by Samantha Adams, Met Office.
Originally purely academic research fields, Machine Learning and AI are now definitely mainstream and frequently mentioned in the Tech media (and regular media too).
We’ve also got the explosion of Data Science which encompasses these fields and more. There’s a lot of interesting things going on and a lot of positive as well as negative hype. The terms ML and AI are often used interchangeably and techniques are also often described as being inspired by the brain.
In this talk I will explore the history and evolution of these fields, current progress and the challenges in making artificial brains
From the FreshTech 2017 conference by TechExeter
www.techexeter.uk
Introduction to Artificial IntelligenceSanjay Kumar
This presentation talks about what is Artificial Intelligence, what are key Algorithms (CNN, RNN, Reinforcement Learning), their applications. AI use cases such as detecting fish species and Spoting Distracted Driver
Webinar on AI in IoT applications KCG Connect Alumni Digital Series by RajkumarRajkumar R
The Artificial Intelligence in IoT Applications. Take your first step towards a bright future with our renowned alumnus,
Prof R. Raj Kumar on AI for IoT Applications.
He is an award wining author of the book, ‘India 2030’.
To get access to the webinar kindly contact your respective department heads.
Looking forward to having you on the webinar.
.
.
.
#KCGCollege #KCGStudentlife #KCGConnect #Education #EmergingTechnologies #ArtificialIntelligence #IoT #MachineLearning #BlockChain #ElectricVehicle #QuantumTechnology #CAD
Our co-founder and CTO, Murray Cantor Ph.D, gave an introductory presentation on the history of Artificial Intelligence (AI). In the presentation he explains what AI means for business today.
by Samantha Adams, Met Office.
Originally purely academic research fields, Machine Learning and AI are now definitely mainstream and frequently mentioned in the Tech media (and regular media too).
We’ve also got the explosion of Data Science which encompasses these fields and more. There’s a lot of interesting things going on and a lot of positive as well as negative hype. The terms ML and AI are often used interchangeably and techniques are also often described as being inspired by the brain.
In this talk I will explore the history and evolution of these fields, current progress and the challenges in making artificial brains
From the FreshTech 2017 conference by TechExeter
www.techexeter.uk
Introduction to Artificial IntelligenceSanjay Kumar
This presentation talks about what is Artificial Intelligence, what are key Algorithms (CNN, RNN, Reinforcement Learning), their applications. AI use cases such as detecting fish species and Spoting Distracted Driver
Webinar on AI in IoT applications KCG Connect Alumni Digital Series by RajkumarRajkumar R
The Artificial Intelligence in IoT Applications. Take your first step towards a bright future with our renowned alumnus,
Prof R. Raj Kumar on AI for IoT Applications.
He is an award wining author of the book, ‘India 2030’.
To get access to the webinar kindly contact your respective department heads.
Looking forward to having you on the webinar.
.
.
.
#KCGCollege #KCGStudentlife #KCGConnect #Education #EmergingTechnologies #ArtificialIntelligence #IoT #MachineLearning #BlockChain #ElectricVehicle #QuantumTechnology #CAD
Our co-founder and CTO, Murray Cantor Ph.D, gave an introductory presentation on the history of Artificial Intelligence (AI). In the presentation he explains what AI means for business today.
In this deck from the HPC User Forum in Milwaukee, Michael Garris from NIST presents: The National Science & Technology Council ML/AI Initiative.
"AI-enabled systems are beginning to revolutionize fields such as commerce, healthcare, transportation and cybersecurity. It has the potential to impact nearly all aspects of our society including our economy, yet its development and use come with serious technical and ethical challenges and risks. AI must be developed in a trustworthy manner to ensure reliability and safety. NIST cultivates trust in technology by developing and deploying standards, tests and metrics that make technology more secure, usable, interoperable and reliable, and by strengthening measurement science. This work is critically relevant to building the public trust of rapidly evolving AI technologies."
In contrast with deterministic rule-based systems, where reliability and safety may be built in and proven by design, AI systems typically make decisions based on data-driven models created by machine learning. Inherent uncertainties need to be characterized and assessed through standardized approaches to assure the technology is safe and reliable. Evaluation protocols must be developed and new metrics are needed to provide quantitative support to a broad spectrum of standards including data, performance, interoperability, usability, security, and privacy.
