On this episode of AI Australia, we’re changing things up a little. With James overseas, Nigel and his friend Sarah Turner from REA Group hosted multiple panel conversations with some of Australia’s best and brightest in the field of AI, computer science, mathematics, and regulation to discuss the launch of OVIC’s new book: Closer to the Machine.

OVIC is the Office of Victorian Information Commissioner and is the primary regulator and source of independent advice to the community and Victorian government about how the public sector collects, uses and shares information.

In this episode, we’re lucky to be joined by:

• Sarah Turner (co-host) - General Counsel, REA Group
• Adam Spencer - comedian, media personality and former radio presenter
• Rachel Dixon - Privacy and Data Protection Deputy Commissioner, OVIC
• Professor Toby Walsh (University of New South Wales and CSIRO’s Data61) 
• Professor Richard Nock (Australian National University and CSIRO’s Data61) 
• Associate Professor Ben Rubinstein (University of Melbourne) 
• Katie Miller (Independent Broad-based Anti-corruption Commission) 
• Professor Margaret Jackson (RMIT) 

In this episode we discuss:

• The degree to which we all take for granted how big a part AI plays in our lives
• The rate of improvement in algorithms in their narrow fields of “expertise”. We discuss how quickly a chess-playing AI went from basic to beating the chess grandmaster Gary Kasparov
• OVIC’s motivation for publishing the book on data privacy and protection
• Grappling with the implications of how AI systems can be misused, or easily breached from a security standpoint. Do we continue to push the boundaries in the face of privacy risks and concerns? Do we pull back?
• The challenge of discrimination. Eventually, machines will need to make decisions that “discriminate” against people in one way or another - but there is such a thing as good discrimination and bad discrimination. Who gets to make those definitions?
• What role should government play in the regulation (or non-regulation) of AI?
• Accountability. What happens when AI “goes rogue”? Where does the buck stop?
• How the conversation has evolved over the years, and become more “realistic” in a sense.

Links mentioned:

OVIC: Closer to the Machine book

Dr Karin Verspoor works at the intersection of Science and Technology, applying computation to analysis and interpretation of biological and clinical data, particularly unstructured text data.

Karin is a Professor in the School of Computing and Information Systems at the University of Melbourne, as well as the Deputy Director of the University's Health and Biomedical Informatics Centre.

She was previously a Principal Researcher at NICTA's Victoria Research Lab and served as the Scientific Director for Health and Life Sciences. Karin headed a research team at NICTA in Biomedical Informatics.

Karin moved to Melbourne in December 2011 from the University of Colorado School of Medicine, where she was a Research Assistant Professor in the Center for Computational Pharmacology and Faculty on the Computational Bioscience Program. She also spent five years at Los Alamos National Laboratory, nearly five years in start-ups during the US Tech Bubble, and a year as a Research Fellow at Macquarie University. She received her undergraduate degree in Computer Science from Rice University (Houston, TX) and her MSc and PhD degrees in Cognitive Science and Natural Language from the University of Edinburgh (UK).

Topics covered include:

• Karin’s take on the future of healthcare and the role will AI play, given her unique vantage point on the topic. We also cover the key building blocks that make this future possible
• Whether there are new risks associated with a more technologically advanced healthcare future
• We’ve seen public outcry over the My Health Record program with people opting out and question marks over doctors willingness to upload data to the system. We discuss what this means for the future of healthcare
• Whether tech companies like Apple and Fitbit will become significant players in the health space
• Karin’s take on what AI will do to the job market in healthcare if AI continues to advance at its current rate
• Is more regulation around data - eg GDPR or similar - a must-have for Australia if we want to embrace a more data-driven approach to healthcare?
• Where Australia ranks in terms of governments and companies leading the charge on transforming healthcare
• What Karin’s colleagues in medicine think about the rapid pace of technology innovation

On today’s episode of AI Australia, we have the pleasure of speaking with Suelette Dreyfus renowned researcher and journalist on digital privacy and whistleblowing. She has a personal connection to these topics and even co-authored a book with Julian Assange. Suelette will explore with us the implications of AI on privacy, the complexity of context able to be extracted from collected data thanks to this technology, as well as tech’s impacts on personal security. She will also cover how technology is changing the world of whistleblowers, what protections citizens have against predatory data collection, and much more. Join us on today’s episode!

More about Suelette Dreyfus:

Suelette Dreyfus is a journalist, technology researcher,  and writer. Her field of research includes information systems, digital security and privacy, the impact of technology on whistleblowing, health informatics, and e-education. 

Her work examines digital whistleblowing as a method of freedom of expression and the right to dissent from corruption. 

