For the 1964 World Fair, science fiction author Isaac Asimov wrote an article for the New York Times, envisioning what the exhibits at the event would look like in fifty years' time. Asimov's predictions were scrutinized and used in numerous think pieces and tech forecasts of 2014, the year that marked the passing of the five decades since the article's publish date. Since a large body of Asimov's work concerned itself with human relationship with artificial intelligence, much attention was focused on the following quote: "If machines are that smart today, what may not be in the works 50 years hence? It will be such computers, much miniaturized, that will serve as the "brains" of robots." Most writers summarized that, while the closest we have to an android housekeeper is a Roomba, Asimov was right to draw the parallel between brains and computers.

These days you just need some artificial intelligence, which can be delivered through your smartphone. Two of this summer's much buzzed-about apps are Prisma, which turns your photos into what look like paintings, and Artisto, which does the same for videos. Both are available for iOS and Android. Whether the results are true works of art can be debated, but millions of the apps' users are having fun trying to imitate the masters. Prisma uses neural networks and deep learning algorithms to process photos and make them look like art, applying styles such as that of The Scream by Edvard Munch or the pop art of a Roy Lichtenstein.

Speech recognition software has struggled to match the precision of typing. But over the last few years, thanks to deep learning and big data collection, those programs have improved dramatically. A new study from Stanford put the two competing interface methods to the test and found that it's now a lot faster to dictate your prose to your phone rather than typing it. In the study, volunteers were asked to either type or speak 100 phrases. The experiment was conducted in both English and Mandarin Chinese.

Nine times out of ten, when you hear about deep learning breaking a new technological barrier, Convolutional Neural Networks are involved. Also called CNNs or ConvNets, these are the workhorse of the deep neural network field. They have learned to sort images into categories even better than humans in some cases. If there's one method out there that justifies the hype, it is CNNs. What's especially cool about them is that they are easy to understand, at least when you break them down into their basic parts. I'll walk you through it.

Originally developed in the late 1990's by Jürgen Schmidhuber, the LSTM block allows a part of the neural network to store a memory cell, and have gates to control whether that memory cell can be overwritten by an input, forgotten, or allowed to be fed to the output gates, kind of like an actual memory cell in a computer. The main difference is that in a computer's memory cell, everything is either one or off (1 or 0), whereas in the LSTM network, the cells will be from zero to one, controlled by a sigmoid function (Although in a memory cell, the actual voltage in the transistors can be closer to a sigmoid function than just 1 or 0). The network can also be trained via stochastic gradient descent, as the entire network can be differentiated and back propagation through time can be applied to train the weights. The advantage of this network is that memories can be stored indefinitely, while normal recurrent networks composed of only sigmoid functions can lose their states (or memory) quickly. Wonders can be done with LSTM especially in the area of speech recognition, and recently in image recognition.

August saw the Defense Advanced Research Projects Agency (DARPA) host its first Cyber Grand Challenge – the first hacking competition not involving people. During this event, teams left their systems alone to single-handedly find, diagnose and fix software flaws in real time. Elsewhere, researchers at MIT are not only developing machine learning systems that automatically mine dark web marketplaces for vulnerabilities and zero-day attacks and reports them back as well as software that automatically fixes buggy code, but also a platform that can predict 85% of cyber-attacks. Machine learning, deep learning, and Artificial Intelligence (AI) are hot topics at the moment, and while there's plenty of research going on, there's also some practical applications that can be deployed right now to make life easier for cybersecurity professionals. A glut of new start-ups, from the likes of Darktrace, Cylance, Deep Instinct, and HackerONE, plus established player such as FireEye, IBM, and Forcepoint, are all working on bringing self-learning systems into the world of security.

Recently we have seen a slew of popular films that deal with artificial intelligence – most notably The Imitation Game, Chappie, Ex Machina, and Her. However, despite over five decades of research into artificial intelligence, there remain many tasks which are simple for humans that computers cannot do. Given the slow progress of AI, for many the prospect of computers with human-level intelligence seems further away today than it did when Isaac Asimov's classic I, Robot was published in 1950. The fact is, however, that today the development of neuromorphic chips offers a plausible path to realizing human-level artificial intelligence within the next few decades. Starting in the early 2000's there was a realization that neural network models – based on how the human brain works – could solve many tasks that could not be solved by other methods. The buzzphrase'deep learning' has become a catch-all term for neural network models and related techniques, as is shown by a plotting of the frequency of the phrase using Google Trends: Most deep learning practitioners acknowledge that the recent popularity of'deep learning' is driven by hardware, in particular GPUs .

Many of the latest machine learning and data science tools purport to be easy to work with compared to previous generations of such frameworks and libraries. Chinese search engine giant Baidu now has an open source project in the same vein: a machine learning system it claims is easier to train and use because it exposes its functions through Python libraries. PaddlePaddle -- "Paddle" stands for "PArallel Distributed Deep LEarning" -- was developed by Baidu to augment many of its own products with deep learning. Baidu touted PaddlePaddle's speech transcription in Chinese, either for transcribing broadcasts or as a speech-to-text system to replace keyboards in smartphones. The company claims it needed 20,000 hours of audio as training material to achieve these results with its framework.

Others are fed enough data and leverage machine learning, using natural language processing in a way that can pick up keywords and phrases to understand human language. We have some great technologies like the IBM Watson API, Facebook bot API, and Microsoft Bot Framework that can help developers create interactive bots based on different taxonomies. Soumith Chintala of Facebook AI noted in a recent article, "Deep learning -- neural networks that have several stacked layers of neurons, usually accelerated in computation using GPUs -- has seen huge success recently in many fields such as computer vision, speech recognition, and natural language processing, beating the previous state-of-the-art results on a variety of tasks and domains such as language modeling, translation, speech recognition, and object recognition in images." Google announced that its natural language processing system called Parsey MacParseface is now able to identify 94 percent of word dependencies within an English sentence.