Watch the video: https://wp.me/p3RLHQ-huZ
Learn more: https://www.nist.gov/topics/artificial-intelligence
Sign up for our insideHPC Newsletter: http://insidehpc.com/newsletter
Transform your Business with AI, Deep Learning and Machine LearningSri Ambati
Video: https://www.youtube.com/watch?v=R3IXd1iwqjc
Meetup: http://www.meetup.com/SF-Bay-ACM/events/231709894/
In this talk, Arno Candel presents a brief history of AI and how Deep Learning and Machine Learning techniques are transforming our everyday lives. Arno will introduce H2O, a scalable open-source machine learning platform, and show live demos on how to train sophisticated machine learning models on large distributed datasets. He will show how data scientists and application developers can use the Flow GUI, R, Python, Java, Scala, JavaScript and JSON to build smarter applications, and how to take them to production. He will present customer use cases from verticals including insurance, fraud, churn, fintech, and marketing.
- Powered by the open source machine learning software H2O.ai. Contributors welcome at: https://github.com/h2oai
- To view videos on H2O open source machine learning software, go to: https://www.youtube.com/user/0xdata
Demystifying Artificial Intelligence: Solving Difficult Problems at ProductCa...Carol Smith
Artificially intelligent systems are becoming part of our everyday lives. This session will answer your questions about artificial intelligence, machine learning, and the ethical conflicts and the implications inherent in these technologies. Topics covered will include: discussions of bias in data; how to focus on the user experience; what is necessary to build a good cognitive computing systems; data needs; levels of accuracy; making safe and secure AI's; and discussions on ethics in AI and our role in leading those conversations. Carol will propose simple models for thinking about these systems and provide time for questions. You will walk away with an awareness of the weaknesses of AI and the knowledge of how these systems work.
Selected by the audience to be presented at ProductCamp Pittsburgh in September 2018
Designing AI for Humanity at dmi:Design Leadership Conference in BostonCarol Smith
As design leaders we must enable our teams with skills and knowledge to take on the new and exciting opportunities that building powerful AI systems bring. Dynamic systems require transparency regarding data provenance, bias, training methods, and more, to gain user’s trust. Carol will cover these topics and challenge us as design leaders, to represent our fellow humans by provoking conversations regarding critical ethical and safety needs.
Presented at dmi:Design Leadership Conference in Boston in October 2018.
UX in the Age of AI: Leading with Design UXPA2018Carol Smith
How can designers improve trust of cognitive systems? What can we do to make these systems transparent? What information needs to be transparent? The biggest challenges inherent with AI will be discussed, specifically the ethical conflicts and the implications for your work, along with the basics of these concepts so that you can strive for making great AI systems.
Data Culture Series - Keynote & Panel - Birmingham - 8th April 2015Jonathan Woodward
Big data. Small data. All data. You have access to an ever-expanding volume of data inside the walls of your business and out across the web. The potential in data is endless – from predicting election results to preventing the spread of epidemics. But how can you use it to your advantage to help move your business forward?
Data is growing exponentially and it’s now possible to mine and unlock insights from data in new and unexpected ways. Empower your business to take advantage of this data by harnessing the rich capabilities of Microsoft SQL Server and the familiarity of Microsoft Office to help organize, analyze, and make sense of your data—no matter the size.
Machine Learning for Non-technical Peopleindico data
Machine learning is one of the most promising and most difficult to understand fields of the modern age. Here are the slides from Slater Victoroff's (CEO of indico) talk at General Assembly Boston for non-technical folks on how to separate the signal from the noise -- stay tuned for the next time he speaks:
https://generalassemb.ly/education/machine-learning-for-non-technical-people
Usama Fayyad talk at Silicon Slopes Technology Summit in Salt Lake City January 31, 2019. The title is "Deploying #AI Technology that Works - #AI Hype vs. Reality: Lessons Learned for Pragmatic AI in the Enterprise. I cover my own version of a brief history of AI and how #BigData is strongly related to making AI work. I cover 5 lessons from the front lines for making AI work in the Enterprise. I conclude with a brief overview of what we are doing at OODA Health, Inc.