Suelette is also a researcher and lecturer in the School of Information and Computing Systems at the University of Melbourne. She is also the principal researcher on an international project testing the impacts of digital technologies on whistleblowing

Additionally, Susan is the author of the ‘97 cult classic Underground: Tales of Hacking, Madness, and Obsession on the Electronic Frontier which covers the exploits of a group of hackers in the late ‘80s and early ‘90s—including Julian Assange who is also the co-author of the book. 

• 6:00 Suelette describes the tradition of privacy in various parts of the world. 
• 7:00 Suelette discusses the upswing of encryption and whether or not law enforcement has always had access to this information, citing the “right to whisper.” 
• 12:00 Can the government read our minds? How much does our Google search history reveal? Can you simply use another search engine? What parts of our privacy and ourselves should we be willing to sell?
• 20:00 Do policies such as GDPR offer protections on things like genetic data? What happens if your data is sold, but it’s wrong? Can you correct it? What legal measures are coming into being surrounding these topics? Are there laws surrounding facial recognition technology?
• 30:00 What sorts of legislation may come about in Australia in regards to data privacy? How can data be used in nefarious ways such as targeting potential gambling customers based on genetic predisposition to addiction? 
• 34:00 Where is the line between reasonable persuasion and manipulation? How does this relate to politics? Social media? How do news filter bubbles normalize ideas? Is fact-checking an issue here?
• 39:00 Suelette discusses whistleblowing and relevant cases. What systems and laws are in place to assist whistleblowers in calling attention to wrongdoings? How are those options changing? Is whistleblowing about fame?
• 48:00 Suelette discusses Australian legislation that allows for backdoor access to data. What was wrong with this recent bill?
• 55:00 How are companies choosing to censor people? 
• 63:00 When did Suelette decide to pursue privacy in her career? How did Julian Assange impact her path? What type of environment and situation is Assange dealing with? What rights are at stake?

Amazon: Underground: Tales of hacking, madness, and obsession on the electronic frontier

Today we’re talking with Data Scientist Katherine Bailey and Professor of Philosophy Oisin Deery about bias in machine learning systems. Katherine and Oisin share their perspectives on why philosophy and ethics of AI is such a hot topic, the scandals engulfing the tech industry and how philosophy could be applied to chart a new course, and how philosophy can help us frame solutions to bias in machine learning systems.  

Katherine Bailey is the Natural Language Processing lead within Accenture Australia’s AI and Automation Engineering group. Originally from Dublin, Ireland, her background is in software engineering and data science, with over a decade in the technology industry, primarily in Canada, the US and Australia. Prior to joining Accenture she was Principal Data Scientist at Acquia, a Boston-based Software-as-a-Service company, where she led the company’s Machine Learning initiatives. Katherine speaks and writes regularly on the topic of A.I. and is committed to dispelling the myths and removing the confusion around it, teasing apart the real from the imaginary implications of these technologies, both practical and ethical.

Oisin is Assistant Professor in the Department of Philosophy at Monash University, in Melbourne. His research interests lie at the intersection of philosophy of mind and action, metaphysics, and ethics. Oisin also works on ethical issues related to artificial intelligence. Oisin completed his Ph.D. in philosophy in 2013 at the University of British Columbia. Katherine and Oisin have a particular interest in bias in Machine Learning systems and have presented at conferences and meetups around the world on this topic, including a presentation at Google in 2018.

In this episode we discuss:

• STEAM versus STEM, how the Arts complements STEM
• Exploring the feasibility of Artificial General Intelligence
• How philosophy can help us think about AI and ethics
• How we define intelligence
• The limitations of the Turing test
• The long history of machines fooling people
• Consciousness and sentience in relation to AI
• Is AI ushering in a golden age of philosophy?
• The tendency in tech to assume a question is being asked for the first time
• Tim Miller’s work on explainable AI and the importance of drawing on social sciences
• How can tech companies best apply philosophy?
• The recent scandal surrounding Google’s external AI ethics advisory council?
• What is ethics washing?
• How tech is impacting democracy and public debate
• Different types of bias in word embedding how it can be addressed

Resources mentioned:

• Life 3.0: Being Human in the Age of Artificial Intelligence https://www.amazon.com/Life-3-0-Being-Artificial-Intelligence/dp/1101946598
• Facebook’s Role in Brexit—and the threat to democracy with Carole Cadwalladr https://www.ted.com/talks/carole_cadwalladr_facebook_s_role_in_brexit_and_the_threat_to_democracy?language=en
• https://ai.google/principles

Today we’re talking with Aire co-founder and mum of the Rita AI, Sarah Bell. Sarah discusses with us the evolution of Rita, humanisation of machines, our sci-fi future, and the ethics of big data and bias-free AI.

