“CODEVID-19 has demonstrated the powerful role community plays in innovation,” says Warren Johnston, Managing Director of Community Engagement at Amii. “Over the past six weeks, we’ve seen some fantastic collaborations that have sparked some big ideas and interesting solutions to the many challenges facing the world as a result of the global pandemic.”
Hackathons provide a ground floor from which multidisciplinary teams can imagine, design and innovate. Similarly, CODEVID-19 has become a catalyst for critical tools and resources that will be used the world over in the weeks and months to come.
First launched on March 16, 2020, CODEVID-19 began as a conversation among members of the Dev Edmonton Society in Edmonton, Canada. The six-week-long event grew quickly, reaching more than 1,000 collaborators in less than seven days.
In its final week, the CODEVID-19 hackathon boasted more than 1,800 contributors from 56 countries across five continents. Over 50 teams offered solutions that met challenges in supporting crisis response, understanding the pandemic, social distancing & isolation, and scarcity & the economy.
First was Charity Shop Exchange, a UK-based platform that applies the increasingly popular subscription box model to charity shops (of which there are more than 10,000 in the UK). Here in Canada, we usually call those thrift stores (places like Goodwill or Value Village). On Charity Shop Exchange, you input the things you like to watch and read, then the site will buy those from the store and deliver them to you. It’s a way to support isolation.
Next was Trusted Locals, a platform that helps locals share information about their current on-site situation “in an organised and audited way.” Users can submit posts such as where they saw toilet paper available. Then other users can confirm or disconfirm it, which informs a confidence score for the post. The idea is to help others around you with more trustworthy information. The developers are also looking to scrape social media sites like Twitter for information.
Next up was Where Have I Been which lets users record all of the places they’ve visited on a day-to-day basis. Check-in apps aren’t new, but what this one lets you do in addition is see how risky the places you visited were. If a user self-reports that they have COVID-19 symptoms, the app will notify other users who visited the same places in the last two weeks. You can also view high risk locations in your area on a map.
“Amii is proud to stand alongside our tech and innovation communities,” says Johnston. “The organizers have done a fantastic job, working with the Dev Edmonton Society, rallying a diverse community of designers, developers and subject matter experts for CODEVID-19. We’re thrilled at the progress teams have made and are excited to see the innovative solutions that have been developed over the past six weeks.”
One of Canada’s three centres of AI excellence as part of the Pan-Canadian AI Strategy, Amii (the Alberta Machine Intelligence Institute) is an Alberta-based non-profit institute that supports world-leading research in artificial intelligence and machine learning and translates scientific advancement into industry adoption. Amii grows AI capabilities through advancing leading-edge research, delivering exceptional educational offerings and providing business advice – all with the goal of building in-house AI capabilities. For more information, visit amii.ca.
Amii has joined forces with CIFAR, Mila and the Vector Institute to offer a coordinated national approach to the fight against COVID-19.
CIFAR, a Canadian-based global charitable organization that convenes extraordinary minds to address the most important questions facing science and humanity, has activated the anchor institutes of the Pan-Canadian AI Strategy through a number of initiatives, task forces, and funding opportunities:
Cam Linke, Amii CEO, is participating in ongoing International Roundtables on AI and COVID-19 with a focus on data access and sharing. Reports from the discussions can be found at cifar.ca/action-on-covid19.
Amii Fellows are actively applying for targeted CIFAR funding for AI and COVID-19 interdisciplinary research collaborations to support innovating, high-risk/high-reward ideas and projects.
AI Against COVID-19
In addition, AI Against COVID-19 Canada is a special task force aiming to map and coordinate AI projects in Canada that can contribute to solve the COVID-19 outbreak and limit its impact on society.
The task force is led by a community of researchers from CIFAR, Mila, the Vector Institute, and Amii. This site is an information-sharing platform and its goal is to stimulate new collaborations and data sharing among the institutes and external collaborators.
As the world continues to adjust to the new normal of the realities of COVID-19, Amii is offering our expertise as part of the Roche Data Science Coalition.
Hoffmann-La Roche (Roche) is committed to working with the global community to develop solutions to the challenges of the COVID-19 pandemic. We have assembled a group of like-minded public and private organizations with a common mission and vision to bring actionable intelligence to patients, frontline healthcare providers, institutions, supply chains, and government. We call ourselves the Roche Data Science Coalition.
The collaboration launches with the UNCOVER challenge (United Network for COVID Data Exploration and Research) on Kaggle.com. The challenge presents a curated collection of datasets from 20 global sources and asks participants to model solutions to key questions that were developed and evaluated by a global frontline of healthcare providers, hospitals, suppliers, and policymakers.
“The Roche Data Science Coalition has come together at a crucial time for the world,” says Amii CEO Cam Linke. “Cross-sector collaboration is key for developing globally-relevant solutions, and Amii is eager to join with other top minds to solidify Alberta’s role in meeting the unique challenges facing the world today.”
Amii scientists will represent Alberta by lending our AI and machine learning expertise to the coalition’s working group alongside public and private organizations like NVIDIA, The Vector Institute and ThinkData Works, among others. Our teams will also open up office hours and idea-generating whiteboard sessions to working group members looking to better understand how to incorporate AI into their solutions.
