- AI for Everybody
Artificial intelligence has so far been mainly the plaything of big tech companies like Amazon, Baidu, Google, and Microsoft, as well as some startups. For many other companies and parts of the economy, AI systems are too expensive and too difficult to implement fully.
AI for Everybody
What’s the solution? Machine-learning tools based in the cloud are bringing AI to a far broader audience. So far, Amazon dominates cloud AI with its AWS subsidiary. Google is challenging that with TensorFlow, an open-source AI library that can be used to build other machine-learning software. Recently Google announced Cloud AutoML, a suite of pre-trained systems that could make AI simpler to use.
Microsoft, which has its own AI-powered cloud platform, Azure, is teaming up with Amazon to offer Gluon, an open-source deep-learning library. Gluon is supposed to make building neural nets—a key technology in AI that crudely mimics how the human brain learns—as easy as building a smartphone app.
It is uncertain which of these companies will become the leader in offering AI cloud services. But it is a huge business opportunity for the winners.
These products will be essential if the AI revolution is going to spread more broadly through different parts of the economy.
Currently AI is used mostly in the tech industry, where it has created efficiencies and produced new products and services. But many other businesses and industries have struggled to take advantage of the advances in artificial intelligence. Sectors such as medicine, manufacturing, and energy could also be transformed if they were able to implement the technology more fully, with a huge boost to economic productivity.
Most companies, though, still don’t have enough people who know how to use cloud AI. So Amazon and Google are also setting up consultancy services. Once the cloud puts the technology within the reach of almost everyone, the real AI revolution can begin.
- Dueling Neural Networks
Artificial Intelligence is an excellent technology at identifying things. Even if you show it a million picture, it will tell you with unnatural accuracy, which one depicts as a pedestrian crossing a street. But when it comes to generating an image of pedestrian itself, then at this point AI becomes hopeless. If it could be possible, it would be able to create gobs of realistic but synthetic pictures depicting pedestrians in various settings, which a self-driving car could use to train itself without ever going out on the road.
Dueling Neural Networks
Creating something entirely new needs imagination and that has puzzled AI until now, and here the problem lies.
The first solution occurred to Ian Goodfellow, and then to a PhD student during an academic argument in a bar in the University of Montreal in year 2014. The GAN (Generative Adversarial Network) mainly takes two neural networks. First is the mathematical models of the human brains which are simplified and that underpin most modern machine learning, and the second is to pits them against each other in the digital mouse-and-cat game.
Both networks are trained on the same data set. One, known as the generator, is tasked with creating variations on images it’s already seen, perhaps a picture of a pedestrian with an extra arm. The second, known as the discriminator, is asked to identify whether the example it sees is like the images it has been trained on or a fake produced by the generator—basically, is that three-armed person likely to be real?
With time, the generator can become an excellent source at producing images that can’t be spot faked by the discriminator. The generator essentially has been taught to recognize and then create an images of pedestrian looking realistic. In the past few decades, this technology has become one of the most promising breakthrough in AI that helps to produce results that cab even fool human beings. GANs have especially been put to use creating photorealistic fake imagery and realistic sounding speech. . In one compelling example, researchers from chipmaker Nvidia primed a GAN with celebrity photographs to create hundreds of credible faces of people who don’t exist. Another research group made not unconvincing fake paintings that look like the works of van Gogh. Pushed further, GANs can reimagine images in different ways, making a sunny road appear snowy, or turning horses into zebras.
We are not always end up with perfect results. GANs can produce up bicycles with two pairs of handlebars, and faces with eyebrow in wrong place. But some experts believe that there is a sense in which GANs are dawning to understand the latent structure of the world that they hear and see, and this is because of the reason that sounds and images are often startlingly realistic. So, along with the sense of imagination, AI may gain a more independent ability to make sense of the things that it sees in the world.