COLUMN # 1
DR. QIN (TIM) SHENG
- MEMBER & ADVISOR-
THE GOOD US AI GROUP
BS, MS, Ph.D
Nanjing University, China
Nanjing University, China
University of Cambridge, UK
Post Doc. UCL, London
Ph.D in Mathematics 1990
BAYLOR UNIVERSITY MATHEMATICS PROFESSOR
Computational and Applied Mathematics
BAYLOR DEPARTMENT OF MATHEMATICS
BAYLOR UNIVERSITY CENTER FOR ASTROPHYSICS, SPACE & ENGINEERING RESEARCH
EDITOR IN CHIEF
TAYLOR AND FRANCIS
COMPREHENSIVE BIO FOR DR. TIM SHENG:
THREE PERSON LEADERSHIP TEAM FOR GOODUSAI.COM
1. DR. QIN (TIM) SHENG
- MEMBER & ADVISOR-
THE GOOD US AI GROUP
2. RUSS (RUSSELL)
- MEMBER & ADVISOR-
THE GOOD US AI GROUP
3. JOHN (JOHNNY) E. MILLER
-- CO-FOUNDER --
- MEMBER & ADVISOR -
THE GOOD US AI GROUP
BE SURE TO REVIEW THE AI CONTENT IN COLUMNS 1, 2, AND 3.
NEW AI CONTENT IS ADDED TO THIS WEB SITE WEEKLY.
COLUMN # 1
Artificial Intelligence (AI) systems will play a major role in every aspect of life on earth. It will change the world we live in. AI will improve the speed, accuracy, efficiency, and safety of human decision making. However, it will carry the risks of some unintended or unplanned results. In fact, AI may turn out to be a greater world problem than any of the future concerns, such as: nuclear security, climate change, civil unrest, chemical- biological warfare, terrorism, cyber-security, asteroid impact, loss of coral and plankton, alien visitors, locus invasion destroying food supply, power grid destruction, ozone depletion, loss of internet, poverty, hunger, lowering water supply, air pollution, sanitation, water pollution, nuclear waste, or de-forestation, , lack of education, loss of rain forests, etc, etc..
Is it possible that AI will be an existential event once it reaches the human level and beyond?
MAIN THRUST OF THIS WEB SITE
THE MAIN THRUST OF THIS WEB SITE IS NOT THE DARK / EVIL SIDE OF AI BUT RATHER RESEARCHING, EDUCATING, TEACHING, PROMOTING, AND APPLYING: (1) SAFETY; (2) INTEGRITY; (3) HONESTY; (4) ETHICS; (5) ZERO BIAS; (6) EMPATHY; (7) RESPONSIBILITY; AND (8) SOCIAL BENEFIT WITH REGARD TO ALL ASPECTS OF GOOD UNITED STATES ARTIFICIAL INTELLIGENCE INCLUDING (BUT NOT LIMITED TO) MACHINE LEARNING, DEEP LEARNING, ROBOTICS, ROBOTIC VISION, NATURAL LANGUAGE PROCESSING, VOICE RECOGNITION, SPEECH RECOGNITION, FACIAL RECOGNITION, DRONES, ARTIFICIAL NEURAL NETWORKS, INTERNET OF THINGS, AND OTHER ASPECTS OF AI,
PROGRESS ALWAYS COMES
AT A COST
Progress always comes at a cost. Paper fundamentally changed the way information was stored and distributed, but its production contributes to deforestation.
Industrialization increased our standard of living, but has led to much pollution and arguably, even some social ills. The benefits brought by the internet are too many to mention, yet viral misinformation, vast erosion of privacy, and the diminishing patience of society as a whole were all unintended consequences. Not even medicine is free from side effects. This should come as no surprise because hindsight is always twenty-twenty. Rarely at the time of invention is a creator the best judge of how their system will be used, or truly knows what good or harm will come of it. Understanding this, as technologists, we ought to give pause and reflect deeply before taking on a project.
SOME BASIC DEFINITIONS:
A. Artificial Intelligence:
1. Artificial Intelligence (AI) is a subset of data science. Data science is a subset of computer science.
AI (as a subset of data science) is the ability of a machine and/or of computer software programs to find, assemble, process, calculate, translate, think, reason, learn, problem-solve, identify risks, create speech recognition, develop human-like speech generation, incorporate feedback, remember, reduce errors, exercise continuous improvement, and potentially act (to some extent). It could be performing in a manner that some people would consider it as ... "intelligent."
AI is a field of science covering how computers can make decisions as well as (and sometimes better than) humans. AI needs to be able to understand humans. Machine learning (ML), a subset of AI, refers to the popular modern-day techniques for creating software that learns from data. It involves learning how to carry out a task from data without being programed to carry out the task. For example, facial recognition is machine learning. Applications of machine learning normally use a neural network. A neural network is a computer system mimicking the human brain.
An example of deep learning is speech recognition.
Included in the AI definition is the difference between weak/narrow AI and strong/broad AI. Strong AI genuinely simulates human reasoning while weak AI is just focusing on getting a system to work without simulating human cognitive
behavior (thinking the way a human would think). Today most of the AI work falls somewhere between weak AI and strong AI. Let's call that Mid-Range AI.
Major Artificial Intelligence Components:
The following items are some of the major components that make up AI:
1. Machine Learning (ML) - Machine learning is an application of AI that provides systems the ability to automatically learn and improve from experience without being explicitly programmed. Machine learning focuses on the development of computer programs that can access data and use it to learn for themselves.
It allows systems to learn from data, identify repeating patterns and making decisions without explicit instructions or human intervention. It enables entities/humans to save time and resources.
2. Big Data - According to Oracle: "Put simply, big data is larger, more complex data sets, especially from new data sources. These data sets are so voluminous that traditional data processing software just can't manage them." It includes structured and unstructured data. There is a growing belief that data is more valuable than oil. Before data can be used in an AI use case it must be cleaned, prepared, and labeled. Data is the heart of AI. Data is the new electricity. Cleaning data involves eliminating duplicates, deleting extraneous data, and working with data by humans to get it ready for use with the appropriate AI algorithm for a use case.
3. Artificial Neural Networks - Artificial neural networks are one of the main tools used in machine learning. As the “neural” part of their name suggests, they are brain-inspired systems which are intended to replicate the way that we humans learn. Neural networks consist of input and output layers, as well as (in most cases) a hidden layer consisting of units that transform the input into something that the output layer can use. They are excellent tools for finding patterns which are far too complex or numerous for a human programmer to extract and teach the machine to recognize.
4. Robotics (including Robotic Vision) - From Wikipedia, the free encyclopedia, it states: “Robotics is an interdisciplinary branch of engineering and science that includes mechanical engineering, electronic engineering, information engineering, computer science, and others. Robotics deals with the design, construction, operation, and use of robots, as well as computer systems for their control, sensory feedback, and information processing. These technologies are used to develop machines that can substitute for humans and replicate human actions. Robots can be used in many situations and for lots of purposes, but today many are used in dangerous environments (including bomb detection and deactivation), manufacturing processes, or where humans cannot survive (e.g. in space). Robots can take on any form but some are made to resemble humans in appearance. This is said to help in the acceptance of a robot in certain replicative behaviors usually performed by people. Such robots attempt to replicate walking, lifting, speech, cognition, and basically anything a human can do. Many of today's robots are inspired by nature, contributing to the field of bio-inspired robotics.”
From Wikipedia, the free encyclopedia it states: "Robot vision or a Vision Guided Robot (VGR) System is basically a robot fitted with one or more cameras used as sensors to provide a secondary feedback signal to the robot controller to more accurately move to a variable target position. VGR is rapidly transforming production processes by enabling robots to be highly adaptable and more easily implemented, while dramatically reducing the cost and complexity of fixed tooling previously associated with the design and set up of robotic cells, whether for material handling, automated assembly, agricultural applications, life sciences, and more.
5. Facial Recognition - Now here is a controversial topic. A facial recognition system is a technology capable of identifying or verifying a person from a digital image or a video frame from a video source. There are multiple methods in which facial recognition systems work, but in general, they work by comparing selected facial features from given image with faces within a database. It is also described as a biometric AI In based application that can uniquely identify a person by analyzing patterns based on the person's facial textures and shape.
6. Speech Recognition - Speech recognition is the ability of a machine or program to identify words and phrases in a spoken language and convert them to a machine-readable format. Rudimentary speech recognition software has a limited vocabulary of words and phrases, and it may only identify these if they are spoken very clearly.
7. Voice Recognition - The ability of hardware and software to recognize the voice of a person (as unique as a fingerprint). It differs from speech recognition 's ability to understand words - not the identity of a person.
8. Deep Learning - Networks capable of learning from unstructured data when the network is unsupervised.
9. Internet of Things (IoT) - This is Computer components embedded in everyday objects with capabilities of sending and receiving data and performing other activities.
The oil & gas industry is especially ripe for the application of IoT with potential benefits in safety, the environment, reduced spills, reduced carbon emissions, improved monitoring, improved drilling strategy, cost reduction, time savings, and increased revenue.
Here below is a good high level description of how IoT works, as so stated in LEVEREGE's ebook "An Introduction to Internet of Things:"
"'An IoT system consists of sensors/devices which “talk” to the cloud through some kind of connectivity. Once the data gets to the cloud, software processes it and then might decide to perform an action, such as sending an alert or automatically adjusting the sensors/devices without the need for the human user .
But if user input is needed or if the user simply wants to check in on the system, a user interface allows them to do so . Any adjustments or actions that the user makes are then sent in the opposite direction through the system: from the user interface, to the cloud, and back to the sensoIrs/ devices to make some kind of change."
10. Drones - An unmanned aerial vehicle (UAV), commonly known as a drone, is an aircraft without a human pilot aboard. UAVs are a component of an unmanned aircraft system (UAS); which include a UAV, a ground-based controller, and a system of communications between the two. The flight of UAVs may operate with various degrees of autonomy: either under remote control by a human operator or autonomously by onboard computers.
Compared to manned aircraft, UAVs were originally used for missions too "dull, dirty or dangerous” for humans. While they originated mostly in military applications, their use is rapidly expanding to commercial, scientific, recreational, agricultural, and other applications, such as policing, peacekeeping, surveillance, product deliveries, aerial photography, agriculture, smuggling, and drone racing. Civilian UAVs now vastly outnumber military UAVs, with estimates of over a million sold by 2017, so they can be seen as an early commercial application of autonomous things, to be followed by the autonomous car and home robot.
11. AI Algorithms are mathematical instructions that provide step by step procedures for calculations. It is kind-of like a set of step-by-step instructions.
12. AUGUMENTED AI:
At this point (all things considered), it seems as though one of the most promising methods of beneficially, expeditiously, and effectively making the best use of artificial intelligence (AI) is to use it to augment the efforts of the user. AI should only be used to assist the human worker. It should not be used to replicate and replace him/her. There’s nothing anywhere that dictates in any way whatsoever that AI can only be used by the user in an automatic manner thus replacing humans. Augmented Artificial Intelligence is the best way to use AI. Augmented Artificial Intelligence refers to the result of skillfully blending human and machine intelligence.
