Deep Learning makes Sci-Fi real
Artificial Intelligence (AI) is a term that Gregory Renard despises because of the sensational Hollywood movie connotation that it brings. However, I will refer to AI here by the narrower branch of it, Machine Learning (ML), and its subset, Deep Learning (DL).
Following Gregory and Louis Monier’s workshop on Deep Learning last week, I can say that the timeline of ML/DL resembles a science-fiction novel; full of crazy ideas, fascination and reluctance from human beings, famous for their resistance to change. Although it’s misattributed to Darwin, I am fond of the following quote:
“It is not the strongest of the species that survives, nor the most intelligent, but rather the one most adaptable to change.”
As a Sci-Fi fan, I want to take the example of the Ousters in Dan Simmons’ Hyperion; creatures that were once humans who refused to cling to the past and successfully adapted to a new environment. Now, I know some would laugh at my idea of comparing ML entering our future daily activities to having wings growing on our backs (as some Ousters did), but my point is: there is no need to resist change, as there is no need to resist ML. It’s coming, whether you care or not. And it’s not about whether or not we should be scared; we may be scared of the unknown, but since ML is real we should embrace it, the same way as we embrace our lives despite the reality of our time on earth being ridiculously limited.
The first exciting thing I heard at the presentation and workshop is that Louis and Gregory’s team do actually take inspiration from Hollywood movies! Some AI models that we see on the wide screen are actually doable. Not the hoverboard though (crying inside), and not as visually mind-blowing for the user experience (UX)… yet. Back to the timeline of AI/ML, Gregory has worked in this field for over 15 years, and for the first 13 years people who crossed his path saw him as insane. Nowadays, many want to listen to him talk about his experience and discoveries.
The two men gave us two reasons for not fearing DL. Gregory does care about helping people and worries about the prospect of jobs being lost. He is the living proof that there are experts in the industry who use ethics to enhance our experience with DL. Louis has optimistic views on the future dominated by DL and is spreading his positive energy. Both emphasize that they do not know what the future holds. This being said, Louis thinks that not everyone will learn DL, nor do they have to. I agree that we cannot expect everyone to understand or even care about DL, but I do expect that the number of people using any type of DL device will be higher than the current number of internet users. Therefore UX is at the heart of our challenge. And that very UX constitutes a common ground between DL and software engineering; it doesn’t matter if you have exceptional hard skills and even soft skills, if your product is not user-friendly or doesn’t solve a problem, nobody will bother to try it.
The one aspect I believe could be improved for future DL meet-ups is the workshop; it was rather short compared to the presentation, and it would have been great to spend more time on testing how to train a model so as to have a better idea on how a total newbie should get started. Of course, I appreciate it was challenging to squeeze so much in a single day.
Later, I asked Gregory for advice on how to get started in DL, and his answer is :
“Find a problem you want to solve, and you’ll see that you’ll learn all about how to solve it.”
When I look at how internet has helped my parents finding information and communicating with family and friends, I can imagine how this has contributed to making their lives better, compared to the time when they had to wait weeks to read news from their loved ones, who lived thousands of kilometers away. The problem I want to solve is building a device that not only makes life more enjoyable, but is also easy to handle for the non-tech savvy users. In the long run, it would be great to have a virtual assistant that can, among other things, do your paperwork compliant to the regulations in place, buy and carry heavy groceries or even cook based on your tastes and diet constraints. Before we get carried away, my first task in making people’s lives better is to start with the smallest problem in order to test a prototype and learn.
What lies ahead of us? Are you scared/excited/couldn’t care less? I hope you’re not part of the last group. It will come; that’s what we know. And it doesn’t have to be the hard way.