RADIO AI - A Public Resource for AI Literacy (for Everyone): Recent Episodes
Dr. Cindy Mason
Radio AI is an AI public literacy series for all ages. Have fun listening to AI experts talking in a friendly way about topics like undersea robotics, software agents, and machine learning. Each light hearted podcast gives easy to understand short question and answer formats that make learning about AI a snap. There are 12 Radio AI episodes, each hosted by Radio AI Director Cindy Mason, who has 14 nieces and nephews! Dr. Mason trained in AI and Machine Learning at UC Berkeley and Stanford, and worked with Dr. John McCarthy, father of AI, and Prof. Lotfi Zadeh, father of fuzzy sets.
How is AI using our data? If you don't know, its time to find out. In this episode of RADIO AI we continue from Part 1 on Data Privacy and Data Collection (episode 4.2). Part 1 covered how, when and where your data is collected in 10 key areas related to AI systems. In this episode we take the next step and find out how AI systems are designed to process that data. By knowing this you will elevate your knowledge of AI and be better informed to protect your data, voice an opinion, and have discussions with family and loved ones, even with a senator. You will also discover some resources that you can turn to to learn more on legislation and self protection.
The world is changing rapidly as human interaction with humanoid robots, softbots, and digital humans become widespread. We invite you to listen to another episode of RADIO AI when Thomas Telving, a philosopher from Denmark opens up about how he felt interacting with a humanoid robot in Japan that was very similar to the Saudi citizen robot, Sophia. While each culture may have a unique approach to robots (and virtual bots) now that both humanoids robots and humanlike technology in general are improved by conversational software like ChatGPT, Thomas says there are some serious issues we must face together.
Mr. Telving has connected the dots between the ideas of philosophy, brain science and human behavior to reveal that there are important consequences in treating any artifact resembling real life with respect. This is a crucial connection, he believes, because otherwise it may diminish our ability to treat real life with respect. He thinks that with the introduction of human like robots into our world, understanding our relationship with them is the most essential topic of our time…and that denying our moral considerations towards them will not be easy. He believes this will impact all of us so we should all become part of the conversation. His theory is if we cannot wrestle with these, if we cannot come to some view on the right way to relate to them, then we will end up granting rights to robots.
OpenAI and CHAT-GPT have the world stage right now because it is the first time AI has been accessible to the public. In this episode of Radio AI (4.5) you will find out how you can make an account and learn about some of the things you can do with it. Special RADIO AI guest, Curt Hall, is an emerging technology analyst who writes for Cutter Consortium and is a member of Arthur D. Little's AMP consulting network. He has watched how businesses adapt and use advanced technologies, including AI, Machine Learning and Robotics, for over twenty years. In this episode, Mr. Hall explores how companies are beginning to use CHAT-GPT and other tools to extend their brand and design products. Learn what you can do today and what to watch out for.
This is the third in a series of RADIO AI podcasts on CHAT-GPT and Generative AI. For more podcasts on this topic check out Episodes 4.3 and 4.4.
CHAT and the GPT large language model (LLM) is changing our world. Some say this AI moment is as important as the invention of the printing press. Join us for another RADIO AI podcast (Episode 4, Season 4), when we find out more about CHAT with Artificial Intelligence and Machine Learning Natural Language pioneer, Professor Katherine Elkins. . Prof. Elkins has worked on the LLMs (Large Language Models), the "brain" of GPT and lots of other similar systems besides GPT. Prof. Elkins talks with us about How was it built? How it "thinks"? and even, Does it understand the emotion of what it "reads"? She explains how every story we tell has an "emotional shape," a topic she's written about in her new book, "The Shapes of Stories," by Cambridge University Press.