More about Sarah:

As a co-founder of the PropTech startup Aire, Sarah Bell is a sought after speaker and expert on the strategic application of artificial intelligence and automation in real estate. She helps real estate businesses harness the opportunity within their data and transition to a blended human-digital workforce using Aire's Digital Employee, Rita.

With a professional and academic background in research and analysis, Sarah joined the front lines of real estate as an agency owner and practitioner for a decade giving her tactical insights, subject matter expertise and just enough street cred to create solutions that are intelligent by design and Cx obsessed.

Sarah is an author, analyst, researcher, project designer, speaker and nerdtrepreneur with research interests and some pieces of paper in strategic planning, change management, planning professional development and strategic application of artificial intelligence with London's Middlesex University, Leeds Trinity University (UK) and recently MIT's Sloan Management and CSAIL Executive Schools. Sarah is also Mum to three human children and a clever little robot named Rita. She has a strong preference for dogs over cats if pushed on the subject.

• How does Sarah see her role? Does she see herself as an entrepreneur? How did she end up in real estate?
• How did the AI Rita come about? How did they avoid creating a solution to a problem that didn’t exist? How did their background help with this? What does Rita actually do?
• What does Rita stand for? How has Rita become humanised among the team? Is this humanization deliberate or does it occur naturally?
• Tech jargon isn’t relevant to customers. How does the team help customers understand what Rita does? How does this let customers connect to her?
• Does Sarah have any practical advice for establishing an AI startup? What role does intuition play?
• Three important questions: How can we connect humans to computers in ways that make them more intelligent? How do we connect humans to each other in ways that are more intelligent? How can we connect people to information in ways that are more intelligent?
• Is it straightforward to hook into the right skills to take a concept or idea to an actual product? What challenges has Sarah faced on this front?
• Do some customers see Rita as a potential risk to their jobs? How can humans and machines work collectively with their best skills?
• Why is Rita built the way she is vs. as a chat bot? How does Sarah feel about chatbots as an interface? Will Rita one day talk directly to customers?
• What is Sarah’s vision of the future in relation to sci-fi—Star Wars, Star Trek, Jetsons, Terminator, Altered Carbon, etc.? Which sci-fi future are we most closely going to end up in?

Aire Homepage

Today, we’re chatting with artist and scientist J.J. Hastings. Her research explores self-experimentation, genome editing, machine learning and the future intersections between tech and the human body. She has long-standing roots as a biohacker—having co-founded two community labs, London Biohackspace and Melbourne’s BioQuisitive—and now has the first garage lab start-up in Australia to be approved by the OGTR to work with genetically modified organisms.

Her work has been exhibited at venues across Europe, India, Asia, North America, and Australia. J.J.’s career in scientific research spans over 15 years. She is an alumna of New York University, Harvard University, the University of Oxford, and Central Saint Martins with advanced degrees in Biology, Bioinformatics, and Fine Art. Her research fuses and folds together the fields of machine learning, bioengineering, space exploration, new media art, and ethics.

In our episode, we’ll probe J.J. ’s mind about the ethics of editing the genes of children, using gene drive to eliminate disease, the possible risks of biohackers. She’ll also cover how close she believes we are to artificial general intelligence (AGI), the impact of quantum computing on biotechnology, and the possibilities of our sci-fi-esque future, and her own lab—Bioquisitive.

• 2:00 How did JJ come to be who she is? How did she get involved in so many interesting projects? Is she an expert in all of these topics? And how did she end up in biohacking?
• 5:00 Has JJ chased being a provocateur? Or is it the product of pure curiosity?
• 7:00 JJ discusses the book Dark Emu (is this right?) and seed modification. How might this relate to our future settlement of Mars?
• 11:00 How are machine learning, CRISPR technology, and biohacking all interrelated?
• 14:30 JJ discusses theoretical biology, quantum computing, and whether or not biohackers pose a risk.
• 17:00 Is it out of control to utilize biohacking to cull species or eradicate disease? Will gene drive be widely used? How might this impact human and ecological health?
• 21:00 What is the history behind CRISPR? How is it still unfolding? JJ discusses the act of editing the genome of children and states her position on the scientist who experimented with exactly this. Is genetic engineering justified?
• 25:00 What are the ethics behind genetic modification? How far is the law behind the technology? How should we concern ourselves with our genetic privacy?
• 32:00 How does JJ feel about sci-fi? Which sci-fi movie most likely depicts our future? Gattaca, Altered Carbon, Star Trek, The Island?
• 38:00 How close are we to Artificial General Intelligence (AGI)? How does the accessibility to CRISPR and AI tech change our research? How much do we really understand and how much of our thoughts around these processes are hubris?