The coalition is focused on support in three key areas: patient support, insights into capacity issues, and research enhancement. The Kaggle challenge is made up of 12 tasks developed and evaluated by global frontline healthcare providers, hospitals, supply chain and public bodies:
Which populations are at risk for contracting COVID-19?
Which populations have contracted COVID-19 and require ventilators?
Which populations have contracted COVID-19 who require the ICU?
Which patient populations pass away from COVID-19?
Which populations assessed should stay home and which should see an HCP?
Which populations of clinicians and patients require protective equipment?
Which populations of clinicians are most likely to contract COVID-19?
How are patterns of care changing for current patients (i.e. cancer patients)?
What is the change in turnaround time for routine lab values for oncology patients?
What is the incidence of infection with coronavirus among cancer patients?
Are hospital resources being diverted from providing oncology care to support the COVID-19 response?
How is the implementation of existing strategies affecting the rates of COVID-19 infection?
The Roche Data Science Coalition is requesting the collaborative effort of the AI community around the world to contribute to the fight against COVID-19 by sharing their modelled solutions using the presented data and if needed, their own proprietary and non-proprietary datasets.
Amii has joined the fight against COVID-19 by lending our support and expertise to CODEVID-19.
This globally-focused hackathon encourages the rapid creation of tools and solutions specifically for solving new problems arising as the world adapts to the COVID-19 pandemic. Produced by the Dev Edmonton Society, the hackathon launched in Edmonton early last week and has grown to include participants from more than ten countries.
The objective is simple: to improve the quality of life of people worldwide during the pandemic. The Dev Edmonton Society believes that fostering rapid development of pandemic solutions is crucial, and by bringing developers, researchers, and domain experts together they can reach their goal of creating useful and sustainable applications that can help the world.
Hackathons provide a ground floor from which teams can imagine, design and innovate. CODEVID-19 seeks to be a catalyst for creating some of the key resources that will be used the world over in the weeks to come. The CODEVID Slack workspace has more than 1,200 active members from across the globe collaborating together.
Amii is proud to offer our expertise in support of this amazing initiative. Our research teams will provide insights to participants to help them better understand how to leverage AI and machine learning as part of their work. We’ll deliver a series of workshops – led by Amii Scientists and Fellows – to explore the use of AI in medical applications and to highlight some of the key considerations teams need to make as they build AI into their applications.
The purpose of the hackathon ultimately isn’t to win prizes, but to build useful apps that people can use to manage and survive during the COVID-19 pandemic. What started in Edmonton, Canada as a discussion amongst developers on the Dev Edmonton Society Slack has exploded into a global initiative to improve people’s quality of life and ease the impacts of the pandemic.
Read more about Amii’s commitments to the global fight against COVID-19.One of Canada’s three centres of AI excellence as part of the Pan-Canadian AI Strategy, Amii (the Alberta Machine Intelligence Institute) is an Alberta-based non-profit institute that supports world-leading research in artificial intelligence and machine learning and translates scientific advancement into industry adoption. Amii grows AI capabilities through advancing leading-edge research, delivering exceptional educational offerings and providing business advice – all with the goal of building in-house AI capabilities.
Building on the momentum of our recent successful launch of AlbertaWomen.ai, we are pleased to take another step forward in encouraging gender diversity in artificial intelligence (AI) with the announcement of a two-week pilot of the AI4Good Summer Lab, an initiative originally launched in Montréal in 2017 by Angelique Mannella and Professor Doina Precup (Canada CIFAR Chair in AI at Mila).
Diversity and Inclusion in AI
Algorithmic bias is a key challenge in artificial intelligence (AI), and we know that investments in diversifying perspectives will enable us to develop more reliable models and more accurate algorithms for our learning systems. With women representing only 27% of researchers in computing science, we still have work to do to incorporate the voices and perspectives that will enable us to build robust algorithms that can provide benefits across all aspects of society.
That’s why we’re so excited to be part of this initiative.
AI4Good Summer Lab
The AI4Good Summer Lab is the first of its kind, focused not only on innovative approaches to teaching and learning AI, but also with the ambitious goals of tackling diversity and inclusion in AI research and development, and inspiring the next generation of technical leaders to develop AI as a force for social good.
The Edmonton pilot, hosted at Amii, takes place over two weeks (July 6 to 17) and brings together a cohort of women from some of Canada’s most engaged AI communities. The program consists of a mashup of lectures, workshops, self-directed team projects. Instructors and mentors include university researchers, graduate students, and practitioners from startups, corporations, inter-governmental organizations and city government. There’s also a strong emphasis on curiosity-driven learning and mentorship both from peers and industry experts.
The program is open to individuals who identify as women who are registered with a post-secondary institution at the undergraduate level (or have graduated within the last 2 years) in STEM fields and who have prior experience with coding and advanced math (calculus, linear algebra, stats).
Apply today and be part of this exciting cohort! Applications close February 29, 2020.
(Please note: This program is only open to those who identify as a woman.)
Edmonton Startup Week 2019 is imminent, and we couldn’t be more excited to participate! From October 21 to 25, attend workshops, socials, and events geared towards building momentum and opportunity around Edmonton’s unique entrepreneurial identity, startup community and culture of innovation.
