AI Business recently interviewed one of the UK's leading experts in AI for Business, George Zarkadakis. George has a PhD in AI, is the Digital Lead at Willis Towers Watson, and the author of "In Our Own Image: will Artificial Intelligence save us or destroy us?". George is also a keynote speaker at The AI Summit London, presenting on the many different ways that AI will be shaping the business of tomorrow. George answered a number of interesting questions giving us a taster of the insightful presentation to come at The AI Summit on the 5th of May. How do you believe AI will impact business overall and in what ways?

With rotors and radios, we are witnessing the birth of a new sport. Aerial acrobatics, sharp turns and staggering climbs that NASCAR can only imagine, all done on the cheap with toys and remote controls, from woods to warehouses to special courses in Dubai. Welcome to the new age of drone racing, as it moves from backyards to broadcasts. Today, ESPN announced it will livestream the U.S. National Drone Racing Championships, a 3-day drone racing event held on New York City's Governor's Island in early August. And then, in October, ESPN will cover the 2016 World Drone Racing Championships at Kualoa Ranch, Hawaii. The drones are flown by pilots wearing goggles and watching through first person vision cameras.

We consider a statistical inverse learning problem, where we observe the image of a function $f$ through a linear operator $A$ at i.i.d. random design points $X_i$, superposed with an additive noise. The distribution of the design points is unknown and can be very general. We analyze simultaneously the direct (estimation of $Af$) and the inverse (estimation of $f$) learning problems. In this general framework, we obtain strong and weak minimax optimal rates of convergence (as the number of observations $n$ grows large) for a large class of spectral regularization methods over regularity classes defined through appropriate source conditions. This improves on or completes previous results obtained in related settings. The optimality of the obtained rates is shown not only in the exponent in $n$ but also in the explicit dependency of the constant factor in the variance of the noise and the radius of the source condition set.

Scientists have been able to develop artificial intelligence (AI) capable of besting humans at their own games, but a new study suggests that people may have the upper hand when it comes to intuitive thinking. A team of researchers led by Denmark's Aarhus University associate professor Jacob Sherson managed to develop a game based around complex theoretical science in which human players were "able to find solutions to difficult problems associated with the task of quantum computing," whereas computerized numerical optimization failed, according to the scientists' findings published in Nature. "The big surprise we had was that some of the players actually had solutions that were of higher quality and of shorter duration than any computer algorithms could find," Mr. Sherson told the Associated Press. The game, Quantum Moves, is available online for the purpose of helping in the development of quantum computing. While it functions as entertainment, Quantum Moves is built to take quantum physics optimization problems and turn them into a game, the results of which demonstrate fundamental differences between human thought processes and the problem solving of computers.

Computers may have us beat at chess and checkers, but new research suggests our brains still have an edge when it comes to solving certain tricky problems thanks to a very human trait: intuition. Scientists in Denmark have found that people who played a game that simulated a complex calculation in physics sometimes did better than their silicon rivals. "The big surprise we had was that some of the players actually had solutions that were of high quality and of shorter duration than any computer algorithms could find," said Jacob Friis Sherson, a physicist at Aarhus University who co-wrote the study published Wednesday in the journal Nature. Experts say the results could advance the quest to develop effective quantum computers, something most major universities and several tech companies are working on as they seek to accelerate processing power. Such computers use individual atoms to store information and it's hoped they could one day outperform even the fastest conventional silicon-based supercomputers.

David Marcus, who heads Facebook's Messenger platform, tells Jefferson Graham how automated chat bots can improve the shopping and customer service experience, on #TalkingTech. SAN FRANCISCO -- Clicking a buy button on a website is so 20th century. Now the next big thing is conversing with a computerized robot. Used by 900 million people, Messenger has a major new look and mission, thanks to former PayPal chief David Marcus, who left the payment giant in 2014 to transform Facebook's hugely popular Messenger app. This week, Marcus unveiled the next big move for Messenger.

I was reflecting on the demonstrations of Deep Mind against Go champion Lee Se-dol, along with Watson at CeBIT and other artificial intelligence (AI) developments. It soon becomes apparent that we are evolving rapidly to a state where data learning through data analytics will be the battleground. In fact, it's clear that the battle over data is already won by those who have data architectures fit for AI. If you don't recognise this chart, please read the series of blogs I wrote in 2014 about such things. In this chart, I outlined the structure of a bank as front, middle, and back office (or retailer, processor, manufacturer), and that this structure is being attacked by technologies.

Ian Burkhart was 19 years old when he broke his neck diving into shallow water on holiday. Since then, he has been unable to move either of his legs, or his arms below the elbow (read Ian's story here). Now, in a world first, he has regained control of one hand and his fingers using a mind-reading device. In the past few years, we have seen paralysed people walk again with the aid of exoskeletons, and by using recorded brain activity to trigger electric stimulations to the leg muscles. Others have trained paralysed people to control computer cursors and robotic limbs by thought alone.

A recent academic study conducted by researchers from Leiden, Ben-Gurion and Toulouse Universities examined the performance of Cortical.io's Compared to the commonly used word-list method, Semantic Folding proved to have greater predictive power. Its other advantages were speed and ease of use. "Like the human brain, our Semantic Folding engine learns a language and understands the meaning of text by making analogies. Like the brain, it is both efficient and accurate. We are thrilled to see these compelling results confirmed by an independent academic study", comments Francisco Webber, inventor and co-founder of Cortical.io.

While video becomes more important as a digital media type, video data often remains dark to analytics. This webinar demonstrates how IBM Bluemix OpenWhisk together with Watson services begins to unlock the value of video data. Dark Vision is an application that uploads video files or streams to the cloud, transcodes video data, extracts and passes frames through the Watson Image Recognition and the Alchemy Face Recognition services, and generates meta-data to use in categorizing the video data for searchability. In the presentation, Andreas Nauerz introduces the basics of IBM Bluemix OpenWhisk and Frederic Lavigne demonstrates how to create OpenWhisk code to accomplish Dark Vision's application workflow. Maybe you want to do that because a particular team already has developed skills in this area, maybe because they do mobile development or have done mobile development in the past, and now want to outsource computer-intensive tasks.