13. There are, of course, other aspects of AI that are in various stapes of creation and development.
FIRST THINGS FIRST:
OK - first things first. We start first with data science (DS), according to Russ Rankin and Dr. Stephen Gardener - Baylor
Magazine - Fall Issue 2018 - "Data Drive" - "Data science is a broadly interdisciplinary field that draws heavily on statistics and computer science and has applications in business, engineering, medicine, law, education, sociology, political science, and other disciplines. Data science is the foundational field for development of AI, robotics, and other technologies that can mimic or transcend many aspects of human intelligence."
This is a very useful video on medical and other AI applications.
ROCK & ROLL:
The term "Artificial intelligence" dates back to about the mid-1950's. About 1956. (The mid 1950's were famous for the development of creative "Rock & Roll" music. Remember Chuck Berry, Little Richard, Jerry Lee Lewis, Bo Diddley, Elvis, Buddy Holly, Johnny Cash, Connie Francis, etc, etc? They also were bursting in on the music scene in the mid-1950's. So, it was only fitting for that time period to also be the time of the start of development of creative AI. AI was first named and identified
AS THE INTENDED AI CIRCUMSTANCE, AI SHOULD BE BENEFICIAL, ETHICAL, SAFE, GUIDED BY COMMON SENSE, RESPONSIBLE, UNBIASED, AND OPERATE WITH COMPLETE INTEGRITY.
AI duplicates the human thought process and behavior. It should act in a beneficial human - like way that is safe, intelligent, brilliant, rational, unbiased, guided by common sense, responsible, reasonable, timely, and (of course) ethical with integrity.
THE TIME IS RIGHT:
AI is possible nowadays due to:
(1) giant increases in computational capabilities; (2) huge growths in data (big data); (3) focusing on specific / unique problem issues to be addressed; (4) being able to timely convert those unique issues / problems into targeted knowledge engineering rules (algorithms) so that AI systems can learn; (5) efficiently plugging the resulting rules into appropriate AI systems; and (6) allowing AI systems to learn the rules automatically.
"AI has the potential to boost rates of profitability by an average of 38 percentage points and could lead to an economic boost of US$14 trillion in additional gross value added (GVA) by 2035."
Accenture further stated in the same post: "Artificial Intelligence could double annual economic growth rates by 2035 by changing the nature of work and spawning a new relationship between both human and machine.
The impact of AI technologies on business is projected to boost labor productivity by up to 40 percent by fundamentally changing the way work is done and reinforcing the role of people to drive growth in business."
Accenture additionally stated in the same post: "AI at its core combines intelligent technology with human ingenuity. Probably the most dramatic change that AI is driving for people is the nature of work. And collaborative intelligence is very important because this is about how we take the best capabilities of a person and the best capabilities of software or machine or AI and put that together to create a new type of job. The
ideal is that there’s something missing
from the discussion around AI today
and it’s what’s happening in the middle between humans and machine. We tend to think about what humans are good at, which is communication, emotion, imagination, generalization, and so forth. And then you think about what machines are good at -- memorization, prediction, transaction, and more. There’s a lot of opportunity in that area in the middle and the discussion that has been missing is about how do you put humans and technology together to do new things and create new capability together.
Human and AI technology are connected and will change the way we work and solve business problems. AI will also re-imagine society, in how we use it to solve social issues, and in how we apply it responsibly. Responsible AI is necessary for AI to come together across business, people and society
and be successful.
There was no better place to talk
about the last 25 years of technology,
and the next 25 years when we will see
epic change driven by AI. The Information Age has brought us to where we are today. The next period will define our future and change the world."
"BENEFITS OF AI CAN BE ENORMOUS:
1. Humans will be safer during disasters;
2. Better care for the aging;
3. More independence for the elderly;
4. Reduction of traffic accidents;
5. Could become a tightly coupled cognitive unit with humans;
6. AI can augment our brains;
7. AI can augment our bodies;
8. It can make humans better at everything;
9. Great improvements to orthotics;
10. Wide applications to exoskeletons;
11. Improvement to human senses (vision, hearing, etc).
12. AI can help to solve climate change;
13. AI can quickly review millions of pages of documents;
14. AI can help to cure diseases;
15. Anticipated to greatly assist with space travel;
16. Improved internet access;
17. Use of robots powered by AI in homes;
18. Wide-spread embedding of AI;
19. Prevent environmental catastrophes;
20. Creation of more free time for humans to be more creative.
21. Improves ability to help customers.
22. Improves e-commerce.
23. Chat boxes.
24. Better decisions and fewer mistakes.
25. Assisting humans with augmented artificial
26. Help compose music.
27. Improved gaming and media.
28. Help in the hospitality and restaurant business.
29. Improve the back office.
30. Bank loan determination.
31. Fraud determination.
32. Assisting in ancestry research.
33. AI serving as an enabler.
34. Potential wide-spread access and use of open-source AI world-wide.
35. Airlines can save money annually on fuel savings in-flight and during take-off and landing (such as Quantas Airlines saving $40M in 2018) due to AI).
36. Assisting air traffic control to improve safety, efficiency, and decision making.
37. Use of AI and ML in the oil and gas business to address safety, exploration, discovery, drilling, production, analysis, and distribution.
POTENTIAL AI RISKS AND THREATS
The numerous legal and other risks that arise out of the use of AI including: machine learning ("ML"), deep learning (DL), robotics, robot vision (RV), internet of things (IoT), signal processing (SP), neural networks (NN), natural language (NL) and unstructured data (UD) or Dark Data (DD) that needs to be anticipated and analyzed.
In the article: "Data, Data, Everywhere", Baylor Arts and Sciences, Special Issue, Research in Arts & Sciences, Fall 2018 (by Julie Engebretson) it states: "It is estimated that about 16.3 Zettabyes of data - the equivalent of 16.3 trillion Gigabytes - is produced in the world each year."
Wow ! Now that's big data !
Dark Data (DD) is sometimes data that is obtained through numerous computer network operations but because it is so disorganized it is generally not used to obtain insights or for decision making .
HERE ARE A FEW OF THE AI RISKS AND THREATS
1. If a motor vehicle accident involves AI, trying to find the liable party is difficult and confusing. (i.e. a autonomous car hits another vehicle).
2. AI often has to identify vehicles,
people, roads, sidewalks, traffic signals, buildings or other items. To do this, AI relies on robotic vision (cameras, radar, infrared, sonar, etc.), sensors, and recordings. Things may not look the same to AI as it does to humans. AI also can reflect the biases of the software/firmware designer/developer.
3 AI is getting getting closer to actual human-like consciousness. This is also often called: Super AI.
4. Robots utilizing AI may sometime be seeking certain civil rights.
5. If AI commits or is involved in a crime who is guilty?
6. Privacy rights of people are potentially being eroded with AI.
7. AI has been developing a very rapid rate. AI has outpaced applicable legislation.
8. How will AI generated information be used in court? (No right of cross-examination, etc.);
9. Mass unemployment;
11. Autonomous weapons;
12. Mass surveillance;
14. Improper actions of robo doctors;
15. Wrongful financial services actions;
16. Constructing algorithms improperly;
17. Improper healthcare issues;
18. Improper insurance issues;
19. Ineffectively regulatory issues;
20. "Reasonable man" and "proximately caused" issues in tort law;
21. Non competitive antitrust AI issues .
22. In addition to the potential benefits of the quantum issues, we also need to keep an eye on the soon to be emerging potentially harmful uses of quantum physics (QP), quantum computers (QC), quantum mechanics (QM) and other quantum issues.
23. Eventuality of AI hacking and subversive activities involving rules, algorithms, data, and other and/or other components.
24. Potential autonomous killing machines (except in very rare US military cases).
25. In the event of consciousness in AI, seek the embedding of ethics, clean data, bias, responsibility, error prevention, legal compliance, integrity, moral values, ethics, continuous human feedback, and proper social manners.
26. Danger of manipulation of data.
AI CHALLENGES AND DANGERS TO DEMOCRACY AND PRIVACY IN THE US
Below is a great link to a very thought-provoking article on the AI dangers to democracy and privacy in the US:
"Artificial Intelligence: Risks to Privacy and Democracy
Karl Manheim* and Lyric Kaplan"
21 Yale J.L. & Tech. 106 (2019)
HOW TO MITIGATE AI RISKS AND THREATS
1. Understand the vast complexities of AI;
2. Track AI reasoning;
3. Participate in rule making, algorithms development, and data gathering;
4. Use good clean data;
5. Allocate AI risk amount applicable to the parties due to benefits received by the applicable parties;
6. Start soon with Augmented Intelligence applications, and then move on early with narrow AI applications before arriving at strong AI applications;
7. Protect AI with IP such as patents and trade secrets;
8. Establish an AI team;
9. Recruit and retain AI talent;
10. Humans will need to get smarter in order to deal with AI;
11. There needs to be a legal revolution with a focus on experiential and lifetime learning ;
12. Participate in government regulation of AI;
13. Make AI secure;
14. Develop strong AI testing prior to deployment; and
15. Think like a robot.
Certain portions of this section on AI legal risks and AI mitigation actions have been summarized, extracted from, and/or derived from:
FOUR INTERESTING VIEW POINTS ABOUT AI:
1. “The automation of factories has already decimated jobs in traditional manufacturing, and the rise of artificial intelligence is likely to extend this job destruction deep into the middle classes, with only the most caring, creative or supervisory roles remaining.”
— Stephen Hawking
2. “What to do about mass unemployment? This is going to be a massive social challenge. There will be fewer and fewer jobs that a robot cannot do better [than a human]. These are not things that I wish will happen. These are simply things that I think probably will happen.”
— Elon Musk
3. “You cross the threshold of job-replacement of certain activities all sort of at once.”
— Bill Gates
4. Here is an interesting comment from the "ProjectAI" web site:
"The coming AI developments may change life and society as we know it faster and more deeply than the industrial revolution, the internet or anything mankind has ever experienced before. It is crucially important to consider the coming AI developments not as a theoretical exercise or academic subject of investigation, but as a very real development which will take a concerted effort by people from all walks of life working together to take action to ensure continued quality of life for mankind."
"Contrary to popular belief, most AI systems currently act as a complement to humans instead of replacing them. According to expert estimates, we are still years away from general artificial intelligence and full automation. But eventually, there will come a day where robots will perform most tasks and the role of humans in the production cycle will be marginal.
It’s very hard to envision the dynamics of a robot-driven economy. But how will humans sustain their lives when robots take all their jobs?
Governments should impose an income tax on robots that replace humans, Bill Gates said in an interview with Quartz. The Microsoft founder proposed that the robot tax could finance jobs to which humans are particularly well suited. This can include taking care of elderly people or working with kids in schools, for which needs are unmet.
Other experts are endorsing the notion of a Universal Basic Income (UBI), or handing out unconditional money to all citizens. The concept has been around for centuries, but it is gaining traction as full automation starts to loom on the horizon.
There are many political, economic and ethical hurdles to the full implementation of the UBI, but pilot programs are underway. Governments as well as private firms are testing the concept in small scale.
We have yet to see how the accelerating evolution of AI will unfold, but what’s for sure is that fundamental changes lie ahead. While we can’t predict the future, we can prepare for its potential outcome as best as we can."