Dark Emu - Book

JJ Hasting’s Homepage

The Island - Movie

Change Agent by Daniel Suarez - Book

Today we’re speaking with Chris Hausler, Senior Data Science Manager at Zendesk, a global customer support SaaS company.

Chris believes that the future of AI and humanity will be a bright one, but humans must adapt and seek new opportunities as old opportunities evaporate. Hausler is working to help clients resolve customer service issues before they crop up by giving users access to the information they need at their fingertips.

Hausler speaks about artificial intelligence growth, development, and trends over the last few years and how Zendesk has helped set the bar for their development on today’s episode of AI Australia

• 02:00 How did Zendesk integrate AI into their already strong ticketing system? How is it changing customer-facing features?
• 02:40 How are Chris and his teams able to find repetitive trends in support conversations,  streamline the information, and add UX elements for self-service.
• 05:40 What was the first machine learning project Chris worked on with Zendesk?
• 09:20 Why does Zendesk keep all data science in Melbourne as opposed to near their home office in San Francisco?
• 14:30 Do you find that talent is more readily available in Melbourne? Is it less competitive than San Francisco? What challenges do you face while recruiting in Melbourne?
• 20:10 What skills and traits do you look for when hiring smart people straight out of university?
• 22:50 When approaching problems, what critical thinking skills come into play? How can these be applied to a machine learning platform?
• 28:00 What technologies have been responsible for machine learning growth over the last four years?
• 33:10 Will humans take action to “hack” self-driving cars? What sorts of action might they take?
• 38:10 Where do you stand on dystopian versus optimistic points of view?
• 40:10 What are some upcoming breakthroughs that are coming up on seriously changing the world?
• 47:00 Can machine learning be used to teach products to change their approach based on the customer experience?
• 48:00 What other outcomes, aside from job loss are possible through machine learning applications in healthcare?
• 49:40 What roles on a team are needed to start hiring for a machine learning business path?
• 50:10 What skills are sought after for early hires in data scientist?
• 50:50 How important is domain knowledge in data science?
• 55:00 Are encrypted backdoors a serious threat to information security? Should legislators be allowed to make these mandatory?

Episode Resources:

https://www.zendesk.com

Today, we’re chatting with Dr Niels Wouters, Head of Research and Emerging Practice for Science Gallery Melbourne and Research Fellow in the Interaction Design Lab at the University of Melbourne.

Dr Wouters’ research focuses on social good and the human element of technology, especially as it pertains to AI and Human-Computer Interaction. A renowned speaker across national and international media, Wouters regularly speaks about the impacts of new technology on urban life. He has been featured on The Sydney Morning Herald, ABC, BBC, The Washington Times, World Economic Forum, Dazed Digital and CNN.

Dr Wouters is the creator of Biometric Mirror, Stories of Exile, Encounters, and Street Talk. His work will be featured in a permanent Science Gallery exhibit in Melbourne.

During today’s discussions, Dr Wouters will explain the purpose behind such fascinating projects as Biometric Mirror and the implications of trusting an AI trained on subjective data. He will also lead us through the journey of Street Talk and the very human, life-changing connections made via technology placed on the outside of homes in Belgium. How can technology bring us together as a community? How might unreliable AI be used against us in the future?

Today, we will explore these questions and more with one of society’s most creative researchers on Human-Machine Interaction, Dr Niels Wouters.

• Street Talk - What happens when you equip family homes with technologies that engage the outside world? Explore the results of note printers for passersby, LED ambient noise detectors, and even a headphone that lets others hear the conversations within the home.
• How do we treat data privacy differently when looking at digital data vs. linking people in analogue ways?
• How did Dr Niels Wouters go from computer science to finance to architecture to find a home here linking science, art, ethics, and the human element?
• How is Science Gallery bringing together artists and scientists to stir interest in STEM among young adults? What’s happening with Science Gallery in Melbourne?
• Biometric Mirror - How do people react to an AI that determines their gender, age, ethnicity, weirdness, aggression, and emotional instability? What if that data were used against them for job selection or insurance rates?
• How simple is it to build an AI from an unreliable dataset? And how open are people to believe an assessment is correct due to the simple fact it was made by a machine?
• What is AI’s impact on human rights? What concerns does a university’s ethics review committee have surrounding a project like Biometric Mirror?
• What are the ethics behind showing AI-generated attractiveness assessments to young adults? Are young people more or less likely to accept a machine-generated assessment as fact?
• Which sci-fi future are we in the midst of creating? The Jetsons? Bladerunner? Black Mirror? Or something else entirely?
• Eliza’s “granddaughter” shows up as a special guest.
• What’s coming in the future? Will there be a Biometric Mirror 3.0? Beyond facial recognition, what would happen if Wouters hacked an Alexa to do unexpected things? How will Biometric Mirror be used as an ethics probe? What’s next for Science Gallery?
• What two or three things could listeners and legislators do now to jumpstart AI ethics in Australia?

