Imagine the first day of your next job. You've never visited the office before or sat down with anyone from the company in person. In fact, the first person you meet face-to-face is the security guard, who hands you your building pass at reception. However, even though you've never been here before, you make your way to your desk, and see familiar faces and spaces. You wave hello to your manager, sit down at your desk, and begin work like you've been with the company for years.

Aon's Steve Petrevski discusses how Covid-19 has affected the analytics industry and what trends he expects to follow the pandemic. As the senior vice-president and general manager of Aon's data and analytics services, Steve Petrevski is an analytics expert who is responsible for driving growth strategy through new capabilities that leverage emerging technologies for digital businesses. "A key goal is developing Aon's data and analytic services platform, which brings together data, technology, expertise and solutions in a single place," he said. "These solutions include new data-driven products, digital distribution, marketplaces to match risk with capital and analytics-as-a-service capabilities." Petrevski has also recently worked on the integration of CoverWallet, a digital insurance platform for SMEs that Aon acquired at the start of this year to enhance the way it engages with small and medium-sized businesses. 'Product people, data scientists, digital marketing experts, data engineers and DevOps will all be in demand' – STEVE PETREVSKI Outside of his role at Aon, Petrevski also serves as an adviser to a number of start-ups in the areas of security, fraud and analytics.

Generalisation of a deep neural network (DNN) is one major concern when employing the deep learning approach for solving practical problems. In this paper we propose a new technique, named approximated orthonormal normalisation (AON), to improve the generalisation capacity of a DNN model. Considering a weight matrix W from a particular neural layer in the model, our objective is to design a function h(W) such that its row vectors are approximately orthogonal to each other while allowing the DNN model to fit the training data sufficiently accurate. By doing so, it would avoid co-adaptation among neurons of the same layer to be able to improve network-generalisation capacity. Specifically, at each iteration, we first approximate (WW^T)^(-1/2) using its Taylor expansion before multiplying the matrix W. After that, the matrix product is then normalised by applying the spectral normalisation (SN) technique to obtain h(W). Conceptually speaking, AON is designed to turn orthonormal regularisation into orthonormal normalisation to avoid manual balancing the original and penalty functions. Experimental results show that AON yields promising validation performance compared to orthonormal regularisation.

After years of promise, more and more businesses are realizing the benefits of artificial intelligence (AI) and machine learning. A 2018 survey showed that 47 percent of companies have integrated at least one AI capability into their business process – up from 20 percent in 2017 – and 71 percent plan to increase their AI investments in the coming years. Change is coming, but that doesn't mean a rise of the machines. Yes, it will mean job changes for some employees: 120 million workers in the world's 12 largest economies may have to be retrained over the next three years (2019–2021), but AI won't replace humans. Instead, industry experts say, AI can help humans make more efficient and transparent decisions.

Insurtechs with expertise in robotics, artificial intelligence, IoT, blockchain and advanced analytics are encouraging insurance companies to rethink product development through collaboration. The report, "Global insurance market opportunities: Reimagining risk management," argues insurtech startups will likely play the role of enabler for insurance innovation rather than disrupt long-standing business models. The report also notes 55% of the more than 550 existing insurtechs worldwide intend to help carriers improve the user experience and communications with policyholders, not compete against them. The growing segment has secured nearly $14 billion in venture capital funding so far, Aon says. "The insurance industry has been relatively slow to embrace digital technology compared with other industries. That reticence has opened a window of opportunity for entrepreneurs to deploy digital technology to improve the customer experience through a host of startup companies," Aon says.

The age of the robot has been predicted for decades. First used as a term to mean automated labor back in the 1920s, and popularised in the classic 1927 slient movie Metropolis, they have been a regular feature of science fiction ever since. The fact that robots and science fiction go hand in hand – and that predictions that we will soon have robot helpers being regular features of future-gazing since the 1930s – has meant that the idea of robots becoming a central part of our lives have become so familiar that we've come to ignore them. Yet robots have been a reality in manufacturing for decades, having become cost-effective production-line solutions by the 1970s. The Roomba – an automated vacuum cleaner that is perhaps the most famous domestic-helper robot – was launched back in 2002, and has sold more than 10 million units worldwide.