The AI Centre of Excellence is a global initiative, established by Think Creative Studios in 2014. It was pioneered to drive a number of key Deep Learning initiatives and to leverage the best minds internationally available to achieve rapid, high quality and astounding results. The AI CofE has thus built productive alliances within Industry, Government and Academia. Its central purpose is to train AI Domains - digital minds providing world class classification and prediction in a specific expert field- and thereby to demonstrate the transformational impact of Artificial Intelligence on society and industry. Projects are managed using the AI GO ^TM Delivery Methodology devised and proven by Think Creative Studios. Developed and refined overs 10 years, this methodology combines the learnings, steps, techniques and milestones requisite to maximise speed of results for AI build, adoption and utilisation. It  ensures the optimum selection of projects, the quality of contributions, and guarantees deliverables and (most important of all) outcomes.

Artificial Intelligence, at this point, is essentially the creation of machine brains. These artificial neural networks - modelled on our best approximation of the way the human brain is structured and functions –  are able to learn just as we do. Such Artificial Neural Networks are not new. Cognitive Scientists conceptualised ANNs in the 1950s, and were hand modelling them in the 1970s. The potential of ANNs to transform society and industry however at the current scale is only possible post-2010 due to the recent convergence of technological innovations. Specifically, this includes the massive increase in computing power, the transformation of parallel processing through GPUs, and the ever increasing persistence of massive data sets. This provides the capability, performance and the information required to train machines to perform a wide range of tasks with the accuracy and reliability that aligns with the most talented human. An ANN however can be scaled to be made available to vast groups simultaneously, 24*7, 365 days a year, whereas the small number of eminent specialists simply cannot. This is not to suggest that ANNs are suited to every task which humans currently perform. They will however begin to multiply in tasks where classification or prediction can be perfected beyond human ability through the consumption and analysis of vast datasets. In this role, AI will save lives, increase health and longevity, and either supplement or replace human decision makers in a wide variety of specialisms.

A revolution is not without risks, as history eloquently warns us. The risks of AI are threefold. Firstly, the transformation of the workforce, including making a range of traditional roles redundant, will create significant tensions between corporate profit and social good. Secondly, AI relies upon information. The greater the information available, the more extensive the uses, and the more accurate the results. Organisations will therefore continue to increase their access and retention of their customer’s data, and this will either become mandatory by default, or be implemented by stealth. This does, and will increasingly pose a threat to the privacy of the individual, and the benefits should be weighed at the State level vs the risks, and the inherent rights of humankind. This noted, the rate of technological advance, and the vested interest within industry to possess data, have resulted in the legislature being left well behind in the sphere of information and data privacy. Finally, with the ever improving ability of the machine to think on behalf of the individual, and even to mimic human relationship, the autonomous, logical, creative and empathatic self becomes subject to forces which can deconstruct and perhaps overwrite it.  All three areas above require a deep philosophical engagement in order to define the ethical and normative framework within which AI should be utilised.  At this point the investment in this domain is insufficient in relation to the complexity and risks implicit.