Here is where you can catch the Amii team:
Tech on Tap: Machine Learning Mishaps with ATB & Amii
October 21, 4 – 5 p.m. | Mercer Tavern (10363 104 St NW)
Start your Startup Week in the best way possible by joining ATB and Amii for a special happy hour event at Mercer Tavern. Join us for some great conversations, some awesome craft beer, and a light-hearted presentation from Cathy King (Director, Amii Educates) and David Chan (Director, Amii Innovates). RSVP Here!
Artificial Intelligence in Edmonton
October 22, 12 – 1 p.m. | CBC Edmonton Stage (123 City Centre Mall, 10062 102 Ave)
Sit down with Cory Janssen (CEO of AltaML), Craig Knox (CTO of DrugBank) and Patrick Pilarski (Amii Fellow and Associate Professor at UAlberta) to learn about the real life ways that any industry can integrate AI into their company. Whether it be to create internal efficiencies, optimize a business process, or advance and scale their operations, these contributors to AI in Edmonton will chat about unsexy AI and the integral role it plays in innovation. RSVP Here!
Machine Learning 101
October 24, 12 – 1 p.m. | Startup Edmonton (301 – 10359 104 St NW)
Heralded by many as the fourth industrial revolution, artificial intelligence has inspired countless news articles, novels, and films. With this deluge of information comes hopes and aspirations, fears and misconceptions – some justified and others not. How can we make sense of it all? Join Amii as we pull back the curtain on AI and lay a pragmatic foundation for what AI is, what it isn’t, and how you can begin to think about the place of AI in your business. RSVP Here!
In Search of Edmonton’s Next Tech Unicorn
October 24, 1 – 5 p.m. | World Trade Centre (#200 – 9990 Jasper Avenue)
Are you interested in investing in tech but don’t know where to start? Don’t know the difference between pre-seed, seed or series A, B or C investment rounds? Would you like to hear from veteran tech investors about how to pick winners or winning teams? Interested in meeting some of Edmonton’s leading tech company founders that are at the forefront of Canadian innovation success? Join us for an engaging and interactive afternoon discussing how to invest directly in some of Edmonton’s best technology companies, featuring Amii Fellow and Professor at UAlberta, Rich Sutton! RSVP Here!
October 25, 12 – 1 p.m. | CSC 3-33, University of Alberta (8900 114 St NW)
These weekly seminars (free pizza lunch provided) give AI enthusiasts a friendly way of engaging with the latest topics and trends in AI research and development.RSVP Here!
Ladies Learning Code: YEGTech Open House (For All Ages)
October 26, 10 a.m. – 4 p.m. | Startup Edmonton (301 – 10359 104 St NW)
This Halloween-themed event gathers together Edmonton tech community groups, meetups, non-profits and businesses – including Amii, represented by Anna Koop (Director, Amii Explores). Try out tons of rad activities while getting to know the people behind Edmonton’s thriving tech community! Look for opportunities to collaborate and possibly recruit new members or employees – plus indulge in some costumes and candy. RSVP Here!
October 24, 6:30 – 10:30 p.m. | Startup Edmonton (301 – 10359 104 St NW)
Many members of the Amii team will be milling about the tenth edition of Launch Party Edmonton, the city’s flagship startup event that celebrates and showcases the hottest startups in town. Launch Party isn’t your typical networking event or trade show. It’s a party designed to celebrate and showcase the rockstar entrepreneurs in our community. Drinks, DJs, and great company all await you at Launch Party!Purchase Tickets Here!
Wrapping up an exciting two weeks of AI excellence in Alberta: 2019 DLRL Summer School
The 2019 CIFAR Deep Learning & Reinforcement Learning Summer School has concluded. Hosted by Amii and CIFAR from July 24 to August 2, 2019, this amazing and inspiring two weeks brought together the world’s brightest graduate students, post-docs and professionals to cover the foundational research, new developments and real-world applications of deep learning and reinforcement learning.
Highlights of the event include:
Over 1200 applicants from 75 countries
300 attendees from 36 countries
113 women attendees
128 participants from outside of Canada (including 28 from developing nations)
Over 25 speakers
More than 90 hours of educational sessions and social events
40 participating organizations at the Career Fair
Over $300,000 in corporate sponsorship
Prior to DLRL Summer School, the inaugural Summer Institute on AI and Society was held from July 21 to 24. Co-convened by CIFAR, the AI PULSE program at UCLA School of Law, and Amii, Summer Institute brought together 40 interdisciplinary experts for engaging conversations and productive workshops on the impacts and effects of AI in society.
This story was featured in the AICan Bulletin. Subscribe to the bi-monthly email publication to keep up to date on AI in Canada.
Conclusion de deux semaines d’excellence en IA en Alberta : École d’été sur l’apprentissage profond et l’apprentissage par renforcement 2019 du CIFAR
L’École d’été sur l’apprentissage profond et l’apprentissage par renforcement 2019 du CIFAR, organisée par l’Amii et le CIFAR, a attiré 300 des plus brillants étudiants diplômés et postdoctorants de 36 pays qui ont pu découvrir les recherches de pointe menées par les plus grands chercheurs du Canada sur l’apprentissage profond et l’apprentissage par renforcement.
