Saturday, October 28, 2017

Who 'da Man 2

The Verge has an interview with Yann LeCun, identified as "Facebook's AI chief ", that ties in well with my own understanding of AI, both as to function and the rate and direction of development efforts. Having read the article, I am more convinced than before that "robots" will quickly come to mean "human-controlled mechanistic devices" and "AI" will mean "non-human controlled systems". Robots will be the machines doing things (under direct - more or less - control of human operators). AI's will be complex constructions controlling devices that semi-autonomously operate systems like vehicle traffic control systems, or shipping/receiving/warehousing systems.
I keep repeating this whenever I talk to the public: we’re very far from building truly intelligent machines. All you’re seeing now — all these feats of AI like self-driving cars, interpreting medical images, beating the world champion at Go and so on — these are very narrow intelligences, and they’re really trained for a particular purpose. They’re situations where we can collect a lot of data.
So for example, and I don’t want to minimize at all the engineering and research work done on AlphaGo by our friends at DeepMind, but when [people interpret the development of AlphaGo] as significant process towards general intelligence, it’s wrong. It just isn’t. it’s not because there’s a machine that can beat people at Go, there’ll be intelligent robots running round the streets. It doesn’t even help with that problem, it’s completely separate. Others may claim otherwise, but that’s my personal opinion.

We’re very far from having machines that can learn the most basic things about the world in the way humans and animals can do. Like, yes, in particular areas machines have superhuman performance, but in terms of general intelligence we’re not even close to a rat. This makes a lot of questions people are asking themselves premature. . That’s not to say we shouldn’t think about them, but there’s no danger in the immediate or even medium term. There are real dangers in the department of AI, real risks, but they’re not Terminator scenarios.
Vehicles in the near-term future will undoubtedly have some form of semi-autonomous "AI" that will operate the vehicle without continuous direct human operator input (in limited applications, at least), but the more likely scenario will be AI semi-autonomous operation of a traffic control system allowing vehicle access to highways, stop light timing, variance of traffic speed limits to conform with road conditions (changing weather or traffic load for instance). The vehicles themselves would still require operator input to function, but the AI traffic control system would have direct access to the vehicles internal control system "AI" to limit the operator's choices (presumably with an emergency override function - factory installed or otherwise :)).

I am also pleased that Mr. LeCun agrees that the virtual assistant application is the most likely near-term market that will most broadly interact with the most people (what I called a "data orchestra")
here.
 I think virtual assistants really are going to be the big thing. Current assistants are entirely scripted, with a tree of possible things they can tell you. So that makes the creation of bots really tedious, expensive, and brittle, though they work in certain situations like customer care. The next step will be systems that have a little more learning in them, and that’s one of the things we’re working on at Facebook. Where you have a machine that reads a long text and then answers any questions related to it — that’d be useful as a function.
Having a virtual assistant that you can verbally (and presumably textually or pictorially too) interact with, one that has internal data bases of reference sources, would be an exponential improvement of the education and skills development (training) activities that are central to achieving any success in a technologic society (education is mathematics, history, languages, art; skills development is the construction/maintenance trades, martial arts/physical fitness, ultimately engineering generally). A personal device that can provide you with immediate access to information, when and where you need it, would drastically alter the way and extent of human learning and personal capability.

One that could also call emergency services (or your lawyer) semi-autonomously would no doubt prove useful too.

Someone is going to arrive at the decision to make a business of improving humanity's individual ability to provide for itself, simply as a means to increase his/her customer base in both total numbers as well as in total market range of participation. Maybe that will be Jeff Bezos, but it obviously doesn't have to be. What it continues to seem more and more we can be certain of is that it will be.

No comments: