer Database (anonymised) • Customer sales history • Customer promotions history • Customer conversation history AI Paradigms • Supervised Learning ○ Learning from examples • Data Clustering • Natural Language Processing Customer Segmentation • Group customers with demographics and purchasing behavior • Benefits: ○Improve transparency ○Improve strategic decisions Churn Prediction • Anticipate the probability of losing customers • Benefits: ○ Preliminary step for targeted marketing ○ Increase the number of recurring customers Chatbot assistants • Automate customer support through helpdesk • Benefits ○ Decrease man-months and reduce costs ○ Improve customer experience ○ Customer support scalability
  • 33. Prof. Ioannis Vlahavas - School Of Informatics - AUTH AI AND SUPPLY CHAIN ❑ McKinsey estimates that logistics companies will generate $1.3-$2 trillion per year for the next 20 years in economic value by adopting AI into their processes ❑ Top AI use cases in the logistics industry ➢ Logistics Planning ✓ Requires coordinating suppliers, customers, and different units within the company ✓ Demand forecasting ✓ Dynamic supply planning to optimize supply chain flow ➢ Automated Warehousing ✓ Warehouse robots, Damage detection, Predictive maintenance ➢ Autonomous Things ✓ Self-driving vehicles, Delivery drones ➢ Analytics ✓ Dynamic Pricing, Route optimization, Sales and marketing analytics, Lead scoring (propensity modeling) 33
  • 34. Prof. Ioannis Vlahavas - School Of Informatics - AUTH AI IN BANKING 34 ❑ AI for corporate banking ➢ automates (time-consuming, repetitive) tasks ➢ boosts customer services through chatbots ➢ detects fraud ➢ optimizes investment, and ➢ predicts market trends ❑ Results ➢ increased productivity ➢ lower costs, and ➢ more individualized services https://appinventiv.com/blog/ai-in-banking/ ❑ Why Must the Banking Sector Embrace AI? ➢ For many years, the banking industry has been transforming from a people-centric business to a customer-centric one ➢ Customers now expect a bank to be there for them whenever they need it – which means being available 24 hours a day, 7 days a week – and they expect their bank to do it at scale ➢ The way banks can do this is with AI
  • 35. Prof. Ioannis Vlahavas - School Of Informatics - AUTH VERIFICATION OF AI SYSTEMS ❑ How can you ensure that your AI system behaves as expected, meets its specifications, and does not cause harm or violate ethical principles? ❑ One way is to use formal methods ➢ A set of techniques that use mathematical logic, models, and proofs to specify, design, analyze, and verify software and hardware systems, increasing trust and transparency ➢ They can also help you to find and fix errors, bugs, or inconsistencies in your AI system, and to provide evidence of its correctness, safety, and reliability ❑ Runtime verification ➢ Is a computing system based on extracting information from a running system and using it to detect and possibly react to observed behaviors satisfying or violating certain properties ➢ Avoids the complexity of traditional formal verification techniques, such as model checking and theorem proving, by analyzing only one or a few execution traces and by working directly with the actual system ❑ Machine Learning applications for Formal Verification ➢ Specification Mining, Automate Troubleshooting 35
  • 36. Prof. Ioannis Vlahavas - School Of Informatics - AUTH ETHICS OF AI 36 ❑ Questions: ➢ Why do we concern only with AI based decision making software? ➢ Do humans make fair decisions? ❑ Proposals to include a human-in-the-loop (HITL) is crucial in catching and fixing a decision-aid system’s outputs ❑ The use of Artificial Intelligence in the EU will be regulated by the AI Act, the world’s first comprehensive AI law (Proposed April 2021 and Updated: 14-06-2023) ➢ Once approved, these will be the world’s first rules on AI. It says that ✓ AI can create many benefits, such as better healthcare; safer and cleaner transport; more efficient manufacturing; and cheaper and more sustainable energy ✓ AI systems used in different applications are classified according to the risk they pose to users ✓ The different risk levels will mean more or less regulation ❑ Recommendation on the Ethics of Artificial Intelligence, by UNESCO ➢ UNESCO produced the first-ever global standard on AI in November 2021 ➢ This framework was adopted by all 193 Member States ✓ The protection of human rights and dignity is the cornerstone of the Recommendation, with respect from policymakers to data governance, environment and ecosystems, gender, education and research, and health and social wellbeing, among many other spheres ❑ Responsible AI: an approach to developing and deploying artificial intelligence (AI) from both an ethical and legal point of view
  • 37. Prof. Ioannis Vlahavas - School Of Informatics - AUTH Concluding 37
  • 38. Prof. Ioannis Vlahavas - School Of Informatics - AUTH HOW AI PRODUCTS SHOULD BE 38 ❑ Human-centric ➢ AI products should amplify and augment rather than displace human abilities ❑ Customizable ➢ AI products should provide a clean interface to configure them to the needs of each client ❑ Affordable and non-complex ➢ The AI industry should provide low-cost and easy-to- use applications for everyone, reducing deployment time and minimizing the need for specialized IT support for clients
  • 39. Prof. Ioannis Vlahavas - School Of Informatics - AUTH INCORPORATING AI INTO YOUR COMPANY ❑ There is no second opinion that AI is transforming businesses ➢ It offers convenience, accessibility, automation and efficiency—all directly related to achieving more productivity and enhancing user experience 39 ❑ AI allows businesses to reach a larger audience and establish long-term customer relationships ❑ Implementing AI-powered tools in business operations requires a solid plan. Here are a few tips: ➢ Remember that AI can’t fix every issue or perform every task ➢ So, you must examine specific use cases of AI that go well with your company’s overall feasibility and ROI ➢ Set realistic expectations and measure the possible outcomes of this strategy (KPIs)
  • 40. Prof. Ioannis Vlahavas - School Of Informatics - AUTH 40 AI – THE PRESENCE OF OUR FUTURE ❑ With sixty years of accumulated research, after several false starts and broken promises, AI is today a reality resulting in a potential for socio- economic impact ➢ With time, AI is getting more sophisticated and powerful ❑ AI can potentially find a better way of solving old problems or solve some unsolved problems ❑ Customer experience (CX and UX) is key to business success in the digital age ➢ The impact of AI and ML on UX design is significant and will continue to grow in the future ➢ These technologies are already being used to create more personalized, engaging, and efficient user experiences ❑ To make the best use of what tech has to offer today, it’s high time you ditch your legacy systems and focus on which processes, behaviors, environments, and supply chains can be made more intelligently automated by integrating AI into your business operations Data is all you need
  • 41. Prof. Ioannis Vlahavas - School Of Informatics - AUTH Intelligent Systems Lab …THE END Questions? Thank you!
  • 42. Prof. Ioannis Vlahavas - School Of Informatics - AUTH GENERATIVE ADVERSARIAL NETWORKS (GANS) ❑ In a GAN, two neural networks contest with each other in the form of a zero-sum game, where one agent's gain is another agent's loss ❑ One model (“generator”) learns to generate new plausible samples ❑ The other model (“discriminator”) learns to differentiate generated examples from real examples ❑ Given a training set, this technique learns to generate new data with the same statistics as the training set 42