Le tout premier Institut d’été sur l’IA et la société s’est tenu du 21 au 24 juillet, juste avant l’École d’été DLRL. Organisé conjointement par le CIFAR, le programme AI PULSE de l’École de droit de l’Université de la Californie à Los Angeles et l’Amii, l’Institut d’été a réuni 40 experts interdisciplinaires qui ont pris part à des conversations stimulantes et à des ateliers productifs sur les impacts et les effets de l’IA sur la société.
“Ken and I had always worked on all kinds of applications of AI to reasoning in any domain. Legal reasoning was an early target for AI […] because it’s a domain in which there’s an attempt to be precise about how you write things in natural language,” says Goebel. “And Ken was so keen on this, that nine years ago […] he got very frustrated with trying to take the next step in applying AI to legal reasoning, so he took a law degree.”
Satoh completed a law degree at the University of Tokyo, while at the same time maintaining his position as a full-time professor at one of the most prestigious computer science departments in Japan. Once completed, he took the Japanese bar exams and failed – five times.
“COLIEE was born out of his frustration at the way the bar exams were administered,” explains Goebel. “He and I decided we could stage a competition. We could use Japanese bar law exams as examples with statute law in Japan and mount a competition to see how people could use AI to answer bar law exam questions.”
Since the competition was first staged, Satoh has passed the Japanese bar exams. Meanwhile, Goebel has accomplished the same, albeit with AI; he led a team which designed a program that passed the exams in 2017. But COLIEE still happens every year, growing in challenge and ambition. This year, Goebel and Satoh are just two of six coordinators, and a total of 18 teams are competing from 13 different countries, from the US to Botswana; Argentina to Germany.
The Tasks at Hand
Since its inception, the challenges have steadily increased in complexity; now, the competition includes two categories classified by legal concepts in statute law and case law. Those two categories each have two tasks: an information retrieval task and an entailment task. Teams have the option of submitting for one task, all four tasks, or any number in between.
The information retrieval tasks challenge a program to take a given test case and retrieve the related statutes or cases. The entailment tasks take it a step further.
“Entailment is really just saying: if this is a statute and this is a question, does the statute entail the question or not? [Editor’s note: the Dictionary.com definition of entail is “to cause or involve by necessity or as a consequence: e.g. a loss entailing no regret”] It’s the legal reasoning required to say, if you are wearing an expensive kimono and a bystander pushes you out of the way of a car that’s about the hit you […] is he liable for damages to the kimono?” says Goebel. “So we want to build computer programs that […] retrieve the appropriate statutes, then they have to find a connection.”
In other words, an entailment task requires a program to perform information retrieval, then create a yes/no argument for the test case based on the returned statutes or cases. In the example above, a program would be required to retrieve statutes related to property damage and liability, then determine whether or not the bystander would be liable for the damaged kimono.
The structure of the competition has accelerated research in this area in a unique way. COLIEE has built a community of people who are tackling similar challenges, and as a result, its participants have naturally developed a vocabulary that has quickened the process of exchanging information and ideas.
“It’s almost to that point at the COLIEE competition workshops […] people talk without even specifying. They say, ‘and here’s our approach to task one’. They don’t describe task one anymore.”
Explainability – the ability to determine how a model arrived at an answer – plays a large role in the entailment portion of the competition. Goebel’s lab, the Explainable AI (XAI) Lab out of the University of Alberta, is dedicated to this exact concept. Goebel’s lab is competing in COLIEE, and in this aspect, he has an advantage.
“Our lab has been tackling these two areas of medical reasoning and legal reasoning; we want to drive the science of AI forward. So we have an advantage to all of the people on the planet who, in an ad hoc way, apply learning of any kind to these domains, because we’ve been doing them for longer, and our focus is on making them explainable.”
Many machine learning models are challenged by the inability to explain how they arrive at decisions. For example, a model can take points of data and group them together based on a similarity it noticed, but it will not be able to articulate the similarity. Many refer to this issue in metaphor, calling them “black box” models.
Black box machine learning models can perform brilliant tasks – but in application, they can have serious consequences. If a model is assisting a doctor to recommend treatment, a judge to recommend sentencing, or a hiring manager to choose resumes, it is important to be able to determine why it is arriving at decisions.
“Explainability is as simple as saying: when I tell you something, please explain,” says Goebel. “The background that I have comes from formal philosophy and building systems to create hypotheses about data. That’s what scientists do. In there lies all of the mechanisms you need to do explanation.”
Inventure$ has arrived! This “unconference,” held from June 5-7 at the Telus Convention Centre in Calgary, joins together entrepreneurs, innovators, investors, researchers and thought leaders to discover and share the latest in innovation. We couldn’t be more excited to be part of what is quickly becoming Alberta’s premier business and technology event.
If you are also in Calgary this week, here is where you can catch the Amii team:
Oil Sands Innovation Summit Panel
When: June 4, 3:20 – 4:20 p.m. Where: Hyatt Regency (700 Centre St SE) Can Alberta be a global leader in Oil & Gas AI? What is the implication for our businesses? Find out at this panel featuring Geoff Kliza, Director of Amii Innovates. The Oil Sands Innovation Summit takes place June 3 – 4 in partnership with Inventure$. The event is sold-out.