Joining us today is Zane Lackey, Co-Founder and Chief Security Officer at Signal Sciences based in New York.

Zane serves on multiple public and private advisory boards and is an investor in emerging cybersecurity companies. He is incredibly well versed in the various trends and advancements in cybersecurity and defending against attacks.

He is the author of Building a Modern Security Program, a how-to in building and scaling effective security teams.

Prior to co-founding Signal Sciences, Zane led a security team at the forefront of the DevOps/Cloud shift as Chief Security Officer of Etsy.

We also have a guest co-host today is Craig Templeton. Craig is the Chief of Information Security Officer REA group and has spent 20 years working in cybersecurity. Safe to say, this episode is stacked with insight into the state of the security industry and what it means for both citizens and businesses.

Here’s what’s discussed in today’s episode:

• Why it’s been a rough year in the tech industry
• How the concept of defence and depth has been turned into expense and depth.
• Why security teams are drowning in too much data and need to focus on what is important.  
• The threat to cybersecurity as a result of automation
• Privacy risks - how do we eliminate or reduce personal information data?
• The implications of data corruption
• Why people may lose trust in machines and avoid using them
• When the most simple methods in designing defence systems can often be overlooked
• Should citizens be able to opt out of automated decisions and would they even know?
• Why legislation needs to catch up with technology
• How can we better adopt and embrace technology such as DevOps, Cloud, AI, and machine learning?

Today, we’re chatting with AI expert, activist, and author Toby Walsh. Toby is a professor of artificial intelligence at the University of New South Wales and Data61 with an educational background from the University of Cambridge and University of Edinburgh in mathematics, theoretical physics, and artificial intelligence. He has also been Editor-in-Chief of the Journal of Artificial Intelligence Research and has chaired multiple AI conferences.

As an activist, Toby helped in the release of an open letter which gained over 20,000 signatures that called for a ban on offensive autonomous weapons. Toby is also the author of multiple books on artificial intelligence, the most recent of which is “2062: The World That AI Made” which we will discuss today.

Dr Walsh opens his thoughts on the future of AI, when and how it could surpass humans, and how little we know about how it could impact our jobs. Toby also discusses with us other ways AI can impact society for better or for worse. How can our data privacy impact AI and how can microtargeting using this data change the course of history? What ethics should we stand by? Who is responsible for decisions made by autonomous machines?  And could government regulation actually help—rather than hinder—innovation?

• Insight into Toby's most recent book, “2062: The World That AI Made”, touted as the book to read, bar none, on AI and society.
• Will AI have consciousness? Is consciousness a biological construct?
• AI's impact on jobs. Will more jobs be displaced than created? Could AI lead to the second Renaissance when people focus on what's truly important?
• Which fictional future will AI lead us most closely to? Blade Runner, A Space Odyssey, Altered Carbon, The Jetsons, etc.?
• How did Toby become involved in AI? How did he become an activist?
• Which legal and ethical issues do we need to look at surrounding AI? AI is being used to target suspicious people for criminal surveillance. Bosses are now sometimes algorithms such as is the case with Uber.
• Should we regulate data monopolies? What is happening to our data privacy and how is it used to manipulate us? Cambridge Analytica and Facebook accused of involvement in manipulating elections through microtargeting using collected personal data.
• What is the regulatory landscape in Australia? How has it been impacted by GDPR? How has data been used illegally to discriminate based on race, gender, and other similar factors?
• What is the potential impact of government regulation on automation in travel? How could the safety of transportation increase through regulated data sharing in automated cars and planes?
• What are the advantages of machines over humans? How will global learning change humanity?
• Do we want humans to be manipulated to such an extreme degree? Can humans be hacked?  Should we regulate weapons of mass persuasion?
• How did AlphaGo create a ”Sputnik Moment” for AI? What impacts has it had? Is robotic soccer Australia's “Sputnik Moment” in AI?

For full show notes and resources head to: eliiza.com/podcast/episode-3