Organization Name: Facebook AI Research (FAIR)
Web references: Facebook AI Research
Technology and social media industry giants like Facebook, Google, Apple and Microsoft are investing heavily in Artificial Intelligence (AI), which they see as an important advance in your online experience.
In reality all that it means is computer code attempts to mimic human intelligence that can be applied in a variety of ways.
Stanford University defines it as:
It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.
Cleverbot is a classic example of AI. Cleverbot uses an artificial intelligence algorithm to have conversations with humans. It learns from humans, remembering words within its AI. Since launching on the web in 1997, the number of conversations held has exceeded 200 million.
AI has been a goal of computer researchers since the late 1940s. Alan Turing famously created the “Turing Test” in 1950 that is still used today to determine whether a computer algorithm is indistinguishable from human interaction based on conversation.
The image below from MIT Technology Review provides a brief timeline on the history of Artificial Intelligence.
Where is Artificial Intelligence going?
Facebook AI Research (FAIR), is Facebook’s high-profile research and technology group that is leading the way in artificial intelligence development.
Unsurprisingly, they have an active Facebook account that provides updates and outreach on the various projects they are working on.
Facial Recognition is considered by many to be a practical application of AI that is achievable in the near future.
FAIR recently launched Moments, a standalone app that uses facial recognition technology to help people privately organize and share their photos with friends. These grouped photos can then be easily shared on Facebook.
Yann LeCun discusses how it works in the video below:
In the longer term, Deep Learning is considered the goal for artificial intelligence.
In technical terms, Deep Learning is based on a set of algorithms that attempt to model high-level abstractions in data by using multiple processing layers with complex structures or otherwise, composed of multiple non-linear transformations.
As described by Yann Lecun:
Animals and humans can learn to see, perceive, act, and communicate with an efficiency that no Machine Learning method can approach. The brains of humans and animals are “deep”, in the sense that each action is the result of a long chain of synaptic communications (many layers of processing). We are currently researching efficient learning algorithms for such “deep architectures”. We are currently concentrating on unsupervised learning algorithms that can be used to produce deep hierarchies of features for visual recognition. We surmise that understanding deep learning will not only enable us to build more intelligent machines, but will also help us understand human intelligence and the mechanisms of human learning.
As an example, in 2013 a Google Deep Learning system that had been shown 10 million images from YouTube videos proved almost twice as good as any previous image recognition effort at identifying objects such as cats.
Deep Learning is considered important as it would allow for efficiency in computer processing that will for us to interact with computers in ways that we can barely imagine.
In an interview with MIT Technology Review, Facebook’s Yann LeCun had this to say:
Language in itself is not that complicated, what’s complicated is having a deep understanding of language and the world that gives you common sense. That’s what we’re really interested in building into machines
Our relationship with the digital world will completely change due to intelligent agents you can interact with. Systems like this should be able to understand not just what people would be entertained by, but what they need to see regardless of whether they will enjoy it.
Facebook is looking towards a future where everybody can have a virtual online personal assistant that helps sort through the vast web of online interactions and possibilities to personalize the experience. The interface for this is not by typing, but rather through conversation.
This may remind you of Siri, Apple’s voice interface, but Siri is really only a clever front-end for an interface that searches using keywords that it is programmed to recognize. It does not learn or understand context.
Siri Co-Founder Matt Cheyer explains the limitations of Siri:
Book a four-star restaurant in Boston seems pretty straightforward,until you realize that Book is a city in the US, and Star is also a city in the US, and there are 13 Bostons, and Star is also the name of a restaurant.
The AI equivalent of Siri would understand the words being spoken and the context.
Director of Facebook Engineering, Mike Schroepfer explains:
You can engage with a system that can really understand concepts and language at a much higher level, I think in the near term a version of that is very realizable.
Eventually it is like this super-intelligent helper that’s plugged in to all the information streams in the world.
What’s in it for Facebook? In addition to providing a personalized online interface, Facebook could also then use this technology to further refine advertising and offers to be specific and of more interest than the standard banner ad of today.
Would you trust an AI assistant?
The advantages of an AI deep learning assistant are clear, you can have on online experience that is personal and tailored to your needs.
Consider how many times you see advertisements online for which you could not possibly have any interest in – if only they knew anything about you. For example, if you don’t own a car, should you see advertisements for winter tires?
Additionally, the online world is so vast that there are surely places to go that would be perfectly suited to your interests that you will never find – but an AI assistant would be able to direct you there.
What About the Downsides?
Consider that if the AI assistant learns about you “as you are” and crafts your online world around that, how can you grow and change?
There is certainly the potential to be placed into buckets and end up never having your views challenged. For example, if your AI assistant determines that you are liberal-minded, how would you know about the conservative point of view on a subject?
Also, how much do you trust Facebook (among others) with your information?
Thinking about a future where software not only collects data on your online activity, but is able to organically learn from you, consider your demographic and really understand who you are and what you want.
This information would be extremely valuable as a way to control what you see and potentially even what you think.
Even if you trust Facebook with this information, how can you be sure that this very personal information about you is safe from hackers, governments and advertisers who would want to exploit you?
The major players in today’s online experience are betting heavily on artificial intelligence as the next step in our digital evolution. Although this has the potential to greatly improve our lives, it also amplifies the concerns that we are today dealing with, in regards to security and privacy.
If you would like to learn more about this balance between hope and fear, the podcast below consists of 2 panel discussions.
The first “This is for everyone: The Hopes of the pioneers” outlines the vision of the early internet pioneers, while the second is a counterpoint “Artificial Intelligence: Are we engineering our own obsolescence?”.
Submitted by: David Pearson
To contact the author of this entry please email at: firstname.lastname@example.org
If you have concerns as to the accuracy of anything posted on this site please send your concerns to Peter Carr, Programme Director, Social Media for Business Performance.