Calgary AI Meetup
When: June 4, 5:30 – 7:30 p.m. Where: Room EN C70, Engineering Complex, University of Calgary (2500 University Dr NW) RSVP Here! Anna Koop from Amii’s science team will be sharing the latest reinforcement learning research happening at Amii, and she’ll be joined by other members of the Amii team to answer your burning questions about machine learning in this latest edition of the Calgary AI Meetup.
Where: June 5, 9 a.m. – 4 p.m. Where: Calgary Telus Convention Centre (120 9 Ave SE) Purchase Tickets Here! Find Amii on the conference floor – we’ll be in our booth answering questions, sharing information and demoing research from our Amii Fellows and Innovation Affiliates:
– Bento Arm: Developed at the BLINC Lab under the supervision of Amii’s Patrick Pilarski, the Bento Arm is a platform for training and research applications of AI-enabled prosthetic arms. The Bento Arm will be on display from 9 a.m. – 4 p.m.
– DeepStack: Developed by a research team led by Amii’s Michael Bowling, DeepStack is the first AI capable of beating professional poker players at heads-up no-limit Texas hold’em poker. DeepStack will be on display from 9 a.m. – 4 p.m.
– SAMDesk: SAMDesk monitors social media to determine when important events are taking place around the world. Visit our booth to see the tool work in real time. SAMDesk will be on site from 9:30 – 11:30 a.m.
– Mikata Health: Mikata Health uses AI to help doctors and staff connect assist patients more effectively through chat technology. Mikata Health will be joining us from 11:40 a.m. – 1:40 p.m.
– Medo.ai: Medo.ai aims to simplify the use of ultrasound for common and critical conditions. Visit the booth to see how they use AI-augmented ultrasound imaging to support diagnosis. Catch Medo.ai at the Amii booth from 1:50 – 3:50 p.m.
ATB Tech on Tap – Machine Learning Mishaps
When: June 5, 5 – 7 p.m. Where: ATB Financial (102 8 Ave SW) RSVP Here! ATB is hosting a very special Inventure$ edition of Tech on Tap happy hour! Drink free beer and listen to Tara and Anna from the Amii science and education teams as they share some memorable Machine Learning fails and mishaps.
Personal assistants. Automation. Smart homes. Hailed by many as the fourth industrial revolution, machine intelligence is bringing remarkable changes everywhere – Canada included. We are fortunate to have places like Edmonton, Montreal and Toronto leading the way as Canada’s major centres for artificial intelligence. Notably, they are also home to some of Canada’s most exciting AI events.
If you want to learn how to harness the technologies that are shaping our future, here are three Canadian AI events you should consider checking out in the coming months:
World Summit AI Americas (americas.worldsummit.ai) April 10 – 11, 2019 | Montreal, QC World Summit AI has made it across the pond! The series with the world’s largest and most active AI community (over 40,000 members from the global AI ecosystem) is hosting the first World Summit AI Americas in Montreal this April. This event promises two full days of mind-boggling innovation, animated discussions on AI4good, applied solutions for enterprise, hands-on workshops and the development of plans for advancing the application of AI in the coming year. To learn more and purchase a ticket, visit the World Summit AI Americas page; use promo “AMII15” for 15% off your ticket.
SingularityU Canada Summit (sucanada.org/summit2019) April 23 – 24, 2019 | Edmonton, AB After a smashing Toronto debut in 2017, SingularityU Canada Summit lands in Edmonton this April. A mixture of keynote discussions, panel presentations, product demos, workshops, and breakout sessions, this conference shares the best of Canadian and international technology. Learn about the transformative impact that technologies like artificial intelligence, nanotechnology, and digital medicine will have on our lives and our world. To learn more and purchase a ticket, visit the Singularity U Canada Summit page; use promo “AmiiSUDIS20” for 20% off your ticket.
Big Data & AI Toronto July 22-24, 2019 | Toronto, ON Big Data and AI Toronto aims to address the greatest business challenges technology leaders are facing today. Industry sessions, keynote discussions and panel presentations explore the future of work in multiple industries. Visit the Big Data & AI Toronto registration page to purchase your ticket.
Whether you are a curious beginner or a seasoned pro, there’s lots of (machine) learning to be had this year!
Bonus Event: The Summer Institute (dlrlsummerschool.ca/the-summer-institute/) July 21-24, 2019 | Edmonton, AB If you are passionate about understanding and shaping the relationship between AI and society, consider applying to attend the Summer Institute on AI and Society. A combination of lectures, panels, and participatory problem-solving, this intimate interdisciplinary event aims to build understanding and action around the most important issues facing AI. Visit the Summer Institute page for more information.
Summer Institute brings together experts, grad students and researchers of all backgrounds to explore the societal, governmental, and ethical implications of AI. A combination of lectures, panels, and participatory problem-solving, this comprehensive and interdisciplinary event aims to build understanding and action around these high-stakes topics.
