The Dawn of Artificial Intelligence

What You Missed From The First Congressional Hearing On AI

Sebastien Dery
5 min readDec 4, 2016

#tldr A Congressional Hearing took place last week with the title “The Dawn of Artificial Intelligence”. What did you miss? Lots of strong words (transformative, innovation, exponential), lots of predictions with high figures, and a few interesting points to think about. Read my takeaways for a quick briefing on what has been discussed.

A Congressional Hearing took place last week with the title “The Dawn of Artificial Intelligence”. Senators and watchers alike were treated to a little more than one and half hours of Q and A with 4 well known (and little less known) individuals within the field of Machine Learning: Eric Horvitz, Andrew Moore, Greg Brockman, Steve Chien. Perhaps an attempt by congress to provide some form of diversity, one is tempted to see from left to right a representative from industry, academia, non-profit and governmental organizations (although they all transcended those boundaries in some way or another at some point in their life). The missing category, women, was rapidly called out by many in the following hours and even by one of the panelist during the session (good men Mr. Brockman). It is admittedly a topic Silicon Valley is very much aware, and sensitive to, all the while struggling with any concrete solution. Baby steps.

But this hearing wasn’t about diversity.

It was about competitive advantages.

Sound like a Déjà Vu?

The year is 1955, when the Soviet Union responded to the US announcement of launching artificial satellites for the International Geophysical Year, by declaring they would also launch a satellite “in the near future”; a competition between the two Cold War rivals ensued for supremacy in spaceflight capability. This Space Race changed the world in many ways; for one thing it offered unprecedented images of our planet, reshaped our vision of humanity, many spin-off technologies but also, and quite importantly in fact, it offered a unifying goal for a world not so long before plunged into the atrocity of war.

We choose to go to the Moon in this decade and do the other things, not because they are easy, but because they are hard; because that goal will serve to organize and measure the best of our energies and skills […]” — John F. Kennedy

Fast forward 60 years and we turn our gaze away from the stars and towards Artificial Intelligence for promises of a better world.

Elon Musk, Stephen Hawking, Ray Kurzwell, Pieter Abbeel, and many others were asked for, or voluntarily shared, their predictions about AI in what ended being a colorful amalgam of thoughts and ideas. Making us safer, becoming cyborgs, turning us into super-humans, solving all our problems, a change so slowly we won’t notice, become our friends, be our enemies, improve medical care, understand us better, allowing us to be more creative, provide more leisure time, revolutionize how we live and work, and take our jobs.

Hard to find a real consensus in this prediction spectrum. The one truth that seems to emerge is that it’s making each individual rethink their core values. What would they do without the need for job? What is worth doing?

Wherever you end up in the spectrum, I think it’s an interesting mental exercise in sanity to replace the word “predicted” by “pulled out of my ass”.

No disrespect to expert opinion of course, it only aims to offer a grain of humility facing seemingly unpredictable complexity and to encourage everyone in educating themselves. Planning is important but plans are useless.

In the meantime a few definition just to clear any confusion

  • Artificial Intelligence: Machine Learning.
  • Machine Learning: Clever use of statistics and algorithms.
  • Algorithm: Process or set of rules to be followed by a computer.
  • Statistics: Discipline concerned with the analysis of data and decision making based upon data.
  • Data: Quantities, characters, or symbols on which operations are performed by a computer.

Here are my takeaway from the hearing

  • The world is facing a talent war for the best and brightest in the field of machine learning.
  • China is considered a real threat to the alleged US leadership in AI.
  • Companies should collaborate on the Science and compete on the market.
  • Technology should not be owned by any company.
  • There is a costly discrepancy in the amount of efforts devoted to hard problems (e.g. homelessness, addiction) versus industry backed endeavors.
  • We should be focusing on encouraging a new generation of computer science literate citizens to fuel the workforce of tomorrow.
  • We need to make tools available to small entrepreneurs so they can compete against much larger Machine Learning shop.
  • Let’s invest in statisticians as an integral part of the Machine Learning development cycle.
  • Focus on global measurement contest (e.g. Kaggle) as the only way to throw hype in the garbage. It’s easy to say you’re doing Deep Learning to predict the future; this recommendation aims to put you to the test.
  • Challenges of ethics, morality and diversity have not been fleshed out from a process standpoint; a crucial point that requires our collective attention.

Senator Ted Cruz deliberately set the stage for this hearing as an historical one when declaring it to be the first of many. In that respect I agree; if we are to shape Artificial Intelligence into the panacea we claim it to be we we’ll need the voice of many, from as many background as possible, and many times before having the semblance of a clear picture.

“To build may have to be the slow and laborious task of years. To destroy can be the thoughtless act of a single day.

--

--

Sebastien Dery

Canadian in Silicon Valley, ML @ Apple #Philosophy #StoneSculptor #Divemaster