We have seen other great developments such as with image recognition, where we can one day expect to see computers that will be able to read X-ray, MRI and CT scans more efficiently than radiologists, enabling the quicker diagnosis of cancer. This is just one example of how the progress of deep learning is rapidly advancing and impacting the world we live in, from the way we shop to predicting energy sources to shaping modes of transport. We asked some of our influential speakers, who will be presenting at our deep learning summits this year, for their predictions for deep learning in 2017. In 2017, we will probably see further rapid exploration of applications of current deep learning techniques, as well as further theoretical advances, improving robustness and sample efficiency. We will also see various fun new applications of deep learning to image and voice resynthesis.

Last week IBM focused attention on "five technologies that [they] believe have the potential to change the way people work, live, and interact during the next five years." They call their vision "5 in 5". The technologies they chose all have to do with enhancing our ability to visualize the world from the micro to the macro level. Here's what IBM sees in our future. Mental and physical disorders with a neurophysiological basis such as Alzheimer's and Parkinson's disease can affect the language processing areas in the brain.

Neural networks have the rather uncanny knack for turning meaning into numbers. Data flows from the input to the output, getting pushed through a series of transformations which process the data into increasingly abstruse vectors of representations. These numbers, the activations of the network, carry useful information from one layer of the network to the next, and are believed to represent the data at different layers of abstraction. But the vectors themselves have thus far defied interpretation. In this blog post I put forward a possible interpretation of these vectors. I argue we shouldn't take these vectors literally, but rather as an encoding for a simpler, sparse data structure.

Last week I went to the workshops at NIPS (biggest ML conference in the world) and I also attended part of the ML and the Law symposium the day before. I found out a little bit too late about the symposia but I was still able to attend two panels on which there were both lawyers and computer scientists. They were very insightful and informative -- did you know that this Spring, the European Union passed a regulation giving its citizens a "right to an explanation" for decisions made by machine-learning systems? The panel discussions were motivated by the problem of explaining ML-powered decisions which have an important impact on people's lives: We need to be able to test how systems get to their conclusions; if we can't test, we can't contest. Individuals are entitled to know which data is being processed of them, and to explanations of how predictions & decisions work, in terms they can understand.

Artificial Intelligence is becoming more and more prevalent in every day life as we see the technology adopted in everything from digital assistants to autonomous vehicles. One sector that has huge potential for AI is retail. You might not know it, but if you've ever submitted an online query to a retailer, then you've probably already spoken to an AI. Brands and companies are quickly beginning to realise the benefits of automation, applying AI not only to behind-the-scenes operations but also to customer services. This is causing huge changes to the way that retail companies work, from tourism to banking. With AI startups now offering adaptable software, it's easier than ever for businesses to integrate the tech into their business strategies.

From winning Jeopardy in 2011 to helping write a sad song last year, IBM's Watson cognitive computing platform is all over popular culture. Press releases fly out about Watson producing a movie trailer, powering a Macy's shopping app, even controlling lights on an internet-connected dress -- along with more serious applications like working on cancer treatments. It seems, from IBM's hype, that Watson can do everything. But Bernie Meyerson, IBM's chief innovation officer, wants to dial back the hype in some ways, calling Watson "just the first step on a very, very long road." Watson can be helpful in a lot of industries, such as medicine, which are awash in data, but it can't replace people, he says.

Chinese social media is currently alight with the news that some of the country's top Go players have been beaten multiple times, not by a human, but by a robot, which has now been revealed to be a software program called AlphaGo run on Google's deep learning neural network DeepMind. In March 2016, DeepMind made international news when the AlphaGo program succeeded in beating the Go world champion player Lee Sedo, 33, from South Korea by 4-1. DeepMind is a neural network – essentially a large web of artificially intelligent classical computers that are trained using computer algorithms to solve complex problems in a similar way to the human central nervous system. The computers are separated into different groups known as'layers' to examine different parts of the problem, and each layer's answer is then combined to produce a final answer. Go is an ancient Chinese abstract strategy game that originated over 2,500 years ago and is considered to be more complex than chess.

One of the undisputable realities of moving into a digital work era is the continuous improvement in technologies which mimic and replicate what human employees are doing. The introduction of'bots' into the workplace to perform logic-based and repetitive tasks is becoming a common occurrence. These task robots are able to perform activities faster, with greater accuracy and more efficiently that their human colleagues. In fact their capacity is close to 700 per cent greater than the human employee, who generally works at a 60 per cent utilisation rate for 7-8 hours a day, can be absent for a variety of reasons, doesn't work 7 days a week, and who's productivity is influenced by a plethora of human frailties. It's no wonder'bots' are attractive to organisations for this type of work.

With the 2016 coming close, Facebook CEO Mark Zuckerberg has completed his personal challenge to built "Jarvis," his AI-powered personal assistant. In his lengthy blog post, Zuck describes the types of tasks his Jarvis bot can accomplish. The personal assistant or what few call a butler, is customized to perform multiple actions at Mark's residence. Mark's simple AI to run his home -- like Jarvis in Iron Man, uses several AI techniques including speech and face recognition, reinforcement learning, natural language processing. It is written in PHP, Objective C and Python. In coming years, Zuckerberg has expansion plans for his AI Bot, he cited building an Android app and setting up the Jarvis voice terminals in the entire house.

India: American technology major Nvidia has launched the Nvidia Inception programme in India, in recognition of the country's budding innovation ecosystem in Artificial Intelligence (AI). Inception is a virtual incubator programme to support startups with revolutionary ideas in AI. Members will receive a custom set of benefits, from hardware grants and marketing support to training with deep learning experts. The Inception Programme was launched in India at the inaugural Nvidia Emerging Companies Summit India, part of the GPU Technology Conference (GTCx), a platform for the brightest minds and greatest ideas in GPU computing. The momentum around AI among Indian innovators is so significant that, at launch, the Inception Programme already has close to 100 Indian startups as members.