Prototyping for Beginners - Pittsburgh Inclusive Innovation Summit 2019Carol Smith
To design for inclusion we often must try out different ideas. In this interactive session you'll learn about all types of prototyping and how to get feedback on your ideas from your users. This session will briefly introduce a variety of prototypes and materials and evaluation methods for early learning.
Participants will have time to build a quick prototype and practice getting feedback on it. We'll cover designing for accessibility and inclusion even at the prototype stage. You'll have the information you need to launch your ideas as early as possible to learn from the experience and improve more quickly.
Presented at the Pittsburgh Inclusive Innovation Summit March 30, 2019 held at Point Park University.
We focus on Invisible Interfaces and their influence on digital experiences. With the advent of 5G creating the foundation for the increased adoption of ‘invisibility’ in our interaction with technology – we’ll discuss what this could mean for the UX and CX industry.
[DSC Europe 22] On the Aspects of Artificial Intelligence and Robotic Autonom...DataScienceConferenc1
Autonomy in targeting is a function that could be applied to any intelligent system, in particular the rapidly expanding array of robotic systems, in the air, on land and at sea – including swarms of small robots. This is an area of significant investment and emphasis for many armed forces, and the question is not so much whether we will see more intelligent robots, but whether and by what means they will remain under human control. Today’s remote-controlled weapons could become tomorrow’s autonomous weapons with just a software upgrade. The central element of any future autonomous weapon system will be the software. Military powers are investing in AI for a wide range of applications10 and significant efforts are already underway to harness developments in image, facial and behavior recognition using AI and machine learning techniques for intelligence gathering and “automatic target recognition” to identify people, objects or patterns. Although not all autonomous weapon systems incorporate AI and machine learning, this software could form the basis of future autonomous weapon systems.
In this deck from the HPC User Forum in Milwaukee, Michael Garris from NIST presents: The National Science & Technology Council ML/AI Initiative.
"AI-enabled systems are beginning to revolutionize fields such as commerce, healthcare, transportation and cybersecurity. It has the potential to impact nearly all aspects of our society including our economy, yet its development and use come with serious technical and ethical challenges and risks. AI must be developed in a trustworthy manner to ensure reliability and safety. NIST cultivates trust in technology by developing and deploying standards, tests and metrics that make technology more secure, usable, interoperable and reliable, and by strengthening measurement science. This work is critically relevant to building the public trust of rapidly evolving AI technologies."
In contrast with deterministic rule-based systems, where reliability and safety may be built in and proven by design, AI systems typically make decisions based on data-driven models created by machine learning. Inherent uncertainties need to be characterized and assessed through standardized approaches to assure the technology is safe and reliable. Evaluation protocols must be developed and new metrics are needed to provide quantitative support to a broad spectrum of standards including data, performance, interoperability, usability, security, and privacy.
Watch the video: https://wp.me/p3RLHQ-huZ
Learn more: https://www.nist.gov/topics/artificial-intelligence
Sign up for our insideHPC Newsletter: http://insidehpc.com/newsletter
Transform your Business with AI, Deep Learning and Machine LearningSri Ambati
Video: https://www.youtube.com/watch?v=R3IXd1iwqjc
Meetup: http://www.meetup.com/SF-Bay-ACM/events/231709894/
In this talk, Arno Candel presents a brief history of AI and how Deep Learning and Machine Learning techniques are transforming our everyday lives. Arno will introduce H2O, a scalable open-source machine learning platform, and show live demos on how to train sophisticated machine learning models on large distributed datasets. He will show how data scientists and application developers can use the Flow GUI, R, Python, Java, Scala, JavaScript and JSON to build smarter applications, and how to take them to production. He will present customer use cases from verticals including insurance, fraud, churn, fintech, and marketing.
- Powered by the open source machine learning software H2O.ai. Contributors welcome at: https://github.com/h2oai
- To view videos on H2O open source machine learning software, go to: https://www.youtube.com/user/0xdata
Demystifying Artificial Intelligence: Solving Difficult Problems at ProductCa...Carol Smith
Artificially intelligent systems are becoming part of our everyday lives. This session will answer your questions about artificial intelligence, machine learning, and the ethical conflicts and the implications inherent in these technologies. Topics covered will include: discussions of bias in data; how to focus on the user experience; what is necessary to build a good cognitive computing systems; data needs; levels of accuracy; making safe and secure AI's; and discussions on ethics in AI and our role in leading those conversations. Carol will propose simple models for thinking about these systems and provide time for questions. You will walk away with an awareness of the weaknesses of AI and the knowledge of how these systems work.
