I had the pleasure of sitting down with Khaoula Chehbouni, PhD researcher specializing in bias and fairness in large language models.
In this episode she shares her thoughts about:
- the superficiality of prominent AI safety research
- the hidden incentives in AI development
- how responsible AI is being quietly discouraged
For more information about Khaoula's research, you can check out website!
One of our most valuable assets is too often ignored in end of life planning, namely our data.
Join me as I talk with two of the researchers behind the Digital Legacy Clinic, Jed Brubaker and Dylan Thomas-Doyle from the University of Colorado Boulder.
In this episode we discuss:
- Types of post-mortem data
- How reasonable it is to feel sentimental about a Facebook page
- The creation of new cultural rituals for data after death
For more information about the Digital Legacy Clinic you can check out this link.
In this episode, Ally speaks with Dane Malenfant, a Metis citizen pursuing a Masters in Computer science at Mila and McGill University. They discuss Dane's work at the intersection of AI and psychology and what it looks like to embed indigenous values into AI systems.
Other points of discussion include:
👆 The culture shock of Western values at McGill University
✌ Why a 1992 algorithm works better than modern AI models at integrating indigenous value sets
🖖 How principles from psychology and neuroscience are influencing AI architecture and how AI architecture influences our understanding of psychology and neuroscience
Links:
- Publication: Causality Mining in Fiction
- Publication: Contrastive Retrospection
- Linc Lab
Australian historian Patrick Wolfe so aptly said "settler colonizers come to stay: invasion is a structure not an event". With the Indigenous Pathfinders Program, Lynnsey Chartrand is working to address the colonial legacy obstructing Indigenous representation in the field of AI. In this episode, Lynnsey sits down with Ally to discuss how she and her colleagues have built a bespoke program to enhance Indigenous representation in the AI pipeline. Lynnsey shares key take-aways from the first cohort and her vision for Indigenous leadership in the AI field.
Indspire: https://indspire.ca/about/
Indigenous Pathfinders in AI Program: https://mila.quebec/en/ai4humanity/learning/indigenous-pathfinders-in-ai
FLAIR: https://mila.quebec/en/ai4humanity/applied-projects/first-languages-ai-reality
OCAP Principles: https://fnigc.ca/ocap-training/
This episode is dedicated to a special person, colleague and friend. Abhishek Gupta was well known in the AI ethics community for his intelligence, kindness and humour. With this episode, we hope to carry his torch of creating accessible content on ethical AI and encouraging the public to participate in shaping the future of technology.
In this episode, Ally sits down with Dr. Rumman Chowdhury, CEO of Humane Intelligence and pioneer of AI ethics with leadership roles in Responsible AI at Accenture and Twitter.
They discuss the re-imagining of human intelligence, the fallacy of the perfectly responsible AI product, and the organizational culture that's needed to support responsible AI product development.
Links:
- Dr. Chowdhury's LinkedIn page: https://www.linkedin.com/in/rumman/
- Humane Intelligence: https://www.humane-intelligence.org/
- "Where Responsible AI Meets Reality: Practitioner Perspectives on Enablers for Shifting Organizational Practices" https://arxiv.org/abs/2006.12358
In this episode of "The World We Are Building", Ally sits down with Adriana Eufrosina-Bora. Adriana is an AI and Modern Slavery researcher, pursuing both her PhD on the topic at Queensland University of Technology and leading an applied research project at Mila - Quebec AI Institute.
In this episode we discuss the passion, perseverance and creativity with which Adriana created a multi-year project using AI to tackle the challenge of modern slavery by going after corporate supply chains.
For more information about the project you can check out the AIMS Github: https://github.com/the-future-society/Project-AIMS-AI-against-Modern-Slavery
In this episode of "The World We Are Building", Ally sits down with Kulani Abendroth-Dias. In addition to being a new mom (you will hear from her newborn in the episode), Kulani is an AI risk and policy specialist with a background in behavioural science.
This episode discusses emerging questions in the field of AI policy including data traps, interoperability and governance gaps.
Kulani shares resources including:
- Blunt Force Regulation by Denise Van Der Kamp
In this episode of "The World We're Building," Ally sits down with PhD student, Kyle Roth. They talk about Mormonism's embrace of AI, particularly tools that make revelation and digital resurrection possible.
These concepts are described with reference to the Wilford Woodruff AI Experience, a large language model that has been trained to output answers to users' questions in a style similar to Fourth President of the Church of Jesus Christ of Latter-day Saints, Wilford Woodruff.
For more information on the topic, see Machines for Making Gods: Mormonism, Transhumanism and Worlds Without End" by Jon Bialecki.