We spoke with one of the co-organizers of Summer Institute, UCLA School of Law professor Edward Parson, to talk about the origins of the event, what themes and topics might be covered, and why you should apply now. Check out what he had to say below:
Please note: this interview has been edited and condensed for space
Tell us about how you became interested in AI and its societal impact.
My main professional background has been in environment, energy and related policy areas. But because of my partial scientific and technical training, I’ve always had a central interest in technology and those areas – what it does, what forces determine how it changes, how, if at all, societies can get the benefits and limit the harms, and how that works. That was the bridge to thinking about AI.
How did you get involved in the Summer Institute?
Last year when I was on a sabbatical year at the University of Victoria, I became aware of CIFAR’s program supporting AI and related initiatives. And in particular, CIFAR’s interest in broadening its support from technical issues of AI out to societal impacts, regulatory, and governance issues. After speaking with them and then consulting with a couple of Canadian colleagues who are more on the technical side of AI – Alona Fyshe and Dan Lizotte – we submitted a proposal, it was approved, and we’re going forward with the three of us co-directing the institute, with joint support from CIFAR and from my project here at UCLA.
Can you tell us more about this project at UCLA?
It’s an outgrowth of a longer-standing activity at UCLA Law School that’s been on Science Technology in Law, called the AI PULSE Program. We’re looking at ways to think through potential impacts that are sort of intermediate in scale and time horizon. We’re looking for ways to get reasonably disciplined hooks on what the impacts might be five, 10, 20 years out, and how to anticipate, assess, and forestall the most disruptive and harmful aspects of those.
This also characterizes my main interest for the Summer Institute. But I’m one of three co-organizers. My two co-organizers’ interests come mainly from the side of technical aspects of AI. They’re more concerned with developing useful ethical guidelines that students and practitioners of AI and machine learning might observe in their current practice. So we expect to be covering a range of issues.
What do you believe the benefit is of the Summer Institute for attendees?
To be involved in conversations on these fascinating topics that don’t have a lot of place for consideration in the normal curriculum. Networking among a bunch of people with similar interests on issues that are likely to be really important and recurrent over time. And I expect it’ll be really interesting and fun.
What important ethics and societal implications should AI practitioners pay attention to?
AI is the weirdest technology in the world. I’ve spent decades studying social impacts of technology in all kinds of domains. AI is unlike any other technology that I’ve thought about before because nobody knows what it is. It is so diffused, so fuzzy in its boundaries, so diverse in the different strains of capability that contribute to what’s going on presently. And so limitless in the things it might be used for.
What might AI do? It might enable things that are not presently possible. It might enable an extraordinary advance in environmental protection management. It might displace human ingenuity, or augment human ingenuity, in dozens of fields of scientific and technological research. Some weeks ago, a new machine learning program out of DeepMind in London won the annual world competition for protein folding projections. It’s sort of like what happened to the Go masters just happened to the protein scientists.
On the other hand, things that become possible through technological advance often get done even if we disapprove. One of my colleagues who thinks about this stuff, Allen Dafoe at Oxford, has thrown out the slogan that “one of the social risks of AI is robust totalitarianism.” Comprehensive surveillance with perfect facial and human individual recognition and omnipresent information about everything you think, do, and say. In the hands of a tyrannical regime.
AI is big stuff. It is big, historical stuff. The possibility of capabilities that really fundamentally disrupt employment and livelihood and labour markets, that fundamentally disrupt the functioning of the state, that fundamentally disrupt the functioning of the economy, and every sub-sector thereof for good and ill.
The potential benefits are enormous, but even they will come with enormous disruption. So if we all get to move to a Jetsons world where we’re at leisure all day and the machines do the work, that might be really nice. But it will explode a bunch of foundations of social order. These are all the things we need to talk about at Summer Institute.
What aspects of its implications do you think are not being paid enough attention?
It’s the medium term – what happens five steps down the line, and how we can get any handle on thinking about that beforehand. To make an environment that makes it likely that people get the benefits and don’t get the worst harms from those rapid changes.
What are you most looking forward to about Summer Institute?
Talking about all this fabulous stuff with a bunch of really interesting and engaged people from all over the place spatially, and from all over the place in terms of intellectual background and how they think.
The Summer School brings together graduate students, post-doctoral fellows and industry professionals to explore the latest AI techniques and advancements, build their research networks and open new opportunities for collaboration. Participants learn directly from world-renowned AI researchers including Richard Sutton, Yoshua Bengio and Martha White. Other programming includes an AI Career Fair and social events around town.
To give this year’s attendees an idea of what to expect for this year’s Summer School, we asked three of last year’s attendees to tell us about their experiences and what they found valuable. Check out what they had to say below:
1. Why did you decide to apply for the 2018 Deep Learning & Reinforcement Learning Summer School?
Katya: The [Summer School] is an incredible opportunity to meet distinguished researchers in the field face-to-face, get the latest cutting-edge research and, no less important, to connect and meet fellow students, find out what they are working on and discuss potential collaborative projects. This past year we had people from the University of Toronto, Université de Montréal, McGill, University of Alberta, CMU, MIT, Stanford, Google, Duke.
Matthew: When the application went up, I was actually informed by several of my lab mates that the Summer School was something that would be extremely valuable in my future research. There was also quite a big push to apply in the RLAI [editor’s note: Reinforcement Learning & Artificial Intelligence] Lab here at University of Alberta. I joined in and was even more excited once I saw the list of speakers.