Don't Miss This Video About a Computer Using Human-Like Intuition
CHECK THIS GREAT LINK OUT !
HERE IS A GREAT LINK TO THE COGNILYTICA WEB SITE WHERE THEY LAY OUT A GREAT ONE-PAGE DESCRIPTION OF THE SEVEN PATTERNS OF AI.
TAKE A LOOK AT THIS INFORMATION
OK. Here are all of the previously-existing seven AI types that were recently identified by Cognilytica on their web site that apparently seem to repeat themselves in different combinations in all of the AI use cases that various entities have created. Some use cases make up just a single AI type for their use case application while others combine a few different AI types together to create their use case.
Here are seven AI types:
A. Goal-Driven System is using machine learning to give the AI system the ability to learn through trial and error in finding the optimal solution to a problem. The AI Chess win is an example of this.
B. Autonomous Systems (AS) are systems that are able to
perform a task, goal, or interaction with minimal to no human
involvement or labor. The primary objective of the autonomous systems type is to minimize human labor. Autonomous vehicles and drones are examples of this AI type.
C. Hyper-Personalization (HP) is defined as using machine learning (ML) to develop a very unique profile of each individual human, and having that special profile learn and adapt for a wide variety of purposes, including content, products, personalized recommendations, personalized healthcare, finance, and other specific one-to-one activities. Advertising, finance, medical/health care, and fitness are some of the AI uses of HP.
D. Predictive Analytics & Decision Support is using ML to
understand how learned patterns of data can assist in predicting the future. It also assists humans in making decisions about the future using what it learned from data and
other information. It is all about helping humans make better decisions. This application of AI is primarily intended to assist
humans to figure out an answer to a problem. The human is still making the decision but it is ML helping the humans make
the better decision.
E. Conversational / Human Interaction is the use of machines
in communicating with humans through natural conversation and including voice, text, images, and other forms. The goal is to create communication between machines and humans. This is enabling machines to interact with humans the way that humans interact with each other.
F. Pattern & Anomaly Detection is the AI type that has the ability to find which one of the things is like the other and which is not. The goal is to find anomalies in data and indicate what looks out of the ordinary such as fraud detection and risk analysis.
G. Recognition is the AI type using machine learning to identify objects or other things within some form of unstructured content. Use cases include image recognition, facial recognition, sound recognition, audio recognition, item detection, handwriting detection, text recognition, and gesture detection. The recognition AI type is one of the most widely used and adopted of all AI use types.
ROBOTS OPERATING IN DANEROUS CONDITIONS:
Understanding the Accident of Fukushima Daiic
Japan Special: Inside Fukushima Daiichi Nuclear Power Plant where robots are helping the clean-up after 2011's meltdown. Plus across the country, robots learn to sumo wrestle, play volleyball and act like humans
AI CAN AUGMENT PUBLISHED MATH RESEARCHERS
Proofs in theoretical math can be overwhelmingly complicated. Even professional mathematicians often do not fully comprehend them. Consequently, researchers just keep faith that the underpinnings of a new proof are correct.
Because of this, when mathematicians reference a published result in their work, readers just take their word for it. Kevin Buzzard, an Imperial College London mathematician, is worried about this. There can be a huge possibility that many existing mathematical proofs are incorrect.
Much of published math results may be wrong because researchers often cannot check background details used. Buzzard considers that a researcher should use artificial intelligence (AI) to do at least the background information checking. Stay tuned.
CONTINUE TO TOP OF COLUMN #2
ABOUT THE GROUP:
The "Group" (The Good United States Artificial Intelligence Group) is not a company, LLC, partnership, corporation, joint venture, proprietorship or any other legal or business entity. It is just an informal number of individuals loosely assembled that has some unifying relationship. In our case, that underlying relationship is the education, safety, ethics, integrity, responsibility, honesty, non-biasedness, empathy, and social good of Ai (in all of Ai's forms) as Ai transitions into more and more of our lives. We look forward to good beneficial Ai --- not bad Ai.
To quote our GoodUSAi.com Co-Founder & Member - Dr. Qin Sheng - "We need to let people know more about our interests, expertise and endeavors. Let’s do the best to support this wonderful country!"
WORK IN PROGRESS
"Do it -- Fix It -- Try It"
Chaotic action is preferable to orderly inaction
This web site will always be a ..."Work In Progress." Changes, updates, additions, and deletions will be made monthly. The last update to this web site was made on February 7, 2020
(C) COPYRIGHT JOHN E. MILLER, RUSS PETERMAN, AND DR QIN SHENG
2018 - 2020
ALL RIGHTS RESERVED (EXCEPT AS NOTED HEREIN WHERE ANOTHER PARTY OWNS A COPYRIGHT INTEREST )
COLUMN # 2
- MEMBER & ADVISOR-
THE GOOD US AI GROUP
BS, MS,Business Management
University of Texas - Austin, Texas
University of Texas - Austin, Texas
University of Texas - Austin, Texas
University of Colorado
THE PRESIDENT OF
PETERMAN CONSULTING ASSOCIATES
AUSTIN, ROUND ROCK, GEORGETOWN, TEXAS
COMPREHENSIVE BIO FOR RUSS PETERMAN
Commercial Glider/Self-launch Sailplane Pilot/Certified Flight Instructor
FAA Issued Jun 1974
COLUMN # 2
HERE IS A LINK TO 10 CURRENT EVERYDAY AI APPLICATIONS:
FOLLOW THE MONEY:
(To some limited extent, some of JohNny's commentary in this "Three Kinds of Money" section just below as well as the "AI Solutions Might Not Work Out" section also just below, may be summarized, paraphrased, derived from or extracted from an "AI Today" oral Podcast: Artificial Intelligence Insights - #57 - titled -"Is VC Funding for AI Over Heated ? "
October 3, 2018:
See: this link:
THREE KINDS OF MONEY
"Follow the money. AI and its subset, machine learning, are getting a huge amount of: (1) start-up ("non-Unicorn and Unicorn") VC funding,; (2) big company funding (Amazon, MS, FaceBook, Google, Apple, Oracle, Accenture, Twitter, IBM, etc), and (3) government funding (US, China, UK, France, Germany, Japan, Russia, and South Korea ( in that order) today.
For example (according to ABI Research) , in 2017 in the USA with its 155 AI VC companies there was US$ 4.4 billion in just VC Unicorn-type funding (not including big company and government funding). Chinese AI VC funding from its 19 companies for 2017 was US$ 4.9 billion.
AI government funding and AI big company investments in AI funding for 2017 is hard to calculate ... but consider this: SenseTime is a Chinese big AI company. Its valued at
US$ 4.5 billion."
AI SOLUTIONS MIGHT NOT WORK OUT:
"Even with all of this AI VC funding, it is possible that the resulting AI solutions might not work out.
Will AI prove itself out? Is it the big wave promised and will the AI bubble remain inflated? AI has tremendous promise. Many think that the US ROI will be met. But there is a lot of competition.
Having the right people in your labor force is important. For example, data engineers and data scientists are both urgently needed. Data scientists do the AI deep analysis. Data engineers do the management and cleaning of big data for use by data scientists. Data scientists do the deep analysis of clean big data. You need more data engineers than data scientists. Obviously, a data scientist needs to have very high STEM skills.
The governments of the US and China. overwhelmingly dominate the AI race. China has both its huge tech giants and government collaborating and working closely. The privacy issues do not hamper AI in China. The Chinese are adding AI to the high school curriculum as a mandatory subject. The US has a promising AI labor force. There are AI centers in the US in places like New York, Boston (MIT, Harvard), Pittsburgh, Seattle ,(MS, Amazon) Washington DC, Austin, northern and southern California.
The global AI race is on. Japan is soaring in AI robotics but has an aging labor force. South Korea is pushing electronics, medical, and health AI and will soon have 5000 new data engineers. UK and France are leading the way in responsible and ethical AI. Russia believes its future rests on AI and has sizable AI funding but lacks the creation of VC funding by companies."
A SAMPLING OF EIGHT CURRENT AI APPLICATIONS
1. ARGOBOT - AI APPLICATION FOR FARMING: Tracks the status and health of each individual strawberry to determine the best berry-specific management of each strawberry.
2. ROBORACE - AI APPLICATION FOR AUTO RACES: Race cars drive themselves in a race tracking and managing each vehicle in race with no human driver.
3. WILDBOOD -AI APPLICATION FOR MANAGING ZEBRAS AND ELEPHANTS: AI enabled system identifies each animal and tracks the individual health, location, and status of each animal in an African country.
4. ICEBURG - AI APPLICATION FOR HOCKEY: Manages the multitude of game statistics and analyzes pros/cons of each player.
5. SMART DRONE - AI APPLICATION FOR CONTROLLING PHOTOGRAPHIC DRONES; Manages the control and positioning of the drone including multiple cameras in order to take the best desired photo..
6. MOTION PICTURE TRAILER DEVELOPMENT: Motion picture companies are using AI to develop motion picture trailers based on the particular viewers of a trailer.
7. UNDERWATER ROBOTS: Underwater scientists are using AI in underwater robots to control the increasing number of harmful star fish.
8. USE OF AI TO ASSIST IN
SPOTTING SHARKS NEAR BEACHES
AI is used on a real time basis to locate and identify sharks that may harm humans near beaches.
GOOGLE AND DARPA
In June 2018, it was reported that Google was offering its resources to the US Department of Defense for Project Maven, a research initiative to develop computer vision algorithms that can analyze drone footage. In response, more than 3,100 Google employees have signed a letter urging Google CEO Sundar Pichai to reevaluate the company’s involvement, as “Google should not be in the business of war,” as reported by The New York Times.
Google then banned its development of AI software that can be used in war, business of war, and weapons.
Accenture has the right idea. As AI rolls in ... managers will have more time for important , strategic, and creative tasks while the time-consuming repetitive work is done by AI. More available time is a huge benefit of AI. They become super-managers."
OPEN AI vs. NATIONAL CLOSED AI
Who will win the AGI race?
The US WWII Manhattan Project is a good example of a successful private closed national program approach. However, today many prefer an open source type approach to the sharing of Ai information so that there is a democracy of Ai algorithms world-wide and other Ai information. An Open Ai Charter has even been established among certain world stakeholders. Will everyone be able to play in the same play box together?
AGI in 2042 ?
It has been said in certain Ai circles that the year 2042 will be when someone reaches AGI. However in the technology world things happen faster than one anticipates. For example, the Wright brothers originally stated that it would take 50 years before someone could fly in a heaver than air device. After making that statement It only took the Wright brother 3 years to make their historical flight.
“Hyper-personalization” is quite valuable in the hands and minds of today’s advertiser employing artificial intelligence (AI). Accenture feels that about 40% of the United States consumers have backed off their preferred products and services brands due to a lack by the advertisers' use of the latest trends of personalization and trust. Plain old “Personalization” is the inclusion with an advertisement that an advertiser sends to a consumer of allowable personal and transactional information in advertising, coupons, and other communications, such as: Name, Address, Title, Organization, Purchase History, Age, Address, Gender, Occupation, other similar items regarding your product or service.