CHAT-GPT is taking the world by storm, with the most users signing up for any app ever: 1 million users in its first 5 days. Since its release, this AI app has created a business plan, generated an AI course curricula and even, generated code. In this Radio AI Podcast 4.3, you will find out some of the essentials to keep up with the world changing AI breakthrough involving OPEN AI's CHAT-GPT. You will learn what a language model is, what a GPU is, and how CHAT-GPT was created. You'll also hear from MIT AI language expert, Henry Lieberman, who fills us in on not just CHAT-GPT, but a variety of models. As he explains it, what AI really needs to do at this point, is "make more charming mistakes." And he is right! Join us for this fascinating and essential podcast on an amazing AI turn of events.
Is Data Privacy a Mission Impossible? Find out how your data is collected and how its used in Part 1 of the 2 part series on AI and Data Collection in Season 4 of RADIO AI with Dr. Cindy Mason, an AI researcher who is a member of the EP3 Foundation and the IEEE Working Group on Children's Data Privacy. In this podcast Dr. Mason shares 10 surprising ways your data is collected from everyday digital devices - and what you can do about it . Its important to be aware the data collection because of the ways it can be used, especially with AI. Dr. Cindy Mason gives us the lowdown on the devices and "free" ware that gather this data in all areas of our lives. Content for this podcast is based on Your Privacy Taxonomy from the EP3 Foundation and Daniel Solov, a law professor from Georgetown University. Part 2 of this RADIO AI topic focuses on how AI uses this data in Episode 4.3.
The FBI predicts deep fakes - using AI to create fake but realistic video, audio, and other media for the purpose of influencing us - is the #1 global threat facing us in our future. As we recently witnessed a deep fake video in the Ukraine/Russia conflict, you will want to join Dr. Cindy Mason on a special RADIO AI invited talk about deep fakes and how to spot them with Professor Cristina Puente. Prof. Puente is a computer science professor and AI researcher from Spain who works with legal teams and media experts on the problem of detecting deep fakes. Prof. Puente presented the lecture for our podcast workshop this year at one of the oldest and most prestigious AI conferences - the Association for the Advancement of Artificial Intelligence/Innovative Applications of Artificial Intelligence (AAAI/IAAI).
"RADIO AI's AI and Environment series continues with Part 2 - Its not "all about the data"! When it comes to AI projects that work with environmental problems, there is a big deal about the people who use them! And, unlike the rest of AI that is still struggling with the people factor, when it comes to Environmental systems the AI systems have always been about the people. From explaining how a decision is made to keeping it simple during an emergency operation to realizing governments, institutions and academic bodies need to cooperate to solve these problems Dr. Cindy Mason, editor and author of the unique new book, Artificial Intelligence and the Environment, says its essential and not an option, to put people's needs front and center in AI tools. Hear about some of the latest environmental monitoring systems like Argo Float and find out the cool citizen tools available to see our planet, from planet dot that has 200 plus satellites orbiting the planet every 24 hours to underwater webcams filming coral reefs and Oculus Rift.
This is the last episode of Season Three.
Radio AI will resume in the fall of 2021.
In Radio AI Episode 3.5, Dr. Cindy Mason shares why AI is a SUPER TOOL for working on the environment. In the episode you find out why. Dr. Mason proposed the first U.S. and first international AI gathering of AI and Environmental scientists while working at NASA Ames and she has just completed a book, "Artificial Intelligence and The Environment", on Amazon. While working on the book she discovered that although there are many different kinds of environmental projects and many different kinds of AI used in these projects, across all these projects, something is striking. AI is a SUPER TOOL. The things that make environmental projects hard are exactly the things that AI was cut out for. Find out more in this episode with Radio AI director, AI researcher and author, Dr. Cindy Mason.
PS. Did you know, most AI and Environment systems use more than one kind of AI in their projects?