Selected by the audience to be presented at ProductCamp Pittsburgh in September 2018
Designing AI for Humanity at dmi:Design Leadership Conference in BostonCarol Smith
As design leaders we must enable our teams with skills and knowledge to take on the new and exciting opportunities that building powerful AI systems bring. Dynamic systems require transparency regarding data provenance, bias, training methods, and more, to gain user’s trust. Carol will cover these topics and challenge us as design leaders, to represent our fellow humans by provoking conversations regarding critical ethical and safety needs.
Presented at dmi:Design Leadership Conference in Boston in October 2018.
UX in the Age of AI: Leading with Design UXPA2018Carol Smith
How can designers improve trust of cognitive systems? What can we do to make these systems transparent? What information needs to be transparent? The biggest challenges inherent with AI will be discussed, specifically the ethical conflicts and the implications for your work, along with the basics of these concepts so that you can strive for making great AI systems.
Data Culture Series - Keynote & Panel - Birmingham - 8th April 2015Jonathan Woodward
Big data. Small data. All data. You have access to an ever-expanding volume of data inside the walls of your business and out across the web. The potential in data is endless – from predicting election results to preventing the spread of epidemics. But how can you use it to your advantage to help move your business forward?
Data is growing exponentially and it’s now possible to mine and unlock insights from data in new and unexpected ways. Empower your business to take advantage of this data by harnessing the rich capabilities of Microsoft SQL Server and the familiarity of Microsoft Office to help organize, analyze, and make sense of your data—no matter the size.
Machine Learning for Non-technical Peopleindico data
Machine learning is one of the most promising and most difficult to understand fields of the modern age. Here are the slides from Slater Victoroff's (CEO of indico) talk at General Assembly Boston for non-technical folks on how to separate the signal from the noise -- stay tuned for the next time he speaks:
https://generalassemb.ly/education/machine-learning-for-non-technical-people
Usama Fayyad talk at Silicon Slopes Technology Summit in Salt Lake City January 31, 2019. The title is "Deploying #AI Technology that Works - #AI Hype vs. Reality: Lessons Learned for Pragmatic AI in the Enterprise. I cover my own version of a brief history of AI and how #BigData is strongly related to making AI work. I cover 5 lessons from the front lines for making AI work in the Enterprise. I conclude with a brief overview of what we are doing at OODA Health, Inc.
Prototyping for Beginners - Pittsburgh Inclusive Innovation Summit 2019Carol Smith
To design for inclusion we often must try out different ideas. In this interactive session you'll learn about all types of prototyping and how to get feedback on your ideas from your users. This session will briefly introduce a variety of prototypes and materials and evaluation methods for early learning.
Participants will have time to build a quick prototype and practice getting feedback on it. We'll cover designing for accessibility and inclusion even at the prototype stage. You'll have the information you need to launch your ideas as early as possible to learn from the experience and improve more quickly.
Presented at the Pittsburgh Inclusive Innovation Summit March 30, 2019 held at Point Park University.
We focus on Invisible Interfaces and their influence on digital experiences. With the advent of 5G creating the foundation for the increased adoption of ‘invisibility’ in our interaction with technology – we’ll discuss what this could mean for the UX and CX industry.
[DSC Europe 22] On the Aspects of Artificial Intelligence and Robotic Autonom...DataScienceConferenc1
Autonomy in targeting is a function that could be applied to any intelligent system, in particular the rapidly expanding array of robotic systems, in the air, on land and at sea – including swarms of small robots. This is an area of significant investment and emphasis for many armed forces, and the question is not so much whether we will see more intelligent robots, but whether and by what means they will remain under human control. Today’s remote-controlled weapons could become tomorrow’s autonomous weapons with just a software upgrade. The central element of any future autonomous weapon system will be the software. Military powers are investing in AI for a wide range of applications10 and significant efforts are already underway to harness developments in image, facial and behavior recognition using AI and machine learning techniques for intelligence gathering and “automatic target recognition” to identify people, objects or patterns. Although not all autonomous weapon systems incorporate AI and machine learning, this software could form the basis of future autonomous weapon systems.