On the other hand, Hyper-personalization goes several levels higher. Amazon is a great example of hyper-personalization. It has access to certain data points such as Full Name, Age, Address, Gender, Job, Search Query, Dwell Time on web sites,, Average Time Spent On Search, Past, Purchase History, Brand Loyalty, Average Spend Amount, among other factors. Using this, the advertiser can create a very accurate profile of you and use it to craft a highly relevant email, mailed ad, phone call, voice mail, text message, etc to the consumer. Some Hyper-personalization experts feel that Hyper-personalization effectively used in an ad contributes more to the consumers' buying decision than the actual content of the ad itself.
AI AND SIMULATIONS
We make use of simulations for all kinds of things. One common example today is teaching pilots how to fly. Another simulation is for doctors to practice various operations. So, how about using artificial intelligence (AI) and simulations to teach autonomous vehicles how to drive and to continue to improve its driving skills eventually from NHTSA Level 2 (vehicle does driving and steering while a human monitors) to NHTSA Level 5 (vehicle does it all without human assistance)?
Sure. However, we will need 5G Networks, advanced GPS, LIDAR, Centimeter-Accurate HD Maps, as well as some other technology advances. It will take millions of simulated vehicle trips but that can be done quickly and concurrently. The hardware and software for Level 5 autonomous vehicles is ready now but the vehicles themselves may not be ready for about 5 years due to NHTSA Level 5 extensive testing and validation.
In the USA there are about 5.5 million vehicle crashes per year that kills about 30,000 people annually. That is just in the USA. Amazing but true.
Some experts believe that by 2040 motor vehicle accidents will go down by 80% due to the use of AI in autonomous vehicles. The motor vehicle becomes safer and safer as it moves towards Level 5 AI autonomous vehicles. This will affect other things like the reduction in the price of auto insurance, reduction of the number of auto body repair shops, reduction in the number of vehicle accident personal injury attorneys, reduction of EMS / hospital services resulting from motor vehicle related injuries, double use of the trucks' 11 hour daily driving limits, driverless RV motor homes, and cost savings in certain aspects of the autonomous vehicles since doors, windshields, seats, AC, heater, and many other items would not be required.
AI WILL CHANGE PRO BASEBALL
Yes - it's true. AI will substantially improve scouting, recruitment, training, performance analysis, strategy, health, fitness, and safety of pro baseball. For example, take the new AI pitching strategy of the New York Yankees. According to long standing baseball traditions and baseball wisdom, the fastball (4 seam fastball or 2 seam fastball) is the very best professional baseball pitch to get an opposing batter out. Since those late 1870's starting during the formative years of baseball in the USA (when overhand pitching replaced underhand pitching), the fastball was considered the best pitch to get a batter out. It is the pitch that most pitchers rely on or go to. Generally, no pro team pitches less than 50% fastballs during a 9 inning game (except for the Yankees and now 4 other teams are doing it). Yankees ... 43.1% fastballs and Houston ... 47.3% fastballs. The Yankees do this despite that their pitching staff throws an average of 94.3 MPH fastballs - the fastest in baseball. Through AI analysis of pro baseball's big data the Yankees discovered that opposing batters were hitting the fastball often and missing the pitches with more deceptive movement, ball placement, and varying ball speeds (like the curve ball, slider, change of pace, fork ball, knuckle ball. etc). This is a good example how AI will challenge and change conventional thinking when conventional thinking may not be accurate.
HERE, IN THE GREAT LINK BELOW, (ACCORDING TO FORBES AND CONTRIBUTOR --- BERNARD MARR), ARE THE TOP 10 AI AND ML NON-MILITARY USE CASES THAT ARE SO IMPORTANT THAT EVERYONE SHOULD KNOW ABOUT THEM:
HOW AI / ML WILL AFFECT PROJECT MANAGEMENT ?
Project Manager (PM) -> Engage -> Approval -> Scope -> Initiate -> Plan -> Schedule -> Sequence -> Design -> Execute -> Monitor -> Report -> Control -> Quality -> Change Orders -> Product Sellers - Service Providers -Validate -> Manage -> Evaluate-> - Close Out -> Review Lessons Learned ->Perform Scorecards
A project manager (PM) is the person who is assigned the responsibility for the creation, operation, completion, delivery, and success of a project by:
(1) obtaining the project initial engagement, (2) obtaining directions/approval of management, (3) defining scope, (4) kicking off the initiation, (5) planning, (6) scheduling, (7) sequencing, (8) designing, (9) executing, (10) monitoring, (11) reporting, (12) controlling, (13) establishing quality, (14) handling of change orders, (15) managing product sellers/vendors, (16) managing service providers, (17) validating, managing, (18) evaluating, and (19) closing the project (including (20) reviewing the lessons-learned, as well as (21) performing a final project scorecard on the project and each seller/vendor/service provider.
Such a PM can be found in all types of projects such as (but not limited to) construction, petrochemical, architecture, environmental, mining, manufacturing, aerospace, information technology, software development, computer hardware. insurance, legal, retail, telecommunications, health care, financial, research, and many other different industries that produce/perform/sell/license/lease products and / or services.
The PM should make sure they control “RISKS”. All topics, issues and risks must be identified, assigned, managed, and resolved. Typically, the PM will manage project tasks under the project by using Microsoft Project Software.
The role of PM is similar to that of a musical band leader. The band leader doesn't need to play all the instruments in order to conduct the band, so the role of PM needs to manage a team of people who have different roles and skills in a project.
However, the PM needs the following PM skills: social, point-of-contact with customer, good communicator to all parties (internal, external, customer, stakeholders, etc), coach, mentor, listener, strong domain, technical, project at-hand knowledge & experience, team-player, persistence leadership, strategy, and business management, proactive, quasi-psychologist, common sense, real-world mentality, problem-solving skills, and facilitator.
AI / ML will probably affect both PMs and projects, as follows: (1) the human PM will have about 50% more time (according to Accenture LLC) to address the serious, and important project issues since the AI / ML will allow the AI / ML project applications to handle and resolve the repetitive minor time-consuming administrative project issues; (2) use of “smart assistants” by the human PM; (3) creation of automatic periodic risks analysis; (4) AI / ML becomes a project valuable team member; (5) use of robots; (6) creative thinking by humans to solve complex problems; (7) valuable tools for reporting and monitoring; (8) use of predictive analysis; (9) evaluating KPI’s; (10) reducing costs and mistakes; (11) avoiding surprises; (12) early warning of surprises; (13) predictive maintenance; (14) risk removal; (15) tracking progress / performance; (16) use of chat boxes; (17) forecasting; (18) use of experimentation; and (19) other similar activities.
PM’s will not be replaced by AL / ML but rather PM’s will be assisted by AI / ML.
A better description is that the PM will be assisted by Augmented Intelligence .
WHAT IS ... “BAT” ?
(It has nothing to do with baseball or flying mammals).
As so reported / blogged on April 2, 2019 by Karen Hao ...
-- (at MY MPCA - mympcapital.
- "China’s AI Industry") --
a little more than 50% of China’s 190 leading AI/ML companies received private funding from the three largest Chinese AI/ML giant tech companies ( Baidu, Alibaba, and Tencent - often referred to jointly as “BAT”).
Each of these BAT companies, although widely focused on many aspects of AI/ML, also has a known expertise. For example, Alibaba is e-commerce; Tencent in social networking; and Baidu in search and information indexing.
China has set a challenging goal to be the World Leader in AI/ML by 2030. However, China appears to be "top heavy". "Top Heavy," in this context, means that China is strong on the AI/ML applications but not as strong as they should be on the fundamentals that support the AI/ML.
China is still behind the US in expanding AI/ML capabilities through fundamental research, algorithms, advanced silicon chips, machine vision, natural-language processing, and other AI/ML fundamental capabilities.
"2019 STATE OF AI REPORT"
(JUNE 29, 2019)
SEE THE LINK BELOW TO THIS GREAT 136 SLIDE REPORT PREPARED BY NATHAN BENAUCH AND IAN HOGARTH:
THE STATE OF AI REPORT REPORT IS PREPARED ANNUALLY BY NATHAN BENAUCH AND IAN HOGARTH
HOW AI / ML IS WORKING IN THE LEGAL SERVICES ARENA
In a recent LexisNexis survey (according to John G. Browning and Christene Krupa Downs in their article - "The Future Is Now" in the July 2019 issue of The Texas Bar Journal) only 20 to 25% of the US in-house legal departments currently use some AI / ML applications. In those low percentage of AI / ML companies, generally only one legal domain area is being addressed. As compared to a company’s in-house Finance and HR departments, those departments of a company are twice as likely to use AI / ML applications than in-house legal.
Take the Coca-Cola’s legal department for example. By using AI / ML applications they have reduced a standard legal agreement for review from 10 person hours to about 15 minutes per document.
In another case, JP Morgan Chase & Co. has saved over 360,000 person hours of legal services time in the last year due to its AI / ML applications and the use of Chase’s own AI / ML platform called “Contract Intelligence”.
Likewise, many outside private counsel firms are using AI / ML to reduce time charged to their clients, be more competitive, perform due diligence, do document review, perform predictive analysis, do AI / ML assisted legal research, use of AI / ML applications for damages models, do verdict prediction, perform predispositions of judges, create cost-saving, and do many other types of legal AI / ML that make lawyers more efficient and effective. This is especially true due to the huge amounts of legal “Big Data” that is now available.
Interestingly, outside legal counsel are the most intrigued with the AI / ML applications related to cost savings and predictive analysis.
The general consensus in the legal community is that, due to AI / ML, legal professionals at the lower service levels may experience declining numbers. However, AI / ML will benefit the legal profession by enhancing what lawyers do and freeing them up for more important, meaningful, creative, and fulfilling work.
18 WHEEL 40 TON TRUCKS
USING DRIVER-LESS AI AUTONOMOUS TECHNOLOGIES
According to Chris Spear -- President and CEO of the American Trucking Associations (ATA) -- as stated on the ATA web site -- (https://trucking.org). -- the current estimates are that there are 3.5 million truck drivers in the USA. In fact, the USA needs roughly 90,000 more truck drivers yearly just to keep up with ever
increasing demand for drivers.
Unlike trucks driven by humans, self-driving trucks using autonomous vehicle (AV) technologies, can drive more than the current 11 hour daily driving limit. Also, driver-less trucks do not need breaks for food, restroom use, and other personal delays/stops.
One company ("Too Simple" of San Diego, CA) is already driving one of their Level 4, driver-less loaded trucks on “dock to dock" regular revenue runs in Arizona three to five times weekly (also carrying a safety engineer and safety back-up driver). They expect fully autonomous, driver-less, Level 5, and commercially- ready 18 wheel 40 ton trucks by the end of 2020.
Alternatively, it may be a future in which self- driving trucks using AV technology drive the long highway miles between what they call transfer hubs, where human drivers will take over for the last miles through complex urban and industrial areas. So, self-driving trucks may be complementing humans, not completely replacing them.
AV technology holds enormous collaborative potential for the trucking industry, its drivers and the motoring public. That could mean improving the truck availability percentage during assigned runs from 50% to 80%.