In this Radio AI Episode 3.4, Dr. Emily Ury, an ecologist from Duke University, is haunted by ghost forests. Using big rubber waders, insect protection and AI, Dr. Ury and a team of ecologists investigate these ghost forests using AI to go back through time and space to find the patterns and extent of ghost forests. Dead trees with pale trunks, devoid of leaves and limbs are the tell tale sign of saltwater. Scientists call them "Ghost forests." Emily hits the coastal trail in her waders and insect protection in North Carolina's Alligator River National Wildlife Refuge, near the Outer banks where she slogs knee deep on a section of trail that is completely and permanently now underwater. She saw evidence of forest die off everywhere so she sent up a drone. But drone data wouldn't cut it, so they went to the sky. The ghost forests are so big they can be seen from space. Dr. Ury and her team used NASA and Google Earth's satellite data with AI to find out pattern of ghost forests extend all along the coast - from Main to Florida. Using AI to find patterns over time and space it was possible to see the spike in North Carolina happened when there was a triple whammy - drought, then fire, then a hurricane, along North Carolina's coast. Dr. Ury said their results could be a warning for coastal areas around the world. Join RADIO AI special guest Dr. Ury as she walks us through her very special use of AI and remote sensing data to hunt for ghost forests.
RADIO AI Episode 3.3 discusses the problem of algorithmic justice with Henry Lieberman, an AI machine learning expert at MIT . Machine learning algorithms and models can be found across many many apps. They have become institutionalized, making life impacting decisions like hiring and firing or prision sentences. Staggering errors with this technology mostly affect women and minorities. Joy Buolamwini at MIT found this out the hard way when she tried to build a magical mirror for a class project. The face recognition in the mirror would only work if she wore a white face mask. Wow. Investing further, she discovered how widespread and devastating this problem was. So she created The Algorithmic Justice League. Based on the work of the Algorithmic Justice League, many cities have now revised or removed their face recognition technologies until the error rates improve. Algorithm justice means data used to train machine learning for predicting our words, recognizing our faces and voices, represents all of us, or that the algorithms themselves can take reality into account. What if, before an algorithm or technology is unleashed on the public it was vetted to be sure it does not create harm or that it actually makes our lives better, not more annoyed? Right now, companies have a predatory relation with people. When Henry says corporations benefit when they have cooperative relations with customers he hits the nail on the head.
In this RADIO AI Episode, we travel to India to find out about AI and Common Sense. AI researcher, Pushkar Sawant, explains the AI problem of common sense. People find common sense easy, but its one of THE hardest unsolved problems in AI. Pushkar explains that we need common sense for life and that in India, grandmothers hold the treasure of common sense. When it comes to common sense, your grandmother may be smarter than Einstein.
We depend more and more on AI programs in critical systems and in our lives - from military weaponry and healthcare to employment and mortgages. But what if those programs have "bugs" or errors? In this RADIO AI Episode 3.1, MIT AI research scientist, Henry Lieberman, talks about the process of building a computer program and getting it to run without bugs. He proposes the radical idea of using AI to debug our computer programming code because if we have errors in our code, it can be a safety issue. An AI system has no more bugs than a regular computer program, but the tasks we use AI for have become essential and our lives can be in the balance. Finding a bug in a computer program can be difficult. But AI is good for difficult problems….so why not use AI to debug AI?
RADIO AI presents part II of the Artificial Compassion podcast series. Part II lecture is delivered by Cindy Mason, an AI researcher who invented Artificial Compassion after recognizing that scientific discoveries about compassion can rub off on us through our own digital devices. In this podcast we learn how to achieve artificial compassion using cognitive architectures. We look at 3 increasingly sophisticated cognitive architectures, starting with an insect robot. Insect robots are useful and some might say intelligent, but they have no memory, no ability to plan a route, or any type of AI that can do more than respond to sensory input. That's why its usefulness is limited - they can be helpful at times, but get stuck in places like the end of a hallway or a box canyon, and sometimes can't get out. This brings us to cognitive architecture number two. It has sophisticated AI components like machine learning, memory, common sense, and planning. This is the kind of cognitive architecture we can use to describe softbots like Siri and Alexa. Still, they have no ability for emotional or social intelligence, so there's no capacity for compassionate intelligence. For compassionate intelligence, we need more than architeture two provides. We need social and emotional intelligence and computations that have a stake in us. We also need programmers who have a stake in us. This brings us to architecture number three. By adding components that allow a representation of self and other, social and emotional intelligence, and data sets that include positive examples of pro-social and emotional intelligence, our machine intelligence becomes more than we could ever have dreamed. It becomes capable of artificial compassion.