Artificial Intelligence and Machine Learning Aditya Singh
Presented By JBIMS Marketting Batch (2017-2020).
Application Artificial Intelligence in MIS(Management Information System). Presented By Trilok Prabhakaran , Aditya Singh , Shashi Yadav, Vaibhav Rokade. Presentation have live cases of two different industry.
Presentation about AI and Libraries. Why should libraries follow technology and be the main information provider and how innovating libraries can reach the AI audience and the increased need for data and information.
We have critically evaluated how AI will shape integration use cases, their feasibility, and timelines. Emerging Technology Analysis Canvas (ETAC), a framework built to analyze emerging technologies, is the methodology of our study.
We observe that AI can significantly impact integration use cases and identify 13 AI-based use case classes for integration. Points to note include:
Enabling AI in an enterprise involves collecting, cleaning up, and creating a single representation of data as well as enforcing decisions and exposing data outside, each of which leads to many integration use cases. Hence, AI indirectly creates demand for integration.
AI needs data, which in some cases lead to significant competitive advantages. The need to collect data would drive vendors to offer most AI products in the cloud through APIs.
Due to lack of expertise and data, custom AI model building will be limited to large organizations. It is hard for small and medium size organization to build and maintain custom models.
20240104 HICSS Panel on AI and Legal Ethical 20240103 v7.pptxISSIP
20240103 HICSS Panel
Ethical and legal implications raised by Generative AI and Augmented Reality in the workplace.
Souren Paul - https://www.linkedin.com/in/souren-paul-a3bbaa5/
Event: https://kmeducationhub.de/hawaii-international-conference-on-system-sciences-hicss/
Modeling Electronic Health Records with Recurrent Neural NetworksJosh Patterson
Time series data is increasingly ubiquitous. This trend is especially obvious in health and wellness, with both the adoption of electronic health record (EHR) systems in hospitals and clinics and the proliferation of wearable sensors. In 2009, intensive care units in the United States treated nearly 55,000 patients per day, generating digital-health databases containing millions of individual measurements, most of those forming time series. In the first quarter of 2015 alone, over 11 million health-related wearables were shipped by vendors. Recording hundreds of measurements per day per user, these devices are fueling a health time series data explosion. As a result, we will need ever more sophisticated tools to unlock the true value of this data to improve the lives of patients worldwide.
Deep learning, specifically with recurrent neural networks (RNNs), has emerged as a central tool in a variety of complex temporal-modeling problems, such as speech recognition. However, RNNs are also among the most challenging models to work with, particularly outside the domains where they are widely applied. Josh Patterson, David Kale, and Zachary Lipton bring the open source deep learning library DL4J to bear on the challenge of analyzing clinical time series using RNNs. DL4J provides a reliable, efficient implementation of many deep learning models embedded within an enterprise-ready open source data ecosystem (e.g., Hadoop and Spark), making it well suited to complex clinical data. Josh, David, and Zachary offer an overview of deep learning and RNNs and explain how they are implemented in DL4J. They then demonstrate a workflow example that uses a pipeline based on DL4J and Canova to prepare publicly available clinical data from PhysioNet and apply the DL4J RNN.
Building Deep Learning Workflows with DL4JJosh Patterson
In this session we will take a look at a practical review of what is deep learning and introduce DL4J. We’ll look at how it supports deep learning in the enterprise on the JVM. We’ll discuss the architecture of DL4J’s scale-out parallelization on Hadoop and Spark in support of modern machine learning workflows. We’ll conclude with a workflow example from the command line interface that shows the vectorization pipeline in Canova producing vectors for DL4J’s command line interface to build deep learning models easily.
Georgia Tech cse6242 - Intro to Deep Learning and DL4JJosh Patterson
Introduction to deep learning and DL4J - http://deeplearning4j.org/ - a guest lecture by Josh Patterson at Georgia Tech for the cse6242 graduate class.
Hadoop Summit 2014 - San Jose - Introduction to Deep Learning on HadoopJosh Patterson
As the data world undergoes its cambrian explosion phase our data tools need to become more advanced to keep pace. Deep Learning has emerged as a key tool in the non-linear arms race of machine learning. In this session we will take a look at how we parallelize Deep Belief Networks in Deep Learning on Hadoop’s next generation YARN framework with Iterative Reduce. We’ll also look at some real world examples of processing data with Deep Learning such as image classification and natural language processing.
