"Our analysis suggests that up to 30% of UK jobs could potentially be at high risk of automation by the early 2030s, lower than the US (38%) or Germany (35%), but higher than Japan (21%)." They join the many such reports from academics, analysts and authors that filled an entire chapter titled "Sum of all Fears" in my new book, Silicon Collar. "The reality, however, is that when Oxford, MIT, McKinsey, and Gartner talk, the person on the street and even business executives typically just read the headlines, and when all of these big brands agree on something, it solidifies readers' overall impression--in this case, pessimism." The PwC analysis suffers from the same "lab rat" mindset of so many such studies. This comprised US data from the Programme for the International Assessment of Adult Competencies (PIAAC) database, merged with automatability data from FO.

Artificial intelligence will be the main way that banks interact with their customers within the next three years, a report from consultancy Accenture has suggested. Banks such as Royal Bank of Scotland (RBS) are increasingly using chatbots to answer customer queries. The report examined the views of 600 bankers and other experts. Many, perhaps ironically, felt that AI would help banks create a more human-like customer experience. "The big paradox here is that people think technology will lead to banking becoming more and more automated and less and less personalised, but what we've seen coming through here is the view that technology will actually help banking become a lot more personalised," said Alan McIntyre, head of the Accenture's banking practice and co-author of the report, Banking Technology Vision 2017.

Swedish transportation company Einride announced its global launch with news that it's working to establish the world's first totally emission-free road transportation system. Their aim: To knock out the current semi truck-based supply chain. Einride's emission-free model is based on an entirely new vehicle class, which it calls the "T-pod," a fully electric self-driving vehicle that can also be controlled by remote drivers. The "T-pod" is around seven meters (23 feet) long, with a cargo capacity of about 15 standard pallets and a weight of 20 tons when carrying a full load. Einride said it should be able to travel 200 km (124 miles) on one charge, which doesn't sound like much -- but the company thinks the system's efficiency and low demand for operators will make up for the loss of range.

Consciousness is a thriving industry. Consciousness is a buzzing business in neuroscience labs and brain institutes. Just a few decades ago, consciousness barely registered as a credible subject for science. Perhaps no one did more to legitimize its study than Francis Crick, who launched a second career in neurobiology after cracking the genetic code. In the 1980s Crick found a brilliant collaborator in the young scientist Christof Koch. In some ways, they made an unlikely team. Crick, a legend in science, was an outspoken atheist, while Koch, 40 years younger, was a Catholic yearning for ultimate meaning. Together, they published a series of pioneering articles on the neural correlates of consciousness until Crick died in 2004. Koch went on to a distinguished career at Caltech before joining the Allen Institute for Brain Science in Seattle. Today, as the president and chief scientific officer, he supervises several hundred scientists, engineers, and informatics experts trying to map the brain and figure out how our neural circuits process information. The Institute recently made news with the discovery of three giant neurons connecting many regions of the mouse brain, including one that wraps around the entire brain. The neurons extend from a set of cells known as the claustrum, which Crick and Koch maintained could act as a seat of consciousness. Koch is one of the great thinkers about consciousness. He has a philosophical frame of mind and jumps readily from one big idea to the next.

Scientists are getting closer to building life from scratch and technology pioneers are taking notice, with record sums moving into a field that could deliver novel drugs, materials, chemicals and even perfumes. Despite ethical and safety concerns, investors are attracted by synthetic biology's wide market potential and the plummeting cost of DNA synthesis, which is industrialising the writing of the genetic code that determines how organisms function. While existing biotechnology is already used to make medicines like insulin and genetically modified crops, synthesising whole genes or genomes gives an opportunity for far more extensive changes. Alexander the friendly robot visits the Indoor Park in London (file pic). Technology pioneers are investing huge sums of money in all manner of innovative ideas.

"Almost one-third of UK jobs could be done by robots in 10 years' time." Warnings of this kind are often softened by the suggestion that new and better occupations will replace defunct jobs. But such is the pace of technological advancement, this is no longer guaranteed. This email address is already registered. By submitting my Email address I confirm that I have read and accepted the Terms of Use and Declaration of Consent.

Given a set of items of unknown utility, we need to select one with a utility as high as possible ("the selection problem"). Measurements (possibly noisy) of item values prior to selection are allowed, at a known cost. The goal is to optimize the overall sequential decision process of measurements and selection. Value of information (VOI) is a well-known scheme for selecting measurements, but the intractability of the problem typically leads to using myopic VOI estimates. Other schemes have also been proposed, some with approximation guarantees, based on submodularity criteria. However, it was observed that the VOI is not submodular in general. In this paper we examine theoretical properties of VOI for the selection problem, and identify cases of submodularity and supermodularity. We suggest how to use these properties to compute approximately optimal measurement batch policies, with an example based on a "wine selection problem".

An inconsistent knowledge base can be abstracted as a set of arguments and a defeat relation among them. There can be more than one consistent way to evaluate such an argumentation graph. Collective argument evaluation is the problem of aggregating the opinions of multiple agents on how a given set of arguments should be evaluated. It is crucial not only to ensure that the outcome is logically consistent, but also satisfies measures of social optimality and immunity to strategic manipulation. This is because agents have their individual preferences about what the outcome ought to be. In the current paper, we analyze three previously introduced argument-based aggregation operators with respect to Pareto optimality and strategy proofness under different general classes of agent preferences. We highlight fundamental trade-offs between strategic manipulability and social optimality on one hand, and classical logical criteria on the other. Our results motivate further investigation into the relationship between social choice and argumentation theory. The results are also relevant for choosing an appropriate aggregation operator given the criteria that are considered more important, as well as the nature of agents' preferences.

Policy gradient is an efficient technique for improving a policy in a reinforcement learning setting. However, vanilla online variants are on-policy only and not able to take advantage of off-policy data. In this paper we describe a new technique that combines policy gradient with off-policy Q-learning, drawing experience from a replay buffer. This is motivated by making a connection between the fixed points of the regularized policy gradient algorithm and the Q-values. This connection allows us to estimate the Q-values from the action preferences of the policy, to which we apply Q-learning updates. We refer to the new technique as 'PGQL', for policy gradient and Q-learning. We also establish an equivalency between action-value fitting techniques and actor-critic algorithms, showing that regularized policy gradient techniques can be interpreted as advantage function learning algorithms. We conclude with some numerical examples that demonstrate improved data efficiency and stability of PGQL. In particular, we tested PGQL on the full suite of Atari games and achieved performance exceeding that of both asynchronous advantage actor-critic (A3C) and Q-learning.

A quarter of voice technology users say they've had a sexual fantasy about their voice assistant, a new study has revealed. Even more users, 37 per cent, say that they love their voice assistant so much that they wish it were a real person. The findings could have an impact on how brands such as Google and Amazon design their voice technologies for human interaction. Pictured is Theodore Twombly, played by Joaquin Phoenix in the movie Her. The Wall Street Journal (WSJ) has recently cited two studies that investigate these allegations, which have discovered that'both women and men find the female voice welcoming and warm,' reports Joanna Stern with WSJ.