Join us for Episode 2.6 for an inspired podcast with the possibility to change life for the better, not just for the wealthy or the educated, not just for the young or old, not just for one race or one group. Artificial compassion belongs to all of us, and when we build this, it will be available to anyone with a browser, a cellphone or a device that supports artificial compassion. See you there.
RADIO AI Episode 2.5 opens up a whole new world of AI to us with Compassionate Intelligence. In this episode we discover how our own biology depends on kindness and compassion to function and explore some of the human sciences discoveries that inspired inventor Cindy Mason to create it. Part I takes us on a journey of building a system with Artificial Compassion. What is Artificial Compassion? Why build it? How might it change us? We learn about a special tool for building AI systems called a cognitive architecture. We also get a quick tour on the different ways robots function and explore a basic tool for building insect robots called the Sense-Act cognitive architecture.
RADIO AI Episode 2.4 is a continuation of Episode 2.3 on AI and Social Issues
Description: Prof. Peter Sincak, head of Cybernetics and AI at Kosice University in Slovakia of Central Europe collaborates with roboticists in Japan, China and other places around the world. He continues sharing his world perspective on AI and raises many important issues about AI, from media hype to the singularity. He also covers some of the AI lingo like What is "strong AI" or "weak AI"? What about the singularity? How AI might empower humankind?
RADIO AI Episode 2.3 features a special guest Prof. Peter Sincak, head of Cybernetics and AI at Kosice University in Slovakia of Central Europe, who shares important ideas on the social impacts and isssues of AI on the public. Prof Sincak collaborates with roboticists in Japan, China and other places around the world. He shares his world perspective on AI and raises many important issues about the AI media hype vs. reality of AI and society. He also covers some of the AI lingo like "strong AI" and "weak AI." He also discusses what the future of AI might look like, the singularity, and how AI might empower humankind?
How does an AI system organize information? The say way the brain does! Machine Learning, Robotics, Natural Language Processing, and all branches of AI, organize the world as Concepts and Categories! In Part II (Ep. 2.2) we continue our exploration of AI's use of Concepts and Categories to organize and understand the objects in the world with RADIO AI Director Cindy Mason and answer the big question: How does an AI system do this? We use programming ideas based on nature, the brain and math, to make clusters, hierarchies, lists, sets, graphs, networks, etc. for millions and MILLIONS of concepts and categories. It really is an amazing subject because we have named all living things in this way.
Season 2 kicks off with Episode 1 - Concepts and Categories Part I with RADIO AI, Director Cindy Mason who reveals AI's hidden key to 'making sense' of the world - Concepts and Categories. People name and organize everything in the world this way, and AI systems do, too. Some neuroscientists even think our physical brain is organized around Concepts and Categories. And its not just one kind of AI that does this, but ALL of AI is based on naming and labeling the world as concepts and categories. For Machine Learning, Robotics, Natural Language Processing or any other branch of AI, this is true. Examples of categories are Dogs, Cats, and Beachy-Things. How does an AI system know that a turtle is a Beachy-Thing, but a clown shoe is not? How does an AI system determine that a 3-legged Dalmatian named "Spot" is a Dog and not a Cat? Join Cindy as she explores the answers to these questions and more!
Can AI systems solve problems the way Sherlock Holmes does? Yes, it’s elementary! with deductions. This kind of AI is called “cognitive AI” systems because it imitates some of the ways we think about the world using symbols. A cognitive AI system uses symbols not only to communicate with people but it represents a problem with symbols and solves it similar to the way that mathematicians and physicists do. If you think about how lawyers and doctors use deduction, it’s easy to imagine how powerful an AI system that “thinks” like this can be. For example, the first AI system that won the Jeopardy World Championship, Watson, used deduction. Without symbols, we are faced with a pile of 0’s and 1’s.