MLConf 2013: Metronome and Parallel Iterative Algorithms on YARNJosh Patterson
Online learning techniques, such as Stochastic Gradient Descent (SGD), are powerful when applied to risk minimization and convex games on large problems. However, their sequential design prevents them from taking advantage of newer distributed frameworks such as Hadoop/MapReduce. In this session, we will take a look at how we parallelize parameter estimation for linear models on the next-gen YARN framework Iterative Reduce and the parallel machine learning library Metronome. We also take a look at non-linear modeling with the introduction of parallel neural network training in Metronome as well.
Have you ever been recommended a friend on Facebook? Or an item you might be interested in on Amazon? If so then you’ve benefitted from the value of recommendation systems. Recommendation systems apply knowledge discovery techniques to the problem of making recommendations that are personalized for each user. Recommendation systems are one way we can use algorithms to help us sort through the masses of information to find the “good stuff” in a very personalized way.
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...John Andrews
SlideShare Description for "Chatty Kathy - UNC Bootcamp Final Project Presentation"
Title: Chatty Kathy: Enhancing Physical Activity Among Older Adults
Description:
Discover how Chatty Kathy, an innovative project developed at the UNC Bootcamp, aims to tackle the challenge of low physical activity among older adults. Our AI-driven solution uses peer interaction to boost and sustain exercise levels, significantly improving health outcomes. This presentation covers our problem statement, the rationale behind Chatty Kathy, synthetic data and persona creation, model performance metrics, a visual demonstration of the project, and potential future developments. Join us for an insightful Q&A session to explore the potential of this groundbreaking project.
Project Team: Jay Requarth, Jana Avery, John Andrews, Dr. Dick Davis II, Nee Buntoum, Nam Yeongjin & Mat Nicholas
Explore our comprehensive data analysis project presentation on predicting product ad campaign performance. Learn how data-driven insights can optimize your marketing strategies and enhance campaign effectiveness. Perfect for professionals and students looking to understand the power of data analysis in advertising. for more details visit: https://bostoninstituteofanalytics.org/data-science-and-artificial-intelligence/
Adjusting primitives for graph : SHORT REPORT / NOTESSubhajit Sahu
Graph algorithms, like PageRank Compressed Sparse Row (CSR) is an adjacency-list based graph representation that is
Multiply with different modes (map)
1. Performance of sequential execution based vs OpenMP based vector multiply.
2. Comparing various launch configs for CUDA based vector multiply.
Sum with different storage types (reduce)
1. Performance of vector element sum using float vs bfloat16 as the storage type.
Sum with different modes (reduce)
1. Performance of sequential execution based vs OpenMP based vector element sum.
2. Performance of memcpy vs in-place based CUDA based vector element sum.
3. Comparing various launch configs for CUDA based vector element sum (memcpy).
4. Comparing various launch configs for CUDA based vector element sum (in-place).
Sum with in-place strategies of CUDA mode (reduce)
1. Comparing various launch configs for CUDA based vector element sum (in-place).
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Subhajit Sahu
Abstract — Levelwise PageRank is an alternative method of PageRank computation which decomposes the input graph into a directed acyclic block-graph of strongly connected components, and processes them in topological order, one level at a time. This enables calculation for ranks in a distributed fashion without per-iteration communication, unlike the standard method where all vertices are processed in each iteration. It however comes with a precondition of the absence of dead ends in the input graph. Here, the native non-distributed performance of Levelwise PageRank was compared against Monolithic PageRank on a CPU as well as a GPU. To ensure a fair comparison, Monolithic PageRank was also performed on a graph where vertices were split by components. Results indicate that Levelwise PageRank is about as fast as Monolithic PageRank on the CPU, but quite a bit slower on the GPU. Slowdown on the GPU is likely caused by a large submission of small workloads, and expected to be non-issue when the computation is performed on massive graphs.
