Blog

3 Things I Learned At AI World Forum in Toronto This Week

aiworldforumlogo-1000x563-3.pngWhen you're honored to speak at a conference, it is appropriate to shut up and listen.

I tried to use my eyes a bit, too.

Here's what I found.

AI is concerning - but we don't know why. 

On the lips of people at the AI World Forum in Toronto this week, there were three main concerns voiced:

  1. Loss of jobs
  2. Loss of control
  3. Loss of privacy (or what shreds remain)

Let's look deeper at each of these.

  • Loss of Jobs. Four different times I listened to the same debate. Party A would say that AI will cost people jobs. Party B would counter with, "But every technological revolution, such as the first industrial revolution, saw a dearth of jobs explode as a result." Statistics accompanied the counter-argument, and if accepted by Party A then both parties would agree that a large social change is on the horizon. Both parties would then agree that haves-and-have-nots would emerge. The conversation would fizzle out. I don't know why, but perhaps because it was the experts and attendees focused on AI who were discussing this and that we could tacitly presume we'd be in the 'haves,' so wait-and-see was a viable short-term solution.

    While I wholeheartedly believe more jobs will be created, I know that it may not be a smooth path. There will be heated debates. And rightly so, because the more people looking at the problem, the faster and better we can solve it.

  • Loss of Control. This concern, also frequently discussed, is easier to explain. Deploying AI, even defining AI, seems to be hard work. The term is confusing and there are spins that journalists, science fiction, marketing departments and probably your author, here, throw into the already-confused discussion. So of course it is out of control - it's a storm of concepts that are changing, that none of us are quite defining, and, oh, it'll take our jobs. It's out of control! But did any of us 'control' the world wide web? Television? Telephony? Did any of us 'control' the printing press?

    If we look at any of the great technologies of our history they've never been 'controlled.' They evolve with us, and that's fine. Even the greatest of our inventions (language or mathematics or governance) evolve with us because they are a part of us, and the stories of our species.

  • Loss of Privacy. I heard, roughly, "AI can watch everything and learn and it never forgets" discussed several times, and a gradually bleak sense of privacy was discussed. After its mention Facebook, Google, and other companies that publicise our private data were noted as the companies with the best AI. This link was then pursued and a thin, dark forensic line led from the death of privacy, like a little stream of blood, to the birth of powerful AI.

    I find this to be a common thread. However, I believe it is a solvable problem, and I continue to strive for new and innovative ways to solve it. One way is to ensure authentication with end-to-end encryption. Reach out if you're interested in learning more. I love this topic.

Conversational User Interfaces are missing, however it is assumed we will - eventually - be talking with something.

aiworldforum2017-2.jpg

We had a veritable zoo of slides on the screen as people talked about AI. Charts, cartoons, diagrams, and maps of the future with upward curves. Curiously many of the presenters' images were of a robot doing something like holding a wooden cube up to the light, or typing at a laptop.  And most of these fictional robots were represented as shiny-white Aryan robots of some sort, dark graphene cogs underneath a shiny sleek carapace of a stern face that looked like a passionless, hairless Clark Kent.

I find this disturbing and won't detail the trend (it was happening at AI Europe, last week as well), but I don't like it one bit. None of these fictional robots were talking, just engrossed in some peurile meditation. Which is not a depiction of reality and fuels the trend of "we're afraid of AI but we don't know why." Time to change that. Let's talk about building humane machines instead.


AI is now fancy and can draw a diverse, black-tie party.

Image uploaded from iOS.jpgToronto is a vision of the future in that when you walk down the street it is a global city. AI World Forum was worldly and, as a conference, benefitted because Toronto itself is worldly.  So we can expect, with such a diverse population from so many walks of life, Toronto will continue to push forward and lead the charge technologically. Oh, and the organizers of AI World Forum did a fine job of throwing a cocktail party at the Carlu. Let's hope they do it again next year.