Artificial Intelligence has permeated almost every industry, either in word or deed, in the last couple of years. From financial institutions to ride-hailing services such as Uber, companies are clambering over one another to take advantage of this technology to stay ahead of the competition. However, one area which Artificial Intelligence has been unable to find a platform in, until very recently, has been the legal sector. There is a belief that the legal sector, particularly when it comes to the courtroom environment, is reserved exclusively for sharp-suited lawyers, who have trained for years to be able to build and present a case in order to persuade a jury of their peers of the validity of their argument. However, the tide might now be turning.

It's said that there was a time when a well-educated individual could be an authority on all the most important areas of science, art and culture. Although that period, if it ever existed, is long passed, you can sometimes believe it's possible to re-enter it by listening, almost indiscriminately, to Radio 4. Did you know, for example, that the word "robot" comes from an old Slavonic word for slave or actually, more accurately, "forced labour"? That was one of the memorable facts recounted by Adam Rutherford in The Rise of the Robots. It was the Czech playwright Karel Čapek who introduced its modern meaning nearly 100 years ago in his play RUR (Rossum's Universal Robots). But as Rutherford noted, these "immigrants from the future" have been haunting the human imagination for much longer than that.

With Brexit underway and Trump in the White House, 2016 was not a good year for forecasters. We will try to buck the trend by predicting which areas of InsurTech will have an impact in 2017. Futurologists have predicted mass adoption of self-driving, or autonomous, vehicles since cars were first invented. We are finally a lot closer to this vision; for example Google and Uber are running trials on public roads. The consensus view is that consumer-ready autonomous vehicles are 3 years away and ubiquity is another 10-15 years down the road (pun intended).

I remember sitting in a conference audience in the late 1990s during the fat part of the first dot-com expansion curve, when everyone was complaining that the Internet was irrationally overhyped. Then, pre-Google Eric Schmidt took the stage and told us that, "I actually think the Internet is underhyped." As a tech journalist in those days, I'd had the privilege of long talks with Schmidt and hadn't wasted the opportunity to learn. Other people laughed, but I knew he was serious -- and he was right. The point is, I've begun to get the sense that most marketers aren't yet taking AI seriously enough.

Machine ethics offers an alternative solution for artificial intelligence (AI) safety governance. In order to mitigate risks in human-robot interactions, robots will have to comply with humanity's ethical and legal norms, once they've merged into our daily life with highly autonomous capability. In terms of technical challenges, there are still many open questions in machine ethics. For example, what is deontic logic and how can it be used for improving AI safety? How do we fashion the knowledge representation for ethical robots? These are all significant questions for us to investigate. In this interview, we invite Prof. Ronald C. Arkin to share his insights on robot ethics, with a focus on its technical aspects.

Autonomous weapons are being increasingly sought my militaries around the world, but experts fear the worst. AUTONOMOUS robots with the ability to make life or death decisions and snuff out the enemy could very soon be a common feature of warfare, as a new-age arms race between world powers heats up. Harnessing artificial intelligence -- and weaponising it for the battlefield and to gain advantage in cyber warfare -- has the US, Chinese, Russian and other governments furiously working away to gain the edge over their global counterparts. But researchers warn of the incredible dangers involved and the "terrifying future" we risk courting. "The arms race is already starting," said Professor Toby Walsh from UNSW's School of Computer Science and Engineering.

Machine learning is at its best when there's way too much information for any human to comb through manually, like making high-volume stock trades or surfacing the best posts from hundreds of friends on Facebook. Now one Estonia-based startup, Teleport, is using this idea, coupled with images from Google Street View, to automatically look around cities and see if people will like them based on their lifestyle preferences. In a Medium post, Teleport co-founder Silver Keskkula walks through an example of the process. First, he plots 10,000 randomized points throughout a city, and grabs images taken by Google Street View. Then those images are run through computer-vision algorithms that identify objects, people, and buildings, and describes them in a short sentence.

Online learning with streaming data in a distributed and collaborative manner can be useful in a wide range of applications. This topic has been receiving considerable attention in recent years with emphasis on both single-task and multitask scenarios. In single-task adaptation, agents cooperate to track an objective of common interest, while in multitask adaptation agents track multiple objectives simultaneously. Regularization is one useful technique to promote and exploit similarity among tasks in the latter scenario. This work examines an alternative way to model relations among tasks by assuming that they all share a common latent feature representation. As a result, a new multitask learning formulation is presented and algorithms are developed for its solution in a distributed online manner. We present a unified framework to analyze the mean-square-error performance of the adaptive strategies, and conduct simulations to illustrate the theoretical findings and potential applications.

On the Web, there is always a need to aggregate opinions from the crowd (as in posts, social networks, forums, etc.). Different mechanisms have been implemented to capture these opinions such as "Like" in Facebook, "Favorite" in Twitter, thumbs-up/down, flagging, and so on. However, in more contested domains (e.g. Wikipedia, political discussion, and climate change discussion) these mechanisms are not sufficient since they only deal with each issue independently without considering the relationships between different claims. We can view a set of conflicting arguments as a graph in which the nodes represent arguments and the arcs between these nodes represent the defeat relation. A group of people can then collectively evaluate such graphs. To do this, the group must use a rule to aggregate their individual opinions about the entire argument graph. Here, we present the first experimental evaluation of different principles commonly employed by aggregation rules presented in the literature. We use randomized controlled experiments to investigate which principles people consider better at aggregating opinions under different conditions. Our analysis reveals a number of factors, not captured by traditional formal models, that play an important role in determining the efficacy of aggregation. These results help bring formal models of argumentation closer to real-world application.

Students code software at desktops, while others assemble odd machines with wires and multi-colored boxes. Earning a spot at this elite university isn't easy; UC-Berkeley accepted a mere 14.8 percent of applicants for the class of 2020. So this young crew will likely be tomorrow's tech leaders and pioneers. Despite all the promise, it appears that BRETT is struggling. BRETT is a robot, and he – or she, or it – is attempting to place a small wooden block into a small hole. Again and again, BRETT swings his arm over the opening, attempts to place the block, but fumbles. Just can't make it fit. However, as robots go, BRETT has a huge advantage: he can learn. Every time BRETT swings his arm and fails, he calculates what went wrong. In essence he's doing what we humans do: he's failing, and in response he's deciding how to improve the next effort. I stand watching for about 15 minutes, and finally BRETT succeeds – a lengthy period given the simple task. But the astounding point is that the robot really did learn.