Data is already a problem for many businesses. Not just in terms of managing the vast amounts already generated alongside the new data arising from analytics and IoT, but understanding what to do with it, and learning how to convert this rich resource into valuable insight that adds to bottom line profitability. IDC is forecasting global data to reach 44 zettabytes of data by 2020; just five years ago it was 4.4 zettabytes. McKinsey in turn highlights how the majority of firms are generating just a small fraction of the potential their data assets can return, focusing on production and top-level insights, then discarding or archiving the remainder. Could AI be the key to unlocking this lost business potential, and what’s the best approach to implement this disruptive technology?
The evidence tells us organisations are already investing, or planning to invest, in AI and cognitive systems to drive returns from their information assets. IDC predicts compound annual growth of 54.4% for AI through to 2020. Drivers for this include AI’s ability to identify patterns in the data that wouldn’t be seen otherwise. These can lead to faster reactions and resolutions to any issues arising, and to the automation of tasks in areas such as production and customer service.
When it comes to developing a strategy for AI, many organisations will be tempted to build around the data they already have. However the optimal approach is to set an agenda based on potential use cases and desired outcomes. Here, organisations shouldn’t be afraid to seek advice from more experienced partners and experiment to see where AI can be the right fit.
Once these objectives are set, the organisation can develop a plan for its existing and new data needs, implement a pilot or proof-of-concept exercise to validate the use case, then identify and recruit the specific skills and resources necessary for execution, in many instances bringing in temporary third-party expertise already available from specialist partners. Ultimately many IT teams and business leaders will scale up their approach, becoming AI led by placing it at the centre of their thinking for product development, logistics, marketing and may other core functions.
We’re still very early in the AI lifecycle
For early adopters, one area where AI is already a reality is Prescriptive Analytics. Organisations are replacing traditional reactive approaches to issues such as machinery refresh and customer service development, with a prescriptive strategy that sifts through real-time data generated by equipment, sensors and applications to identify issues before they impact, then automating the response. The data in play comes from both structured and unstructured sources: from Excel spreadsheets to online geodata, cloud and SQL databases.
AI is creeping inexorably into our lives, from Amazon’s Alexa in the home, to Libratus – the almost celebrity poker playing AI that took on the world’s best and won comprehensively – to a plethora of warnings that AI will threaten humanity, AI is on the agenda for business and consumers alike.
On the horizon is HPE’s revolutionary reinvention of computing, The Machine. Built specifically to addresses the problem of being data rich but insight poor, this HPE initiative is exploring memory driven computing as a means to provide the massive compute capabilities AI will need to conquer Big Data.
Right now HPE is offering organisations the first steps to AI in the enterprise, in the shape of its Deep Learning initiative centered on Apollo systems and HPE Pointnext services, and also in its HPE Nimble and InfoSight storage offering. InfoSight brings machine learning based AI right into today’s datacentre, by analyzing not just all the arrays and compute in a single enterprise environment, but all the arrays and compute HPE has deployed, to look for patterns and trends in real-time and convert these into insight and automated measures that almost fully prevent storage downtime. In the grand scheme of things this might be a small step, but it’s definitely the shape of things to come for computing, AI and Big Data.