In this episode of "The World We're Building," Ally sits down with Dr. Alex Hanna, the Director of Research at the Distributed AI Research Institute (DAIR). Dr. Hanna is a computational sociologist by training, working to tackle some of the most pressing socio-technical challenges with a focus on the data. Dr. Hanna shares quality data practices and both the necessary and sufficient criteria for achieving "good AI". Dr. Hanna also debunks the myth of AI hype and describes the harm it can cause.
For further information about the DAIR team that Dr. Hanna references, you can see their website here: https://www.dair-institute.org/team/
In this episode of "The World We're Building," Ally sits down with Manon Gruaz, a trailblazing UX/UI designer and advocate for ethical, user-centered AI design. Manon's journey from France to Canada, overcoming burnout and depression, has fueled her passion for integrating empathy and human connection into technology.
Join us as Manon shares insights on the importance of understanding user contexts, the intersection of mental health and design, and innovative approach to creating calming, compassionate user experiences. Through their story, we explore how personal challenges can inspire impactful design and why putting mental health first is crucial for reclaiming our well-being.
Tune in to learn about the concept of "healing by design," the responsibilities of designers in the AI era, and how empathy-driven design can transform technology for the better. Whether you're a designer, a tech enthusiast, or someone interested in the human side of AI, this episode offers valuable perspectives and thought-provoking discussions.
Links:
Manon on LinkedIn
Manon's website
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.
Ally brings us in on a story where we explore the captivating story of the Turadic Ocean, the origin of all life on the planet. Journey with us through various landscapes where water not only flows but transforms and adapts to its environment, inspiring innovation across different industries.
From the crystalline streams of the northeastern highlands, purified by ancient trees and moss-covered riverbeds, to the nutrient-rich rivers of the vast southern plains that sustain both people and ecosystems. We also venture into the western desert, where a small stream creates a vital oasis filled with rare minerals essential for the survival of the draconids.
But the story doesn't end there. Learn how this natural inspiration is leading to groundbreaking innovations in today's world. In Silicon Valley, entrepreneurs are mirroring the Turadic Ocean’s adaptability, customizing foundational models to meet the specific needs of their industries. In healthcare, agriculture, and finance, data scientists and experts are refining models to enhance everything from disease diagnosis to crop yields and financial advice.
Did we pique your interest? Listen to the full episode with Carolyne Pelletier and discover how foundational AI-models can be transformed and specialized to tackle unique challenges across various sectors, paving the way for a sustainable and innovative future.
Listen in to find out how this might just be the world we're building
In this insightful episode of "The World We Are Building," we dive deeper into the world of AI with Carolyne Pelletier, a principal machine learning engineer at Reliant and a trailblazer in the field of artificial intelligence. Carolyne shares her unique journey from studying anatomy and cell biology to becoming a key figure in AI development. She discusses her transition into tech through self-taught coding, further education at MILA, and her impactful roles at various startups.
Carolyne emphasizes the importance of multidisciplinary collaboration, rigorous science, and the value of iterating quickly. She provides a nuanced perspective on the challenges and evolution of natural language processing (NLP), highlighting the persistent issue of gender bias in AI and the necessity for responsible AI practices. Carolyne also discusses the scaling of language models, the associated ethical considerations, and the future of AI development.
Join us as we explore the incentives guiding the startup ecosystem, the advancements in language models, and the crucial role of transparency and expert collaboration in building a better AI future. Whether you're an AI enthusiast, a tech professional, or simply curious about the world of AI, this episode offers valuable insights and thought-provoking discussions.
Tune in to gain a broader understanding of AI beyond the hype, through the lens of a leader striving for a more inclusive and responsible technological future.
Links:
The Biasly Project at Mila
The World We Are Building
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.
This is an intro to our episode with Kasia Chmielinski
In this episode, we travel to the quaint town of Carberry, Manitoba, where we uncover the story of Alera, a young woman whose distinct characteristics have caught the eye of leading AI companies worldwide.
Amidst a future where personal data sovereignty is paramount, Alera leverages her unique traits to negotiate groundbreaking deals with tech giants. From SynthCorp's humanoid robots to Cognitech's virtual assistants and NeuroVision's virtual realities, her data promises to bring an unparalleled authenticity to AI technologies.
Tune in as we explore Alera's journey through the lens of the Iterative Trust License, a pioneering concept that ensures ongoing compensation, transparency, and insight between individuals and the tech world. This narrative not only challenges our understanding of data privacy but also provokes a deeper conversation about the ethical deployment of AI.
Join us in conversation with Kasia Chmielinski to discover if this is indeed the world we are building.
In this episode Ally and Peter is joined by the multi-faceted Kasia Chmielinski, a trisectoral professional with significant roles in industry, academia, non-profits, and government. Kasia’s career includes impactful positions at notable organizations such as Google, United Nations, McKinsey, and the United States Digital Service, alongside their co-founding role at the Data Nutrition Project.