It is anticipated that the "Perception" distance for trucks driven by humans could increase from about 1/4 mile achieved by current human drivers to 1/2 mile for very high perception AI driver-less systems. In other words, all things considered, the Perception System to be employed on the driver-less truck is expected to much better vision than a human driver.
With 94% of highway accidents attributed to human error, the successful deployment of AV driver-less technology can drastically reduce fatalities on the road.
Moreover, the AI driver-less technology can deliver other significant returns by reducing traffic congestion, improving driver productivity and decreasing emissions through lower fuel burn.
CAN YOU PLAY
"G O" ?
AI ... or ... AAI ?
We currently seem to be a good distance from establishing advanced intelligence systems that can independently process, reason, and create in the same capacity and capability (or better) of the human brain (i.e. artificial intelligence - AI).
However, is matching (or improving upon) the capacity and capability of human intelligence the most useful AI application (all things considered)? Should AI be the conscious singularity wide ultimate intelligence goal for businesses and other entities?
How about considering the growing worthy alternate concept of: “Augmented Artificial Intelligence” (AAI)? AAI focuses on supplementing human intelligence so as to augment humans. AAI maintains humans at the center of the appropriate management and decision making.
The innovative technologies implementing AI and AAI are the quite similar, however, the applications of AI and AAI are quite different. AI aims to create systems that run without humans. On the other hand, AAI aims to create systems that make humans better. Isn't our goal to make humans better? What do you think is the best approach? AI or AAI?
WILL THE UNITED STATES WIN, PLACE OR SHOW IN THE AI / ML
HORSE RACE ?
As of this point in time in late 2019, despite the persistent hype, there appears to be several countries currently struggling to achieve a global advantage and become the leader in the artificial intelligence (AI) and machine learning (ML) “horse race..
Those AI / ML -focused countries clearly understand that AI / ML is composed of several foundational technologies that can sky-rocket competitiveness, increase productivity, improve health care, protect national security, improve transportation, care for the environment, favorably address climate issues, provide adequate energy, control space, and help solve many other societal challenges (just to name a few of the many current world challenges).
It is quite interesting to objectively compare China, the European Union (EU) countries (currently including the UK, Germany, Norway, and Sweden ), and the United States (US) in terms of their ranked standing in the AI /ML economy by examining the following eight categories: (1) talent, (2) research, (3) scholarly journal articles, (4) development, (5) adoption, (6) big data, (7) hardware, and (8) committed AI / ML strategy.
The bottom line summary is that despite China’s bold AI / ML initiative (claiming it will become the most dominate AI / ML force in the world BY 2030), the US still leads the world (at least for now).
China comes in second, and the EU countries lag further behind in third place. Ranked behind the EU countries (in no order of priority) are Russia, Japan, South Korea, Singapore, India, Israel, Canada, Indonesia, Brazil, Mexico, and Taiwan.
This AI / ML ranked order could change in the next few years as: (1) China appears to be making more rapid progress than either the US or the EU countries (28 member states); and, (2) the China AI / ML labor force becomes even larger. That’s 1.43 billion general population in China compared to 323 million general population in the US. China’s 1.43 billion general population makes a sizable dent in the total world general population of 7 billion and creates massive amounts of big data (the driving force behind AI / ML).
HOW TO SUBSTANTIALLY REDUCE HARMFUL AI BIASES -— AKA (“GARBAGE IN … GARBAGE OUT")
Harmful human biases can: (1) be innocently and inadvertently found inserted in an artificial intelligence (AI) system; or (2) be formed intentionally with potential intended harm and found inserted in an artificial intelligence (AI) system. Ideally, such AI biases, once discovered and identified are then corrected by humans, AI systems, and/or by humans and AI systems in a joint collaborative manner.
Although bias is considered by many to be AI's “Achilles Heel” or AI’s “weak link in the chain”, once the AI bias is located it can in most cases be mitigated.
Unfortunately , often the AI bias passes into active use in an AI use case application without notice, identification, analysis, and correction … “like a large ship quietly passing in the night.”
Additionally, an AI system itself can be creatively and efficiently used as a virtual software tool to help humans to identify the remainder of AI bias system problems (not earlier identified and previously cured by humans). This preventive procedure using AI typically may reduce the harmful impact of human biases. However, it should be noted, unless AI biases are properly and promptly addressed in a timely manner by appropriate humans and AI, then such a less-than-needed amateur type effort can potentially make the resulting AI bias problems even worse.
For the purposes of this article the example of correcting AI baises in the hiring/employment activity will be used.
Miscellaneous Take Away Guidelines:
1. AI biases can potentially creep into AI algorithms and data in several ways and should be deleted, mitigated or substantially reduced as follows: The take-away guideline here is to initially remove from algorithm and data consideration, analysis, and discussion of certain sensitive variables such as age, gender, race, pregnancy, disability, health, religion, politics, sexual orientation, and other similar applicable variables. This is the initial first step in mitigating AI biases.
2. AI systems typically learn to make their AI decisions (including other similar non-AI decisions) based primarily on big data, which can include pre-existing “baked-in” biased human decisions or reflect historical and social inequities. The take-away guideline here is to be observant, discerning, and aware so as to delete from further algorithm and data consideration any pre-existing “baked-in” biased human decisions or biases that reflect historical and social inequities.
3. A third takeaway guideline here is for your business leaders to stay up to-date on this fast- moving field of research of AI biases.
4. Perhaps a fourth takeaway here is to consider using operational practices such as internal review such as : Internal AI Bias Review Teams, “Red Teams” or Third-Party Audits to help mitigate AI biases.
5. The fifth takeaway is that since algorithms that tend to favor certain male employment applicants (based on the use of seemly innocent words like “executed” or
“captured” that are more commonly found on men’s resumes) need those words need to be corrected by mitigating/revising such words.
6. Your sixth takeaway here might be to define and develop a way to measure “fairness” between employment applicants such as using AI models that can be more readily evaluated.
7. A seventh takeaway here is to consider that AI can improve on traditional human decision-making. Machine learning systems disregard variables that do not accurately predict outcomes (in the data available to them). This is in contrast to humans, who may not even realize the factors that led them to, say, hire or disregard a particular job candidate.
8. As an eighth takeaway, AI biase mitigation approaches that only look for a one-time “fix” for fairness oversimplifies the complexity of social systems. Within each domain – such as education, healthcare or criminal justice – legacies of bias and movements toward equality have their own histories and practices. Legacies of bias cannot be “solved” without drawing on domain expertise. Addressing fairness meaningfully will require interdisciplinary collaboration and methods of listening across different disciplines.
The purpose of this article is not to list all of the viable ways to reduce, mitigate, or eliminate AI biases. That could almost be endless for a vast multitude of domains. Rather, the goal of this article is to encourage meaningful and immediate AI bias reduction research and development.
ARTIFICIAL INTELLIGENCE (AI) AND MACHINE LEARNING (ML) ARE FOR THE BIRDS !
It has been said (by some) that gliders don’t really fly — they just fall elegantly. However, is that really correct?
No ! As so described by the US astronaut Neil Armstrong: “Gliders, sailplanes are wonderful machines. It’s the closest you can come to being a bird. Silent flight – it’s wonderful.”
Neil’s comparison of birds to gliders is a good comparison. Birds, for example, don’t always flap their wings to fly. They often soar and glide by using rising invisible columns of warm air known as … “thermals.”
Birds can stay airborne for quite some time while using little body energy. Bird experts are using AI and ML to figure out just how the birds perform this elegant gliding skill.
Recently some Italian scientists used machine learning (ML) to train an algorithm to control a glider to navigate thermals. Their efforts suggest that autonomous glider aircraft could use thermals in a similar manner as the birds. Wow ! Autonomous gliders.
When training the AI algorithm, the scientists found that some logical factors, such as: (1) using vertical wind acceleration; and (2) using side-to-side torque might be used to teach the glider to fly correctly. Birds may handle it this way too.
The bird scientists used AI’s reinforcement learning techniques creating repetitive trial and error actions in a way that maximizes a certain AI- favored reward. Starting with no prior knowledge of the required task(s), AI learns how to function correctly over time using trial and error. In this application, the input to AI was composed of the glider’s pitch, yaw, groundspeed, airspeed, and flight information. Likewise, the AI’s doggy-like reward was to maximize its climb rate (a speed causing the glider to gain height). Strange reward - right?
The algorithm trained first in a simulator. Later it was trained in real life. Guess what ? It took about 240 flights near Poway, California at about three minutes of training time on each flight.
The pilot steered the glider to a fixed location using a manual controller. Then the AI took control, using the air currents from thermals (which can travel as fast several meters a second) to climb into the wild blue yonder..
During good situations the glider could stay in the air for about 45 minutes. Some flights, where the wind was too strong, the pilot had to prematurely fly it back home.
Thermals are just one type of updraft that birds use. Other updrafts are created by air currents rolling over mountain ridges. In some cases, updrafts can be created by the collision of air masses in “convergence zones sometimes found on shores and over desert boundaries.
AI and ML seem to to work OK with thermals. Also, AI and ML will probably also work OK with future autonomous gliders that use AI and ML to navigate thermals over long distances. It could eventually be used for long-term scientific surveys and for ambitious projects, like tracking bird migrations wingtip to wingtip.
Also, if we learn how to fly and soar with the birds using gliders not only will we learn more about efficient gliding and autonomous gliding, but we will also learn more about the fascinating topic of how the lives of birds can substantially enhance the lives of humans in so many important ways.
Algorithmic Future ?
Artificial intelligence (AI) and drones work very well together in a variety of applications and in various sectors including defense, agriculture, natural disaster relief, security, sports, travel, recreation, construction, medical and others. AI-enabled drones are helping humans do things previously thought impossible. One way to think about AI drone information is that there are two basic parts: (1) Drone Software; and (2) Drone Hardware.
AI Drone Software Examples:
With the new AI software for drones, it can visualize its surroundings and provide analytical feedback on a real time basis. There are, of course, many applications for these types of capabilities.
AI drones are used to combat elephant (and other wildlife) poaching in Africa by use of an image recognition software technology. It monitors the condition of elephant herds and locates possible poachers.
There is an AI software platform for public safety officials that takes unstructured data harvested by drones and turns it into key structured clean data for police, fire, emergency teams, and other first responders.
AI drone software has been used by many teams to combat a variety of terrorist and other public safety threats.
AI drone software is used by SWAT teams in gathering and analyzing crime scene intelligence.
AI software can assess damage after hurricanes, floods, tornadoes and employs thermal imaging to locate missing persons and pets.
A deep learning neural network helps drones hover slowly through crowds to find and identify key persons of interest. In order to scan large groups for an individual, its AI-powered software needs about 20 minutes prior to its in-flight operation to understand the image(s) of an individual.
Machine learning software on the drone platform is
often used for training purposes by various companies
such as: (a) insurance companies to review and
analyze insurance claims; (b) high schools and
colleges to teach STEM - related topics; (c) military to
teach flight training; (d) US Space Force training; and
AI Drone Hardware Examples:
1. Drones are used for surveying and analyzing equipment (such as those big windmills and remote oil/gas wells) at greater and more accurately than humans.