2. Central Thesis for “What is Artificial Intelligence?”
“Artificial
Intelligence” is a
term for algorithms
that increase user
productivity
•Quite artificial, no where close
to being “alive”
•Historically we get over-excited
about AI
•Because its an existential threat
•Accelerated in our minds by our
fixation on social media fueled
over-marketed narratives
•Can be hard to define
•Goal posts tend to move
•Much like Data Science
3. History and Definitions
Of Artificial Intelligence
One of the best books written on the subject of AI (if not the best) is
Stuart Russell and Peter Norvig’s Artificial Intelligence: A Modern Approach.
I can’t recommend this book enough for you to get a more complete idea of the depth and history of AI.
4. The Study of Intelligence
•Study of intelligence formally initiated in 1956 at Dartmouth
•Yet is at least 2,000 years old.
•The field is based on understanding intelligent entities and studying
topics such as these:
•Seeing
•Learning
•Remembering
•Reasoning
•These topics are components of what we’d consider intelligent
function
•to the capacity we have to understand intelligence
5. Building Blocks of Intelligent Study
•Philosophy (400 BC)
•Philosophers began to suggest the
mind as a mechanical machine
that encodes knowledge in some
form inside the brain.
•Mathematics
•Mathematicians developed the
core ideas of working with
statements of logic along with the
groundwork for reasoning about
algorithms.
•Psychology
•This field of study is built on the
ideas that animals and humans
have a brain that can process
information.
•Computer science
•Practitioners came up with
hardware, data structures, and
algorithms to support reverse
engineering basic components of
the brain.
6. Defn. “AI”
•Methods considered AI
•Linear modeling
•Neural Networks
•Random Forrests
•Expert Systems
•Rule Bases
Algorithms that automate parts, or all of, tasks
Russell and Norvig’s book on AI:
The intellectual establishment, by and large, preferred to believe that a
“machine can never do X.”
Problem is, AI researchers have
systematically responded by
demonstrating one X after
another.
7. Why is AI Popular Now? Deep Learning
•Three major contributors are
driving interest in AI today:
•The big jump in computer-vision
technology in the late 2000s
(Hinton’s team, others)
•The big data wave of the early
2010s
•Advancements in applications of
deep learning by top technology
firms
8. Quick History of Neural Networks
•Rough approximation of biological neuron
•1950s saw perceptron developed in
hardware
•Changes in activation function allowed for
non-linear functions
•Multi-layer perceptron becomes more
modern version of “neural networks”
12. Practical Use Cases of AI Today
•Computer Vision Applications
•Is there damage on this asset?
(Insurance)
•Sensor Applications
•Which machine on this assembly
line needs maintenance?
•(Automotive, manufacturing)
•Control and Planning
•Robots that navigate a warehouse
(logistics, retail)
•Machines today are getting
better at
•Making sense of images from
cameras
•Working w sensor data
•Making a plan to interact with the
world based on vision and sensors
13. •Developed Deep Reinforcement
Learning techniques to play
different types of games
•Atari
•Go
•Developed system that beat the
world’s best Go player
•AlphaGo beat Lee Sedol the world
champion in a five game
tournament
“Alpha Go Zero.”
Later on deepmind developed a
new variant that was able to
defeat Alpha Go at its own game
only 40 days later
Deepmind
14. IBM Watson Cancer
Reality check:
• uses expert system (programmed by
Memorial Sloan Kettering) to
recommend treatments
• uses NLP to discover and summarize
relevant studies, research, etc., to
back up recommendation
• does NOT automatically learn from
patient record data
• apparently uninterpretable despite
being an expert system!
Pros
• high concordance with common practice --
at hospitals similar to MSK
• can make medical decision making more
efficient
• can improve patient outcomes in regions
with limited resources, no experts
Cons
• recommendations are biased toward MSK
doctors, patient population
• cannot take, e.g., local regulations into
account (maybe recommended treatment not
covered by insurance)
• human-in-the-loop training is slow, labor
intensive; makes adaptation to local
population difficult
15. Why Does the term AI
attract so much Attention?
•Existential
•Will it take my job?
•Will robots take over the world?
•More subtly: “can it answer all of my
questions?”
