Whether we call them Deus Ex Machina (from The Matrix) or Terminator,
artificial intelligence systems have been at the center of multiple science
fiction movies. The box-office success of such movies makes it safe to say that
the public’s interest is sparked by the idea of a “robotic superintelligence”
that could exceed human cognitive capacities. Still, the artificial
intelligence field remains unknown to many and it is an exciting domain in
which Canada (and Montreal in particular) is starting to emerge as a
significant player. In this context, I sat down with Dave Anctil, a philosophy
professor at College Jean-de-Brébeuf, who teaches a class entitled “Penser
l’intelligence artificielle” (Thinking artificial intelligence).
1) What are your qualifications regarding
the subject of AI? Have you ever worked directly in the field?
First of all, as an ethicist and a political scientist, I have been
studying closely military intelligent systems for about 15 years. More
recently, as a cognitive scientist, I work in many fields of AI and social
robotics. For example, I contributed recently to an experimental study in the
emerging fields of “digital sexualities” and “erobotics”. My expertise in
ethics, law, politics and the cognitive sciences is useful to help understand
and solve some of the more complex issues surrounding the rise of AI and
robotics in the 21st century. But I am no computer scientist! I understand well
enough the technology – I even have a few programming projects of my own! –,
and I hope I can help with some of the difficult issues surrounding autonomous
systems, to make the technology beneficial to society as a whole.
2) What is deep learning?
Deep learning (DL) is just a general approach to machine learning (ML).
DL techniques are loosely inspired by the way the brain works, in particular,
the general idea of neural networks. AI scientists call it “deep” (or
many-layered) in reference to the many layers of neural nets that are used to
generate the spectacular results of AlphaGo and AlphaStar, as well as the many
other powerful AI systems you can read about every day now. There is a big hype
surrounding DL right now. But you must never forget that there are many other
techniques of ML used in the AI field!
3) Will artificial
intelligence steal our jobs? Which economic sectors could be impacted?
And will we, as a society, benefit from these technological advances?
Any major technological revolution massively impacts the economy and
society as a whole. Think about the industrial revolution and the rise of the
information age. We currently live in -
by far! – the wealthiest period of human history. But we also live in a
pretty unjust world of massive economic inequalities. The wealth we create
mostly benefits the top 1%. Our globalized capitalist system is unable to bring
real prosperity for the bottom 99%. We live in a world organized for – and
controlled by – billionaires. If we actually lived in a fair and just world, I
would be very concerned about the rise of AI and robots. But we don’t. Our
world is on the brink of ecological collapse. That is the only issue we should
be really scared about. I think we should fear climate change and the
concentration of wealth instead of computers and robots taking over.
Controlling the development of new technologies is a matter of democratic
decision-making. It is also a matter of ethics and law. We can make the current
transition a success, but only if we, the citizens of the world, mobilize to
bring beneficial changes in the world. We only (!) need wise leaders and
informed citizens…
4) What studies should a student pursue if
she/he wants to work in the field of AI?
Soon enough, everybody will work with new intelligent systems. The
future of AI is a future of human-machine collaboration. Think about it: 40 or
30 years ago, very few people worked with computers and fewer still had them at
home. Now we all work, play and socialize with computers all the time. I think
it will be the same with AI systems, digital and physical robots. These agents
will be ubiquitous, like your smartphone and apps that you use without even
thinking about it. AI systems, in particular, will help us get better at many
complex tasks in many fields. So I would say to any student interested in AI:
don’t become a computer scientist if you don’t like coding or advanced maths!
Instead, follow the evolution of technology, learn the basic concepts, be
open-minded about it all… and it will be easier for you to adapt. Education
systems are currently trying to figure out how to adapt. Soon enough, applied
AI courses will become available (or even mandatory) in most college and
university curricula.
5) Is there a possibility
of an arms race with the Chinese government concerning AI?
Well, it’s not just a possibility. It’s a reality. Most leaders in the
AI field – Joshua Bengio, Stuart Russell, Yann LeCun – are very preoccupied
with the development of intelligent systems in warfare. Not only military
drones and “killer bots”, but also AI systems capable of massively disturbing
our digital societies like powerful engines of mass manipulation taking over
social media or intelligent agents capable of corrupting or bringing down our
infrastructures. Cyber warfare is nothing new. Bu AI systems and robots are
scary new tools in the hands of dictators. China is a new superpower but with
no democratic or judiciary oversight. You can bet that many military projects
are currently developed by Chinese computer scientists and roboticists working
for the military. They already have supercomputers and all the means and
infrastructures to beat the Americans and Europeans…
6) What are the main
ethical factors concerning AI?
They are well known: programming beneficial, unbiased AI systems with
implemented democratic, ethical goals and rules; protecting the workforce
against the new industrial revolution; limiting the use of AI in warfare; etc.
You can all read the Montreal Declaration for a Responsible Development of AI.
It’s all there. But I would give my own personal, different take about that.
I have dedicated the biggest part of my life to studying the “dark
sides” and ethical limitations of humanity: war, violence, domination,
exploitation, corruption, genocide, ecocide… As a philosopher and a scientist,
these have been the major subjects of my research for many years. For me, the
real ethical factor is the following: can we create better versions of
ourselves to help protect mankind against itself? We need intelligent systems
to help us make better decisions than those we have made and that we currently
make. We already do this in many domains (finance, science, medicine…). Can we
do it elsewhere to protect the environment and make the economy great again?
Can we make our democratic and judiciary institutions better? Can we become
wiser and – yes – more intelligent in our collective decision-making? I don’t
know if we can. But we must try. And AI could help us do it. Because right now,
despite all of our knowledge and capabilities, we suck!
7) What are the challenges
that you face while teaching about AI?
I just started and I’m having a blast! Soon enough, I will discover my
many failings… But for now, I am just teaching this awesome stuff, you know? As
a big reader of science fiction… it’s like… making science-fiction real!
- Marc-André Otis (Sciences Humaines)
Comments
Post a Comment