2. Drones are fully-autonomous and are used in aerial real estate surveying, pipeline inspections, disaster response and search and rescue. An average commercial drone can travel over 100 miles and reach speeds of up to 90 miles per hour. Military drones are much faster and can go much further. (See URL at the end of this article).
3. With no human interaction, there are autonomous drones that combine AI supercomputers and many cameras to capture various styles of video footage.
4. AI-and machine learning-enabled drones are used for greenhouse management that can fly around greenhouses and collect data on temperature, humidity and carbon dioxide levels to ensure that plants are growing in an optimal ecosystem.
5. Other AI drones can analyze soil and crop health to ensure that plants are disease-free and unhindered in their growth.
6. Some AI drones assist with mapping and the designing of large-scale warehouses and factories.
7. Some applications use drones and AI analytics to scan shelves and automatically track and order new inventory. Drone flights can even be programmed to perform tasks during off-hours and do it themselves thereafter.
8. There is an autonomous Drone used by businesses and policymakers to collect environmental analytics.
9. There are drones that can fly over forests, mountains, quarries or farms to assess damage from natural disasters.
10. Climate change information is gathered by drones and then reported back using the analytics platform.
11. A drone can be used to inspect railroads, power lines, oil pipelines, and solar panels for quality-control and upkeep purposes. The drone can then travel the required route, film what it sees and create a report based on its findings.
12. An AI drone can assist ground forces and first responders in exploration and data gathering. Equipped with certain software it can identify individuals in an emergency situation.
13. AI drones assist law enforcement and military personnel in reconnaissance missions. The robots can access GPS-denied areas such as building interiors and underground facilities to gather ground-level intelligence.
14. Some drones are used for military recon and some carry a precision strike warhead(s) for military operations.
15. Other drones are being used in agriculture to map field acreage, spot crop health issues and determine irrigation, cultivation, herbicide use, pesticide use, fertilizer application, and picking issues.
16. Drone software is used to inspect and analyze the physical condition of dams, sky scrapers, Statue of Liberty, monuments, radio and cell phone towers.
17. The MQ-9 Reaper is a hardware military drone that can fly 300 mph and can fly re up to14 hours. For a description of the “Reaper” Drone - See: URL below.
So, what does the future hold for AI drones? Check out these 3 URLs.
CONTINUE TO TOP OF
WORK IN PROGRESS
"Do it -- Fix It -- Try It"
Chaotic action is preferable to orderly inaction
This web site will always be a ..."Work In Progress."
Changes, updates, additions, .and deletions will be made monthly. The last update to this vertical column of this web site was made February 7, 2020
(C) COPYRIGHT JOHN E. MILLER , RUSS PETERMAN,
AND DR QIN SHENG 2018 - 2020
ALL RIGHTS RESERVED (EXCEPT AS NOTED HEREIN WHERE ANOTHER PARTY OWNS A
COLUMN # 3
JOHN (JOHNNY) E. MILLER
-- CO-FOUNDER --
- MEMBER & ADVISOR -
THE GOOD US AI GROUP
BA, JD, LLM, CPCM
University of Memphis
Cecil C. Humphreys School of Law
University of Missouri (KC)
National Contracts Management Association
State Bar of Texas
MISSOURI BAR ASSOCIATION
Inactive - Out of State
THE JOHN E. MILLER LAW OFFICE
STATE BAR OF TEXAS
MISSOURI BAR ASOCIATION
COMPREHENSIVE BIO FOR JOHN ("(JOHNNY") E. MILLER
United States Army,
Security and Intelligence, S-2,
442nd Signal Battalion,
29th Signal Group,
Nahon Ratchasima Province,
1969 to 1971
1. HERE IS A SUMMARY OF A PORTION OF THE ARTICLE: "GAME CHANGING TRENDS TO LOOK OUT FOR WITH AI ."
THIS IS AN INTERESTING PLAIN-ENGLISH ARTICLE ABOUT THE FUTURE OF AI
The author advises: "A recent report by McKinsey states that Baidu, the Chinese equivalent of Alphabet spent $20 billion in AI last year.
At the same time, Alphabet invested roughly $30 billion in developing AI technologies. The Chinese government has been actively pursuing AI technology in an attempt to control a future cornerstone innovation. Companies in the US are also investing time, money and energy into advancing AI technology.
The reason for such interest towards artificial intelligence is that artificial intelligence can enhance any product or function. This is why companies and governments make considerable investments in AI."
The author concludes by stating:
"Therefore, the best approach is not to wait until AI leaves you unemployed, but rather proactively embrace it and learn to live with it. As we said already, AI can also create jobs, so a wise move would be to learn to manage AI-based tools. With the advance of AI products, learning to work with them may secure you a job and even promote your career. ... The wisest strategy is to embrace artificial intelligence and let it work to maintain our well-being. "
2. HERE IS ANOTHER SUMMARY OF AN INTERESTING ARTICLE ABOUT : "HOW TO PREPARE YOURSELF AND YOUR CAREER FOR AN AI - FUTURE"
Kasper Nymand kaspernymand.com/
Mr. Nymand states: "The future is here. It’s time to rethink your position in the job market. It’s time to question your current career path, and consider if it would be wise to make a few changes. Bots and intelligent assistants / agents are on the rise, ready to disrupt and turn the job market on its head. Now, it has always been key in life to stay up-to-date, grow and develop yourself and your knowledge base. But now more than ever this is definitely true.
AI is expected to eliminate 6% of jobs by 2021, according to Forrester, and a further 47% is expected in the next two decades, according to a US Federal report. Those are current types of jobs, but the evolution will very likely create new types of jobs that we haven’t really been thinking about yet."
The author further states: "What this all means is simply that you need to keep yourself in the loop. You need to keep following the newest trends and tendencies in your market, and just generally beware that things are going to change. But remember, it's most likely not going to be the end of your career."
The author offers the following advice:
"We should all take the time to do this simple, yet comprehensive, exercise every now and then. Nobody is safe in a time with intelligent machines. We’re all in the same boat. Some are just closer to the edge than others. Get up and move as close to the wheelhouse as possible. Use your imagination, be creative and open to new combinations and collaborations within and across industries. Innovation and disruption will happen eventually, it’s just a matter of whether you’ll be in charge or somebody else will."
Author: Kasper Nymand:
ABOUT THE AUTHOR OF THE ABOVE ARTICLE
UNILEVER HR USE OF AI
Unilever hires many people and sells a multitude of products worldwide. It has massive duties in dealing with employee recruitment, interviewing, employee selection, and managing employees, Thus the Unilever HR departments have huge efforts in dealing with this very important area of Unilever. See the following link to understand how Unilever is creatively and efficiently addressing this HR issue using AI (machine learning and neural networks): https://www.businessinsider.com/unilever-artificial-intelligence-hiring-process-2017-6
Check out this link to a good WSJ article about how savvy lenders are using AI to better analyze potential borrowers. Lenders, with the help of AI, are completing a more thorough risk management analysis of potential loans and as a result are making more risk-appropriate loans to more borrowers who meet the acceptability standards: https://www.wsj.com/articles/ai-helps-auto-loan-company-handle-industrys-trickiest-turn-11546516801
OPAQUE BLACK BOX
In XDNET's article "Inside the Black Box: Understanding AI Decision Making" Charles McLellan states:" AI algorithms are increasingly influential in peoples' lives, but their inner workings are often opaque." This article discusses what is being done about that problem. We need to be able to peek under the hood and understand why AI makes decisions. Of course, not everyone agrees with that sort of visablity because humans may not be able to understand and explain.
FUTURE OF AI, ANI, ACI, AND ASI (SEE LINK AND TEXT BELOW)
Here is a very good article that addresses the possible future of AI, ANI, AGI, and ASI in as early as 2029:
Title: "The Risks of Advanced AI Are Real. We Need to Act Before It's Too Late, Warn Experts"
BY PAUL SALMON, PETER HANCOCK & TONY CARDEN, THE CONVERSATION
FEBRUARY 01, 2019
"Artificial intelligence can play chess, drive a car and diagnose medical issues. Examples include Google DeepMind's AlphaGo, Tesla's self-driving vehicles, and IBM's Watson.
This type of artificial intelligence is referred to as Artificial Narrow Intelligence (ANI) – non-human systems that can perform a specific task. We encounter this type on a daily basis, and its use is growing rapidly.
But while many impressive capabilities have been demonstrated, we're also beginning to see problems. The worst case involved a self-driving test car that hit a pedestrian in March. The pedestrian died and the incident is still under investigation.
The next generation of AI
With the next generation of AI the stakes will almost certainly be much higher.
Artificial General Intelligence (AGI) will have advanced computational powers and human level intelligence. AGI systems will be able to learn, solve problems, adapt and self-improve.
They will even do tasks beyond those they were designed for.
Importantly, their rate of improvement could be exponential as they become far more advanced than their human creators. The introduction of AGI could quickly bring about Artificial Super Intelligence (ASI).
While fully functioning AGI systems do not yet exist, it has been estimated that they will be with us anywhere between 2029 and the end of the century.
What appears almost certain is that they will arrive eventually. When they do, there is a great and natural concern that we won't be able to control them.
The risks associated with AGI
There is no doubt that AGI systems could transform humanity.
Some of the more powerful applications include curing disease, solving complex global challenges such as climate change and food security, and initiating a worldwide technology boom.
But a failure to implement appropriate controls could lead to catastrophic consequences.
Despite what we see in Hollywood movies, existential threats are not likely to involve killer robots.
The problem will not be one of malevolence, but rather one of intelligence, writes MIT professor Max Tegmark in his 2017 book Life 3.0: Being Human in the Age of Artificial Intelligence.
It is here that the science of human-machine systems – known as Human Factors and Ergonomics – will come to the fore.
Risks will emerge from the fact that super-intelligent systems will identify more efficient ways of doing things, concoct their own strategies for achieving goals, and even develop goals of their own.
Imagine these examples:
an AGI system tasked with preventing HIV decides to eradicate the problem by killing everybody who carries the disease, or one tasked with curing cancer decides to kill everybody who has any genetic predisposition for it
an autonomous AGI military drone decides the only way to guarantee an enemy target is destroyed is to wipe out an entire community
an environmentally protective AGI decides the only way to slow or reverse climate change is to remove technologies and humans
that induce it.
These scenarios raise the spectre of disparate AGI systems battling each other, none of which take human concerns as their central mandate.
Various dystopian futures have been advanced, including those in which humans eventually become obsolete, with the subsequent extinction of the human race.
Others have forwarded less extreme but still significant disruption, including malicious use of AGI for terrorist and cyber-attacks, the removal of the need for human work, and mass surveillance, to name
only a few.
So there is a need for human-centred investigations into the safest ways to design and manage AGI to minimise risks and maximize the benefits .
POTENTIAL USES OF AI AND MACHINE LEARNING IN USA EDUCATION
It is estimated by Cognilytica, in their #87 "AI Today" Podcast, that by 2024 six billion US dollars will be spent annually in the USA on educational AI and machine learning.
The types of educational case studies that are anticipated are: augumation of teachers on non-instructional tasks; tutoring; hyper-personalized learning; AI classroom assistants / robots; voice systems; and possibly students advancing at their own individualized learning pace with hyper-personalized academic content regardless of a student's grade level.
