During my time at UNESCO as a representative of Quebec, I had the chance to meet experts in artificial intelligence (AI). I am fascinated by this technology which, for me, will change the face of the world.
The ethical debates at UNESCO, but also in the AI community, then essentially concerned its impact on the job market and its malicious use to manipulate people, in particular by making fake videos and audio recordings of great credibility.
This particularly worrying phenomenon, called deepfakewas born in 2017. Already in 2018, Deeptrace, a cybersecurity firm, identified more than 8,000 such videos and 15,000 in 2019. The vast majority of these videos concerned pornography and featured personalities, mostly women, but also ordinary citizens.
We may worry about the fact that this technology can hijack elections, but it is still pornography that dominates. Although in 2020, the American elections were the field of experimentation for this type of video, the fact remains that its impact has not been validated by empirical data.
Murders and pedophilia
My fascination with AI led me to reflect on its criminal potential. I talked about it at length with several experts who looked at me a bit like a weirdo. They thought I was surfing a little too much on science fiction, since artificial intelligence is still in its infancy.
My concerns? Robots as identical as humans with high-performance artificial intelligence that would be used to satisfy the vilest perversions. Robots in the likeness of children or women that it would be possible to rape or even kill, again and again.
I even talked about it to a criminologist, a professor at a university in Quebec. And you know what he answered me…? This could prevent rapes. My arms fell, especially since this observation came from a professor of criminology. I felt like I was back in the prehistoric era where it was still believed that prostitution would protect women from sexual assault. Bullshit !
AI coaching
When experts raise red flags and call for a framework for AI, they are raising the same concerns I heard in 2018 at UNESCO: the labor market and the deepfake.
A short-term view that is anchored, of course, in the current advances in this technology, but which obscures other fundamental ethical questions that will determine the kind of humanity that we want to bequeath to future generations.
Are we going to ban the production of child robots? Are we going to give legal status to the machine as we do for animals in order to protect the mental health of humans? Are we going to introduce AI hygiene into the education system as we are currently trying to do with social networks?
Several questions remain and we must not wait to be overwhelmed by technology to answer them.