Raksha: I was aware of various summer schools, but I’d never gotten a chance to attend one. So in late winter of 2018, when I heard about the [Summer School] happening in Toronto, it seemed like the perfect opportunity. The list of speakers for the school were distinguished researchers in the area, plus the RL school component, the positive reviews from peers who had been to the 2017 version, and encouragement to attend by professors in the department, made it very exciting – and I decided to apply to it!
2. What did you enjoy most about the event?
Katya: People. The brain power and diversity of research backgrounds in the room were fascinating. It was also valuable that the timetable had plenty of time for networking, i.e. over lunch.
Matthew: I was struck by the amazing talent of all the participants and how many were from fields outside machine learning and reinforcement learning. Some of my favorite conversations were with physicists who were looking to apply deep learning or reinforcement learning to model the dynamics of particular physical systems. There was also a lot of opportunity to connect with people from around the world. Learning about the many avenues of research explored is awe-inspiring and a reminder that my own interests are a minuscule part of the picture.
Raksha: As the focus is both on deep learning and reinforcement learning, it was a great opportunity for me, a person who works in reinforcement learning, to get insight into state-of-the-art research in deep learning and hear different perspectives about reinforcement learning. Additionally, it was really nice to meet and interact with the extended peer/research community in a more school-like setting!
3. Did the Summer School affect your career and/or research trajectory? If so, how?
Katya: Terrific impact. During one of the breaks, we sat down with Rich Sutton and discussed a new project. As a result of that discussion, four months later, I am visiting the RLAI group at the University of Alberta and working with Rich on this project, which becomes an intersection of the NLP and RL. This is a great example where a conversation turned into a mind-blowing experience – and not because of the winter in Alberta (I am from Siberia!) – but by being in the “mecca of RL” and having an opportunity to learn from Rich and the group.
Matthew: I would say it emboldened my research trajectory. My interests are still the same — i.e using reinforcement learning to make predictions of the world through interactions — but I am more excited about this topic and how it relates to the wider AI community. The scope of my research has also widened. While before I was very narrow in what I thought was the way forward, I am now looking towards many communities I had not considered (or was even aware of!).
Raksha: [The] chance to discuss my research with some leading researchers, listen to their experiences and thoughts about what’s to come and where we are headed, meet and interact with peers who are pursuing interesting problems in their research, etc., has been very inspiring!
4. What was the most valuable thing you learned or experienced at the event?
Katya: Being able to connect with people and learn from them … [S]uch as having an opportunity to ask Graham Neubig hundreds of questions and get most practical answers, learning about RLAI Lab from the people in that group and connecting with people there, being deeply inspired by Martha White and Jamie Kiros.
Matthew: I found the most value in how my view on the field was expanded. The exposure of ideas and topics that I hadn’t yet seen sparked many ideas that I want to explore in my future research. It gave me perspective on the massive amount of work that is still left, but with that the many interesting topics still unexplored.
Raksha: I have always heard about the general [Deep Learning/Reinforcement Learning] community being large and diverse, but this was my first hands-on experience of it. It was invaluable to meet and interact with peers from various countries! The Summer School was filled with energy from day one — and we got a sneak-peak into the breadth of research in the community, in a classroom-like setting, which was really valuable!
L’École d’été permet à des étudiants diplômés, à des boursiers postdoctoraux et à des professionnels de l’industrie d’explorer les plus récentes technologies et avancées de l’IA, de développer leurs réseaux de recherche et de multiplier les occasions de collaboration. Les participants apprennent de chercheurs en IA de renommée mondiale, dont Richard Sutton, Yoshua Bengio et Martha White. Le programme comprend aussi un salon de l’emploi en IA et des activités sociales en ville.
À quoi doivent s’attendre les participants de l’École d’été 2019 ? Pour leur donner un aperçu, nous avons demandé à trois stagiaires de l’année dernière de nous parler de leur expérience et de ce qu’ils en ont retiré. Voici ce qu’ils avaient à dire.
1. Pourquoi as-tu participé à l’École d’été sur l’apprentissage profond et l’apprentissage par renforcement 2018 ?
Katya : [L’École d’été] est une occasion incroyable de rencontrer en personne d’éminents chercheurs dans le domaine, de connaître les plus récentes recherches de pointe et, tout aussi important, de rencontrer d’autres étudiants, de savoir sur quoi ils travaillent et de discuter de projets de collaboration éventuels. L’année dernière, il y avait des gens de l’Université de Toronto, de l’Université de Montréal, de l’Université McGill, de l’Université de l’Alberta, de l’Université Carnegie Mellon, du MIT, de l’Université de Stanford, de Google et de l’Université Duke.
Matthew : Au début de la période d’inscription, plusieurs de mes partenaires de laboratoire m’ont dit que l’École d’été pouvait s’avérer extrêmement utile pour mes futures recherches. Il y avait aussi beaucoup de pression pour s’inscrire au Laboratoire sur l’apprentissage par renforcement et l’intelligence artificielle (RLAI) de l’Université de l’Alberta. Je me suis inscrit et j’étais encore plus excité lorsque j’ai vu la liste des conférenciers.