SOME INITIAL IDEAS ON HOW TO MONETIZE AI and ML. ... NOW
Obviously, a company that has invested a lot of time, energy, and resources in getting its own company ready for AI and ML applications in order to benefit its own organization will want to implement beneficial applications internally for its own company.
However, why not get started now with some of that "low hanging fruit" type ROI.
At the right time (not too soon or too late) start the process of initially monetizing ROI. (Keep in mind that it's ..."the second mouse that gets to eat the cheese").
1. Provide tools for people to build there own solutions.
It is kind-of like how during the gold rush in the western USA in the 1800's the first folks to profit from the gold rush were not the gold miners but rather the merchants selling wheel barrels, picks, shovels, and other tools to facilitate the gold mining. So selling AI and ML related tools (products), like various machine learning foundational platforms, can be profitable. The key word is here "tools."
2. Leverage the power of AI and ML for numerous AI and ML activities involving ordinary consumer activities, such as conversational chat bots, special Siri-like domain-specific products, and chat boxes, so as to give the ordinary consumer a taste of AI and ML. Other examples here are things like Uber, on-demand insights, and language recognition.
3. Explore the intersection of AI / ML and enterprise / company business processes so as to help the business save time, to make more money, to reduce costs, to improve safety, to be more efficient, to gain more insights, to improve compliance, to improve customer services, and to examine and improve company processes.
4. Provide AI and ML services and consulting in unstructured data, structured data, big data, algorithms, artificial neural networks, robotics, robotic vision, drones, math services , statistical services, legal services, facial recognition, voice recognition, speech recognition, deep learning, signal processing, and other services related to ML and AI.
5. Sell ownership or license access of Big Data owned or legally acquired by your company to customers for use with customers' AI and ML projects.
6. Create an indirect sales channel by engaging resellers, value added resellers,
integrators, and sales reps that will market and sell your Big Data.
7. Create direct and indirect sales channels of all other individual AI and ML components.
8. Establish education and training programs for AI and ML .
9. So as to proactively encourage world-wide equal-access democratization of certain open source non-proprietary AI and ML information, data, hardware products, software items, and other Al and/or MI relevant things, the Parties state the following:
The Parties will attempt to create a non-binding informal Joint Cooperation Agreement (JCA) between certain companies or other entities (i.e. universities, government organizations, research organizations, etc) that have relevant AI and ML experts working for or engaged by such companies or other entities. The JCA companies or entities would meet periodically(electronically or in-person) to discuss open source non-proprietary AI and ML issues.
10. Focus your monetizing efforts on AI now and not on AEI.
THE FACIAL RECOGNITION ASPECT OF AI AND MACHINE LEARNING
Now here is a controversial topic. According to the Gemalto web site (now part of the Thales Group) [See: https://www.thalesgroup.com ] it states - ”Facial recognition is the process of identifying or verifying the identity of a person using their face. It identifies, captures, analyzes, and compares patterns based on the person's facial details. The face detection process is an essential step as it detects and locates human faces in-person and in images and video. The face capture process transforms analog information (a face) into a set of digital information (data) based on the person's facial features. The face match process verifies if two faces belong to the same person.”
Also, on the Gemlto web site (See link above), it is stated: “Biometrics are used in assisting one in identifying and authenticating a person using a set of recognizable and verifiable data unique and specific to that person. 2D or 3D sensor "captures" a face. It then transforms it into digital data by applying an algorithm, before comparing the image captured to those held in a database. T hese automated systems can be used to identify or check the identity of individuals in just a few seconds based on their facial features: spacing of the eyes, bridge of the nose, contour of the lips, ears, chin, etc. They can even do this in the middle of a crowd and within dynamic and unstable environments.”
Apple, Facebook, Google, IBM, Oracle, Accenture, MIT, Harvard, Stanford, Amazon, USA, UK, Russia, China, India, Japan, Canada, and others are rapidly researching, developing, and applying facial recognition products and services.
The following recognition signatures using the human body are also applied: fingerprints, iris scans, voice recognition, digitization of veins in the palm of the hand, and other behavioral criteria. Those biometrics (for the most part) are used to protect internet based monetary transactions where illegal/unlawful cyber actions have been rapidly expanding.
Why is facial recognition so enamored by Ai personnel today? Facial biometrics continues to be the widely preferred biometric benchmark. That's because it's easy to plan, prepare, deploy, implement, and analyze feedback. There is no hampering or interfering physical interaction required by the facial recognition user. Additionally, face detection and face match procedures for implementation, identification, and verification are so amazingly quick.
Facial recognition use cases today currently include law enforcement identifications, lost persons identification, military identification uses, social media identification / tracking, identification of threats / risks in large venues ( i.e. stadiums, airports, stations, festivals, political gatherings, and other facial recognization uses.
To be successful, one must eliminate, as much as reasonably possible, events of mistaken indentity.
5G and AI CAN WORK TOGETHER FOR GOOD
· 5G is the next-generation wireless mobile communication technology. It will improve internet and telecom. 2G included texting. 3G included browsing. 4G allowed for wireless video conferencing.
· Qualcomm, Verizon, Sprint, AT&T, Erickson, and other 5G providers are currently rolling out 5G in 22 USA test cities.
· 5G will create huge improvements in traditional mobile communications systems and traditional networks.
· 5G will increase the efficiency and innovation in both the business/industrial and consumer sectors.
· Much of the innovation 5G will enable is currently unknown.
· 5G will be robust and will enable quick downloads. With 3G, you would be able to download an average HD movie in about 25 hours, with 4G it would be less than 10 minutes, and with 5G, it would be about as 4 seconds.
· 5G will enable “smart” cities and“clever” countrysides.
· AI will soon enable machines and systems to function with intelligence levels similar to (or better than) that of humans.
· Artificial intelligence (AI) and Machine Learning (ML) promise many beneficial use applications, however, the required AL/ML processing speed is currently a limiting challenge.
· With 5G wireless technology, the current AI processing challenge will be much less of a barrier.
· 5G will soon be ready and able to provide the requisite speed that will increase the processing capabilities of AI.
· As a result of 5G, AI gets to analyze data much faster and to learn much faster.
· AI applications are currently being integrated into devices, rather than waiting for 5G to be developed and deployed.
· With 5G working online enabling simulations for analysis, reasoning, data fitting, clustering and optimizations, AI will become more reliable and accessible.
· 5G and AI integration will happen on the same chips on mobile smartphones, making those phones even more intelligent.
· 5G will serve as the basic technology for future Internet of Things ( IoT ) technologies.
· 5G will support the IoT applications of in various fields, including business, manufacturing, healthcare, academia, and transportation.
· 5G will enable the future massive network plans.
· Currently device-based processing is being used with AI.
· 5G will narrow the divide between processing in the cloud and processing on devices.
· 5G will have the speed to provide the services needed in the cloud.
· The narrowing the divide between cloud and on-device processing will be reduced with 5G.
· USA 5G will face strong competition from Huawei of China.
· 5G and AI need to be prepared for initial and on-going robust IP piracy and cyber-security threats.
POTENTIAL FUTURE OF AI / ML
The US AI / ML market place is projected by some knowledgeable individuals to potentially (all things considered) reach $70 billion by 2020 with a potential for AI / ML to also be unbiased, responsible, ethical, safe, legal, exhibit common sense, and good … not bad or evil. (Note: Gartner says that by 2021 AI Augmentation will generate $2.9 trillion in "business value").
Generally, in the US, about 63%, according to Price Waterhouse Cooper (Pwc), believe AI may allow humans to spend more time engaged in high-level thinking, be allowed to have more fulfilling activities, exhibit more creativity, and be involved in logical decision-making. On the other hand in the U ,(according tp Pwc), about 46% believe AI will harm people by taking away jobs and about 23% believe it will have serious, negative results.
The majority (or near majority) of US consumers, business executives, and subject-matter experts generally believe that AI will: improve health (including diseases such as cancer, heart related disease, diabetic issues, etc.), improve access to needed medical care, better access to legal assistance, improve transportation issues, encourage effective clean energy, improve gender equality, improve cyber security, better safeguard privacy, improve education, enhance economic growth, improve solutions for climate change, provide for more income equality, promote financial security, better track down fraud, and proactively provide fraud protection.
The business case for AI / ML is more productivity and return on investment. AI needs to be monetized.
It looks like the effect of structured clean Big Data means less repetitive tasks and more opportunities for collaboration.
Big Data is quickly becoming the new oil.
As a particular AI / ML user case becomes more and more important - it then becomes less and less noticeable over time.
P.S. AI / ML may also be able to improve traffic conditions in Austin.
AI / ML WILL ENHANCE CONTRACTS MANAGEMENT
THE NATIONAL CONTACT MANAGEMENT ASSOCIATION (NCMA) describes “Contracts Management” as follows: "Contracts Management is a profession that includes many positions along the buying and selling chain, including jobs within the federal government, state and local governments, industry, commercial businesses, academia, and more."
Contracts management professionals strive to: (1) Manage customer and supplier expectations and relationships, (2) Manage budgets, (3) Control risks, (4) Manage costs, and (5) Contribute to organizational success.
Contracts management integrates a broad set of business disciplines and involves working closely with all areas and departments within an organization.
Contracts management involves the proactive management of a wide variety of commercial, international, government, collaborative, consortial, relationship, and other agreements in accordance with best practices, applicable entity policies and procedures, compliance programs, applicable laws, and the other requirements of the parties.
It generally encompasses contract planning, contract designing, contract drafting, contract negotiation, contract execution, contract summary document, brief personnel on contract contents, contract insurance management, subcontracts, creation and management, modifications, contract administration, use of best practices, export compliance, contract close-out, contract assessment, lessons learned activities, and other similar items.
AI/ML aspects will improve contracts management as follows:
1. Digitizing contracts and important letters and other documents;
2. Creating a contracts central repository;
3. Improving organization strategy; and
4. Better managing legal, contractual, operational, and performance 'RISKS' (with contracts management software, 5. limiting project access, 6. limiting contractual access, 7. limiting other access, 8. tracking milestones, 9. complying with confidentiality requirements,,
10. complying with IP requirements.
11. complying with insurance requirements, 12. understanding the applicable legal compliance issues,
13. tracking KPI's, and other similar activities.
Contracts Management personnel will not be replaced by AL / ML but rather Contracts Management personnel will be assisted by AI / ML.
USE OF AI RESEARCHING VERY LARGE ANTARCTIC ICE HOLES (“POLYNYAS”}
After a portion of the antarctic sea areas freeze, there are certain sizable open areas in that sea area that never freeze. These unfrozen (or very thin ice) sea areas are called “polynyas.”
Polynyas are quite important in antarctic sea areas for: (1) helping to maintain marine life in, or near, those open sea areas and (2) creating ocean cooling in those polynyas sea areas by expelling ocean heat.
So, a sizable number of whales, walruses, seals, polar bears, and other sea animals, as well as certain birds, tend to live in, or near, those open polynyas sea areas. Also, the unfrozen polynyas sea areas maintain the seas, in the area of polynyas, at temperatures above freezing in the polynyas area.
Some researchers believe that heat escaping through polynyas may potentially impact atmospheric temperatures, wind patterns, and rainfall world wide.