Kasia shares insights on developing AI tools during Trump's presidency, emphasizing the importance of transparency in data sets used to train AI models. We dive deeper into the Data Nutrition Project, an initiative they co-founded to provide a "nutrition label" for data sets. A project that aims to improve the transparency of data used in AI systems by providing clear, accessible insights into the contents and quality of data, similar to nutritional information on food packaging. This approach helps developers and stakeholders make informed decisions about the data they use, promoting more ethical and unbiased AI applications.
Kasia also discusses the critical intersectionality in AI development, drawing from personal experiences growing up in a multilingual family. That illustrates how these diverse perspectives have shaped Kasia's views on technology’s role in society and stresses the necessity of incorporating these views to foster ethical AI development.
Join us as we dive into a conversation that challenges dominant narratives about AI, exploring how ethical considerations and transparency are pivotal in the evolution of AI technologies. This episode is a must-listen for anyone interested in the future of AI governance and its societal impact.
Tune in to gain a broader understanding of AI beyond the hype, through the lens of a leader striving for a more inclusive and responsible technological future.
Links:
Kasia on LinkedIn
Berkman Kleine Center - Kasia
The World We Are Building
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.
This is an intro to our episode with Krystal Kauffman
Step into a world where data is the new gold, and those who refine it hold the keys to the future. This episode of "The World We're Building" tells the dramatic sci-fi story of the Orange Run, an audacious act of civil disobedience by data labelers in 2024. These unsung heroes, once sidelined, banded together through the subreddit RejectU to challenge mass rejections and expose the biases riddling their work.
Through meticulous coordination and solidarity symbolized by their copper necklaces, these artisans not only showcased the critical flaws in AI datasets but also ignited a transformative movement in the AI industry. Discover how their strategic defiance brought about a profound shift in recognition, compensation, and career opportunities, turning data labelers from invisible cogs into respected collaborators in the AI development lifecycle.
In this episode, we explore the often-overlooked human element behind artificial intelligence, focusing on the rights and experiences of gig workers who contribute to AI development through data annotation platforms like Amazon Mechanical Turk. Ally is joined by Krystal Kauffman, whose unique journey from political campaigning to advocating for gig workers provides a critical perspective on the intersection of technology and labor rights.
Krystal has a rich background in politics and community organizing, which she has leveraged in her roles at Turkopticon and DAIR to advocate for the rights and fair treatment of gig workers. Her experience as a gig worker herself, coupled with her academic pursuits in geology, provides her with a unique insight into the challenges faced by workers in the digital economy.
Learn about the significant yet often invisible role that gig workers play in the AI industry, the challenges they face, and the potential for systemic change. Krystal discusses the ethical considerations of AI development, the impact of big tech's power dynamics, and her efforts to ensure fair treatment and equitable pay for gig workers globally.
This episode is a must-listen for anyone interested in understanding the critical intersections of technology, labor, and ethics. Tune in to gain a deeper appreciation for the contributions of gig workers and the pressing need for advocacy and reform in the tech industry.
Links:
The World We Are Building
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.
This is an intro to the episode with Sacha Alanoca
This story introduces "Citizen Collect," a virtual reality platform used in Japan to engage citizens directly in policy-making through immersive VR games that simulate complex ethical scenarios.
This innovative approach has skyrocketed political engagement rates, transforming traditional democratic participation and leveraging AI to synthesize public opinion into actionable policy measures, specifically around the use of facial recognition technology and AI in consumer pricing and mental health accommodations in educational settings.
This episode explores how such technologies might redefine civic engagement and democracy, reflecting on the balance between technological potential and ethical governance.
Tune into the full episode to delve deeper into these themes with insights from AI governance expert, Sacha Alanoca.
Welcome to the second episode of The World We Are Building, the podcast which aims to bring hope, agency, and critical thinking into the AI hype cycle.
Today’s guest, Sacha Alanoca, has extensive experience in AI governance, currently studying at Harvard and holding significant positions related to AI policy and technology.
Sacha underscores the importance of ethical considerations in AI development, calling for policies that not only address the immediate impacts of AI technologies but also their long-term effects on society. Her work aims to balance innovation with responsibility, promoting an AI future that is inclusive and fair for all.
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.
Welcome to the first episode of "The World We're Building," podcast! Forget the headlines about AI taking over the world, this podcast is designed to shed light on the people, ideas and stories that rarely make the front page. We´ll be speaking with people who are working with passion and intention to make the world we're building a better place for all.
Our guests will include data labelers, activists, designers, and policymakers (among others) and our interviews will be bolstered with sci-fi narratives to promote collective engagement in and criticism of the world of tomorrow.
If you want to get in touch with us, feel free to reach out to Allison Cohen and/or Peter Kurzwelly on LinkedIn, or both of us on Twitter/X.