Raksha : Je connaissais différentes écoles d’été, mais je n’avais jamais eu la chance d’en fréquenter une. À la fin de l’hiver 2018, quand j’ai entendu dire que l’École d’été aurait lieu à Toronto, j’ai pris cela comme un signe du ciel. Les conférenciers au programme étaient tous des chercheurs renommés dans le domaine. De plus, la composante en apprentissage par renforcement, les commentaires positifs de ceux qui avaient participé à l’édition 2017 et les encouragements des professeurs du département pour qu’on y participe rendaient cela vraiment intéressant. J’ai donc décidé de poser ma candidature !
2. Qu’as-tu aimé le plus ?
Katya : Les gens. Le calibre intellectuel et la diversité des recherches des participants étaient fascinants. J’ai aussi bien aimé que l’horaire laisse beaucoup de temps au réseautage, notamment durant le dîner.
Matthew : J’ai été frappé par les aptitudes incroyables des participants et par le nombre de personnes issues d’autres domaines que l’apprentissage automatique et l’apprentissage par renforcement. J’ai eu certaines de mes meilleures conversations avec des physiciens qui souhaitaient appliquer l’apprentissage profond ou l’apprentissage par renforcement à la modélisation de systèmes physiques particuliers. L’École nous a aussi fourni de nombreuses occasions d’établir des relations avec des gens du monde entier. En apprendre plus sur les nombreuses avenues de recherche explorées est une grande source d’inspiration et un rappel que nos propres intérêts ne sont qu’une infime partie du tableau.
Raksha : Comme l’École met l’accent à la fois sur l’apprentissage profond et l’apprentissage par renforcement, j’ai trouvé que c’était une excellente occasion pour moi, qui travaille en apprentissage par renforcement, d’avoir un aperçu de l’état actuel de la recherche en apprentissage profond et de découvrir d’autres perspectives de l’apprentissage par renforcement. De plus, c’était vraiment agréable de rencontrer des pairs et des chercheurs, et d’interagir avec cette communauté élargie dans un cadre plus scolaire !
3. L’École d’été a-t-elle eu des répercussions sur ta carrière ou ta trajectoire de recherche ? Si oui, comment ?
Katya : Des répercussions incroyables ! Pendant une des pauses, nous discutions avec Rich Sutton d’un nouveau projet. Quatre mois après cette discussion, je visitais le Laboratoire RLAI à l’Université de l’Alberta et travaillais avec Rich sur ce projet, au croisement du traitement automatique des langues et de l’apprentissage par renforcement. C’est un bon exemple d’une conversation qui se transforme en expérience marquante – pas seulement en raison de l’hiver albertain (je viens de Sibérie !) –, mais parce que je me suis retrouvée dans la mecque de l’apprentissage par renforcement et que j’ai eu la chance d’apprendre de Rich et du groupe.
Matthew : Je dirais que cela a consolidé ma trajectoire de recherche. Mes intérêts sont toujours les mêmes – l’utilisation de l’apprentissage par renforcement pour faire des prédictions sur le monde au moyen d’interactions –, mais ce sujet et ses relations avec l’univers de l’IA dans son ensemble m’allument davantage. Le champ de mes recherches s’est également élargi. Alors qu’auparavant, je pensais que je devais suivre une voie étroite pour aller de l’avant, je me tourne maintenant vers de nombreuses communautés que je n’avais pas envisagées (ou que je ne connaissais même pas !).
Raksha : [La] chance de discuter de mes recherches avec les plus grands chercheurs, de les écouter parler de leurs expériences et de leurs idées à propos de ce qui s’en vient et de l’endroit où nous nous dirigeons, de rencontrer des gens et d’interagir avec des pairs qui rencontrent des problèmes intéressants dans le cadre de leurs recherches… Tout cela a été très inspirant !
4. Quelle est la chose la plus utile que tu as apprise ou expérimentée ?
Katya : Être capable d’entrer en relation avec les gens et d’apprendre d’eux… comme avoir la chance de poser des centaines de questions à Graham Neubig et d’obtenir les réponses les plus pratiques, en apprendre plus sur le laboratoire RLAI des membres mêmes du groupe et interagir avec eux, être profondément inspirée par Martha White et Jamie Kiros.
Matthew : Ce que j’ai trouvé de plus bénéfique, c’est que ma vision du domaine s’est élargie. En étant exposé à des concepts et à des sujets que je ne connaissais pas, j’ai eu accès à de nombreuses idées que je veux explorer dans mes futures recherches. Cela m’a donné un aperçu de l’énorme quantité de travail qu’il reste à faire, mais aussi des nombreux sujets intéressants encore inexplorés.
Raksha : J’ai toujours entendu dire que la communauté [de l’apprentissage profond et de l’apprentissage par renforcement] était vaste et diversifiée, mais c’était la première fois que j’en faisais vraiment l’expérience. Les rencontres et les interactions avec des pairs de divers pays étaient inestimables ! L’École d’été a débordé d’énergie dès le premier jour, et nous avons eu un aperçu de l’étendue des recherches dans la communauté dans un cadre scolaire, ce qui a été extrêmement intéressant !