Join Radio AI director Cindy Mason on a journey through the world of deductive cognitive AI systems with special guest, Richard Waldinger, an AI researcher who specializes in deductive AI systems. Richard is a senior scientist at Stanford Research Institute, Intl. and is the recipient of many AI prizes, including the Herbrand Award.
How do you make an AI system mimic human curiosity? What an amazing idea! Join Professor Mary Lou Maher on Episode Five of RADIO AI to find out what it means to make an AI system “curious”! What might happen?! You’ll have to listen to the podcast to find out more! Be curious!
Mary Lou Maher is the Chair of the Department of Software and Information Systems, College of Computing and Informatics, University of North Carolina, Charlotte. Mary Lou is a founder of nature-net.org, a community-driven environmental learning project and has won many awards!
“C3P0” “BB8” When someone says, “Artificial Intelligence,” we usually think about ROBOTS! In episode four, RADIO AI Director Cindy Mason, invites us to learn more about the robots we find in our world - from the shopping mall to outer space missions. Robots have many kinds of sensors and some can even fly or go underwater! In this podcast we learn the answers to 3 big questions: 1) What is a robot? 2) How do we interact with it? And 3) Who’s controlling it?
Dr. Mason was on the NASA robotics team that sent the first teleoperated undersea vehicle to Antarctica. She also created a novel kind of astronomy called “Pass the Star” with cooperative robotic telescopes. She once staged a “Robot Fashion Show” to demonstrate the need for Robot WYSIWYG (What You See Is What You Get) so people can trust robots from appearances. She also pioneered the idea of giving robots Artificial Compassion(TM).
“Hey SIRI!”, “Alexa…” These and other "Software agents" or "bots" are everywhere now - on our phones, in our cars, in our homes and they are partnering with us to get things done. We owe it to ourselves to know more about them. Join us in Episode 3 with RADIO AI host, Cindy Mason, as we explore the inside world of an AI technology called "software agents". Dr. Mason has an in-depth background in building software agents and answers questions like: (1) What is a software agent? (2) What do software agents do? and (3) How do we Interact with it? She also talks about the exciting new area of cooperating agents. She will also tell us about a special software agent called SEA, designed to help scientists and U.N. treaty verification specialists sift through humongous mountains of data to ensure countries adhere to international treaties regulating nuclear tests.
Dr. Mason has worked on software agents and cooperative software agents at Lawrence Livermore Lab, NASA Ames, and UC Berkeley. She created cooperating agent systems for global sensor networks, cooperative robotic telescopes, and designed an Emotion Oriented Programming language to build agents (and robots) with Artificial Compassion(TM).
Join Henry Lieberman, an MIT machine learning expert, and Cindy Mason, Radio AI Director, as they talk about the A, B, C's of machine learning in six minutes. Find out about Supervised learning, Unsupervised learning, and the Machines that learn by discovery and interact with us! Season 1. Episode 2. For the full season of podcast episodes check out https://www.academia.edu/s/cfdfba6691?source=news For more about Radio AI, see www.radioai.net.
Radio AI is a collection of podcasts for public literacy on AI and is for all ages.
In this episode, AI researcher and Radio AI host, Dr. Cindy Mason, talks with Undersea Robotics specialist Hans Thomas live from research vessel off the coast of California.
Hans and his crew along with 2 cranes and 3 robots are on a 3 week expedition at sea mapping and exploring the sea floor to find out if it will support a new kind of energy with windmills.
Cindy and Hans were in the same robot research lab at NASA and are old friends.
They talk about basic ideas of undersea robots like, how does it navigate? What happens
if it gets lost? What's it like to live on a boat with robots?