Recently this year, a team from University of Washington has discovered how polynyas are created as a result of the Team’s analysis of the data (Big Data) that polynyas provided due to: (1) a combination of intense storms; (2) an underwater mountain; (3) saltier water, and; (4) certain unusual ocean conditions. The team used floating artificial intelligent (AI) robots, satellites, and specially trained elephant seals (that can dive up to 2 kilometers) fitted with harmless temporary head sensors to gather the data.
Be Sure To See: https://www.youtube.com/watch?v=90YxgaTL12M
Here are a few additional facts and issues about polynyas:
The size of polynyas varies from small holes to holes that suddenly appear equal in size to Connecticut, Maine, New Zealand or South Carolina.
The many holes are in the open mid-seas away from shore. Other holes are coastal in nature.
The holes seem to be appearing about every 40 years (such as the Waddell Polynya).
The exact location and size of the polynyas are closely monitored by the US Navy and other navies of the worid (friend or foe)
since they are needing to surface periodically for one reason or another.
The traditional method for a nuclear sub to surface through solid sea ice is to nuzzle up slowly to the bottom of the ice and then, while resting just below the ice, suddenly burst through 3 to 9 feet of solid ice. Surfacing through an existing polynya is much simpler, faster, and safer. Of course some of the sizes and locations of polynyas are military classified information.
Of course, polynyas appear in Arctic seas as we'll as off the coast of Greenland, Iceland, Canada, and other locations.
Check Out This Great Video (Linked Below) About the Polynyas Discussed in This Note Above:
More and more AI robots are being used in Antartica today ( as well in the Arctic areas) for various civilian, research, and military purposes due to the lengthy extreme cold weather conditions and the on-going dangers to humans.
Facebook, Google and Stanford University have created “AI Ethics Research Centers”.
Canada and France have jointly created an international panel to discuss AI's "responsible adoption."
The European Union (EU) has its own seven (7) guidelines seeking a "trustworthy AI.”
A summary of the EU's seven (7) guidelines is stated below. One can read the full PDF of the guidelines and the below summary here at:
- EU Guideline Summary -
• Human agency and oversight: AI systems should enable equitable societies by supporting human agency and fundamental rights, and not decrease, limit or misguide human autonomy.
• Robustness and safety: Trustworthy AI requires algorithms to be secure, reliable and robust enough to deal with errors or inconsistencies during all life cycle phases of AI systems.
• Privacy and data governance: Citizens should have full control over their own data, while data concerning them will not be used to harm or discriminate against them.
• Transparency: The traceability of AI systems should be ensured.
• Diversity, non-discrimination and fairness: AI systems should consider the whole range of human abilities, skills and requirements, and ensure accessibility.
• Societal and environmental well-being: AI systems should be used to enhance positive social change and enhance sustainability and ecological responsibility.
• Accountability: Mechanisms should be put in place to ensure responsibility and accountability for AI systems and their outcomes.
This 7 item Summary of the EU Guidelines can be found
on the Engadgetb web site - "The EU Releases Guidelines To Encourage Ethical AI Development"
Earlier this year on February 19, 2019 President Trump issued an Executive Order on US AI leadership that can be found on the below Harvard Law web site:
Please note that the US federal government did not provide any additional AAI funding when it issued these guidelines
Powerful “Dual Use” AI Technologies Used for Good or Evil
There is a growing debate in AI circles today around an urgent question: What do we do with increasingly powerful “dual use” AI technologies that can be used for good or for evil?
The conclusions of this scientific debate will largely determine how future AI technologies that could cause widespread harm — or could cause widespread good: (1) be released into the world for wonderfully useful democratic research purposes; or (2) be withheld from the world due to the potential for drastic societal impacts.
Some academic AI researchers say that it is too early to be withholding any such research in new AI technology. They believe that withholding would slow down the AI research community. and that we are very far from possible AI risks.
However, some researchers say that a decision to withhold such research is the responsible and ethical thing to do and they feel that such withholding is vital to long term safety.
The time is now for you to take part in the discussion on the dual use of emerging AI technologies.
AUGMENTED ARTIFICIAL INTELLIGENCE (AAI) AND MACHINE LEARNING (ML) WILL SOON IMMENSELY AFFECT SPORTS
The Good United States Artificial Intelligence Group
(Founders - Johnny Miller, Dr. Tim Sheng, and Russ Peterman)
The changes made by the use of Augmented Artificial Intelligence (AAI) and machine Learning (ML) in the very near future will be enormous. It will revolutionize sports. It will elevate sports. It will be much more than just the pioneering use of statistics and analysis used now in sports. It will help owners, managers, coaches, players, trainers, and other sports personnel to make smart, astute, wise, and timely decisions before, during, and after the applicable sports game.
It will use a multitude of novel AAI and ML platforms using highly sophisticated AAI and ML technologies. It will collect huge amounts of structured (“Clean Data”) applicable data ("Big Data") to create and power the necessary algorithms to implement the AAI and ML needed.
It will predict the chances of success in numerous sports situations for various sports game tactics and plans. Football, baseball, basketball, hockey, soccer, wrestling, field & track, tennis, swimming, motor vehicle racing, and other sports coaches will be turning to AAI and ML to help them call the right plays/moves in real time and to use the right planning, strategy, and execution during a sports event/game. It will also greatly assist in individual player/participant thinking, focusing, analyzing, positioning, execution, improvement and performance.
It will assist managers, coaches, trainers, and other sports personnel to devise better education, equipment, training, health, injury recovery, nutrition, strength, speed, and conditioning programs for each of their individual players/participants. It will greatly improve player development.
It will greatly enhance player recruitment, analysis, and player /participant scouting by managers, coaches and other sports management personnel.
It can even enhance the performance of referees, judges, umpires, and other sports competition rule providers.
It will disrupt the future of sports - in a very good way. While humans may have somewhat peaked in analyzing sports, AAI and ML are just getting started in the sports arena. Together - humans and AAL / ML can collaboratively develop and implement a highly effective model for future applications of good AAI and ML in sports.
One can often hear people today say: “Big Data is a grand prize.” It is referred to by many folks as the … “new oil.” Some say it is the world's most valuable resource. The overwhelming opinion today among artificial intelligence (AI) and machine learning (ML) experts is: “The entity with the most data can build the best AI and ML.” Others say that those who rule Big Data will rule the world. Big Data, of course, is an absolutely essential item in using AI and ML..
The purpose of using Big Data is primarily for training AI and ML applications to function very efficiently. You could say that today there is clearly a strong preoccupation of thought by entities to use Big Data. Just take a look at all the companies that are today feverishly trying to re-brand themselves as data companies.
However, check this new concept out: The future of AI and ML is likely to be far less data-intensive than it is today. What ? I s that true? Yep. Some of the AI and ML leaders in this current wild west style of AI and ML frontiers this are now developing improved forms of AI and ML that do not require such massive amounts of labeled data. Some leading AI and ML companies are currently developing highly reliable innovative methods to fabricate high quality data at very low costs. Such data will be molded to the exact needs, such as, creating billions of AI-ML required alternative scenarios. It is called Synthetic Data and it does not require such massive amounts of labeled data as are being used today. Synthetic Data is approaching real-world data accuracy.
In order for machine intelligence (AI and ML) to approach human intelligence in its capabilities, it should be able to learn and reason from a handful of examples much the same way that humans do. This is the goal of an important new field within AI and ML known as "Few-shot Learning" or “One-Shot Learning” where much smaller amounts of data are needed. Few-Shot Learning or One-Shot Learning refers to problems where the model is given only one instance (or several instances) for training data and has to learn to re-identify that instance in the testing data. A popular example of Few-Shot or One-Shot Learning is found in facial recognition systems.
In another new area of AI and ML called Reinforcement Learning, an AI and ML model does not learn through currently used brute-force data ingestion. It learns through AI and ML self-guided trial and error. It is let loose to experiment with different actions in a given environment. It gradually’ optimizes its behavior as it receives feedback about which actions are advantageous and which are not.
Here Are Are Some Reasons Why Synthetic Data Is Used:
*Overcoming real data usage restrictions: Real data may have usage constraints due to confidentiality, privacy or security rules or other regulations. Synthetic Data can replicate all important statistical properties of real data without exposing real data, thereby eliminating issues.
*Creating data to simulate not yet encountered conditions: Where real data does not exist, Synthetic Data is the only solution.
* Immunity to some common statistical problems: These can include item nonresponse, skip patterns, and other logical constraints.
These Synthetic Data benefits (just as few of the many) demonstrate that the creation and usage of Synthetic Data will only stand to increase as our data becomes more complex and more closely guarded. As AI and ML get smarter and smarter in the three AI and ML sub-fields of (1) Synthetic Data, (2) Few-Shot Learning or One Shot Learning and (3) Reinforcement Learning, AI and ML will then require less data not more data.
In providing artificial intelligence (AI) outsourced services by the AI contractor / seller for the enterprise buyer or functional department head buyer, the following AI contractor performances objectives might be required:
1. Positive completion of an initial small “Sand Box” type Phase demonstrating how valuable insights and new options can be analyzed and extracted from buyer’s data;
2. Upon positive completion of the positive Sand Box phase, the contractor might next be required to do a POC (proof of concept) Phase on a larger scale than the Sand Box Phase, that substantially confirms the viability of obtaining valuable insights and new options by analyzing the buyer’s data; and
3. Upon completion of the POC Phase, the definitive Final AI Project is completed by analyzing a large volume of buyer-data and clearly extracting valuable insights and various new useful options from the data. I am assuming that the buyer would provide the seller/contractor with the applicable buyer data adequately scrubbed so that it is structured (not unstructured) data and ready to be analyzed. Also, I am assuming that the work would normally be performed on a "time and materials" basis (not firm fixed price). The parties would agree on a period of performance. The buyer could cancel the Project at the completion of the Sand Box Phase or POC Project Phase. The seller/contractor would only be required to exert reasonable efforts to complete the work within the budgetary limit and period of performance. The parties would agree on a monetary/budgetary limit. Once the seller/contractor spending reaches the budgetary limit or the end of the period of performance, the seller/contractor could contractually suspend performance unless the buyer provides additional funding and/or extends the period of performance. Warranties would be minimal since it is unknown if the seller/contractor will discover will discover anything of value in buyers' data. Seller/contractor might want to have a non exclusive, royalty free, paid up, world-wide, sub-licensable, transferrable license from buyer in order for seller/contractor to potentially use (not own) the buyers' data in future seller/contractor projects with its other customers. There should be reasonable limits of seller/contractor liability. There should be a mutual confidentially provision. There would of course be other provisions - as required.
"Do it -- Fix It -- Try It"
Chaotic action is preferable to orderly inaction
This web site will always be a ..."Work In Progress".
Changes, updates, additions, and eletions will be made weekly. The last update to this web site was made on:
February 7, 2020
(C) COPYRIGHT JOHN ("JOHNNY") E . MILLER, RUSS PETERMAN, AND QIN ("TIM") SHENG
2018 - 2020
ALL RIGHTS RESERVED EXCEPT AS NOTED HEREIN WHERE ANOTHER PARTY OWNS A COPYRIGHT INTEREST