•So often people expect technology to “give
them the answers”
•Being human is about the intuition to know
which questions to ask
•And leveraging the right technology to answer
these questions
17. Irrational Fear in a Non-Linear World
•When we start out with existential threats on our way of life
•Coupled with irrational exuberance in marketing narratives
•We arrive at some crazy end games, which tend to not be realistic as its hard
to project outcomes in the non-linear world (that we live in)
•Ever notice that in horror movies the narrative always requires the
characters to make irrational choices to move forward?
•“how about we *not* land on that planet?”
•Society tends to self adjust in emergent fashion
•Lots of agents make small local decisions to adapt to changing environments
•Tends to make global changes in the system in ways that are hard to predict
18. Also: Sort of hard to regulate games of linear
algebra and methods of optimization
Sorry, Elon.
20. Modern AI Marketing Narratives
•Narratives such as
•Requires a phd to do anything related to
“AI” (“expert gating”)
•Only top 5 tech shops can hire anyone
good at AI, NFL-level salaries
•Techniques are impossible to fathom,
basically
•A lot of hype
21. The Hype Cycle
•Previous hype cycles:
•cloud, smart grid, big data
•Why is AI different than most hype cycles?
•Existential threat
•Companies tack on the big terms of the cycle to get
marketing and funding
•Most have no real claim to the term they tack on
•“AI spreadsheets”, “AI calendars”, “AI for HR”, etc
•What do they really mean?
•Most of the time: “we use linear modeling in our product in some
tangential way”
22. All of This Has Happened Before
Periods of interest have been the result of the sector being unrealisti‐
cally overhyped followed by a cycle of predictably underwhelming
results.
AI Winter I: (1974–1980). The lead-up to the first AI winter saw
machine translation fail to live up to the hype. Connectionism (neural
networks) interest waned in the 1970s, and speech understanding
research overpromised and underdelivered.
AI Winter II: Late 1980s. In the late 1980s and early 1990s, there was
overpromotion of technologies such as expert systems and LISP
machines, both of which failed to live up to expectations. The
Strategic Computing Initiative canceled new spending at the end of
this cycle. The fifth-generation computer also failed to meet its goals.
23. AI, Experts, and Commodity
•Hadoop, 2009: only a few experts at the top shops can do this
(Google, Yahoo, FB)
•2017: Hadoop becomes commodotized
•Although the narrative is that AI-experts are compensated like NFL
stars today, and that only the big-tech co’s can get them…
•Reality is that it will get commoditized through the combined forces of
tooling, open source, integrated AI-apps
•Along a long enough timeline, all technology converges on “Big 6 Consultant”
24. Post Trough AI-Technologies
We’ve seen this with the following:
•Informatics
•Machine learning
•Knowledge-based systems
•Business rules management
•Cognitive systems
•Intelligent systems
•Computational intelligence
The name change might be partly because they consider their field to be
fundamentally different from AI.
25. The Red Queen Always Wins
•We Survived the spreadsheet
from the 90’s
•Radiologists will survive AI
today
Automation will shift and evolve
the workforce as it has for the
past 1000 years
“My dear, here we must run as fast as we can, just to stay in
place. And if you wish to go anywhere you must run twice as fast
as that.”
27. A Style Guide for Writing About AI
•Don’t talk about “AI” as if it is a noun
•Its not
•Don’t take anything Elon Musk says about AI too seriously
•He’s great at cars and rockets, but for other things --- Its marketing,
Editor's Notes
The study and application of AI techniques we see today are based on these fundamentals. We typically see the study of AI broken into a focus on either behaving or thinking in simulated intelligent systems.
In 2006, Geoff Hinton and his team at the University of Toronto published a key paper on Deep Belief Networks (DBNs).2 This provided the industry with a spark of creativity on what could possibly improve the state of the art. We’ve seen a tsunami of deep learning publications at top journals over the succeeding decade.
Go was previously too hard for the same techniques that solved checkers and chess (A*-variants)
This down period is referred to as an “AI winter” and involves cuts in academic research funding, reduced venture capital interest, and stigma in the marketing realm around anything connected to the term “artificial intelligence.”
2009: only a few experts at the top shops can do this (Google, Yahoo, FB)
We did it at TVA
Cloudera comes along, begins the commodotization process
Top end people in distributed systems see this
Ex: Joe Hellerstein co-founds Trifacta
2017: Hadoop becomes commodotized
Not nearly as exciting as it used to be
Table stakes in most shops