The first steps to AI
Artificial Intelligence (AI) is splitting opinions around the world. Signaling the start of Industry 4.0, could AI be a force for good, set to increase the UK economy by 10% over the next ten years as projected by PwC, or the start of a major downturn for a range of physical and knowledgeable workers? Given the comprehensive list of technologies that appear to fall within the definition of AI, it is difficult to affirm any projections at this point.
AI is not a simple one-step solution that will completely remove manual tasks or overhaul day-to-day operations – it is one more step towards harnessing large volumes of data to better comprehend the business and therefore encourage new efficiencies. However, this is where the problem lies. Businesses are continuing to struggle to achieve insight from current data sources – where are the reliable and efficient data gathering systems that allow organisations to review insight rapidly and effectively? Where is the inherent trust in data? Businesses must take significant strides in trust and confidence in data before any of these innovative technologies – from machine learning onwards – can obtain any practical foothold within the day-to-day corporate business operations.
Chairman at Zizo, Peter Ruffley, reviews the significance of Digital Transformation (DX) and realising a data-driven culture to ensure that businesses can optimise within the world of AI..
The Promise of AI
AI has rapidly infiltrated into everyday life and has taken some by surprise. While we are still some way off the realisation of autonomous vehicles eliminating road traffic accidents; robots utilising X-Rays, MRI scans and medical research to transform diagnostics; even robot surgeons, innovations in machine learning, speed and visual recognition technology are already infiltrating AI in day-to-day life.
From supply chains using the Internet of Things (IoT) to reduce food wastage to the translation engines revolutionising worldwide communications, the notion of intelligent automation is becoming increasingly popular. If fulfilled, the promise of AI will completely alter each aspect of daily life.
But, while some have described a future dominated by machines, a world where 150 million knowledgeable workers will have their roles replaced by cognitive robots, at this point in time, most businesses have no clear comprehension of, or strategy for, effectively making the most of AI in the future. And in fact, they are not even close.
So, how can a business prepare for the intuition led processes that can and will be fuelled by AI? The essential change will be accomplishing a cultural willingness to trust and believe the data. The principle of AI is that technology is trusted to do get the job done, based on the data at hand. If the information provided is partial or incorrect, the AI cannot perform effectively.
Undoubtedly, if the AI is given a restricted subset of vital patient data, any complicated diagnostic processes will be unconvincing to say the least. ilots will be hesitant to depart in a aircraft that is using an AI-based predictive maintenance system if they are not completely confident in the quality of the data provided.
The evolution from where we are today to completely trusting in AI will be significant. When a business today does not fully trust in its data resources to even make data driven decisions, it is unlikely to fully embrace AI, a model founded on data driven activity. For organisations that are not yet making decisions founded on trusted information are going to find it increasingly difficult to fully encompass AI.
Digital Transformation (DX)
Businesses that are able to follow a digital transformation process and fully take on data driven evaluations will be perfectly positioned to discover the array of opportunities that AI can deliver. From regression algorithms and predictive models, that can precisely forecast shopping patterns and allow retailers to place the appropriate store and security personnel on site; to completely automated tills, discounted prices automatically produced based on stock availability, even web analytics and product affinity, AI will totally alter every level of the retail model. Trials have already begun in stores without the need for checkouts, instead using sensors and cameras to monitor customer selections and automatically debiting accounts via Smartphone applications.
However, the comparison to today’s operations is evident this is not a simple transition. Just contemplate the sheer volume of data that is not currently being captured and applied effectively. Supply chain optimisation in real time, responding to customer demand should be a given bearing in mind the quantity of EPOS data and inventory information. Yet many grocers continue to trust in human individuals to constantly examine the shelves throughout the day in every store to recognise any breaks in stock availability. A report is then generated, dispatched to the head office and, finally, the stock arrives in store.
Nevertheless the information exists. So why are these retailers not using the data collected from sales patterns across every store to forecast regional demand? Why are they not considering seasonality, demand variation by day of week and resupplying accordingly? Enhanced insight can revolutionise the precision of stock quantities, minimise stock outs and increase the overall customer experience. However, without a willingness to move to the next step – specifically recording availability of stock against sales and fully embrace data driven decision making – the retailer is far from realising a fully automated business model by being left with an inefficient and inaccurate manual method.
Looking too far into the future is where the mistake lies. There’s no doubt that AI is on the way, but there are a various steps that businesses must take first in order to get closer to AI in practice. And the effective application of current data sources is the first step in which to take.
At present, regardless of the board level dedication to DX processes, how many businesses can really state that they are truly data driven? To have a user base with complete trust in the information and an ethos that has real confidence in data led decisions? Without that fundamental foundation, the extent of using any type of AI or predictive or machine learning will be severely minimised. AI may signal the beginning of the fourth industrial revolution; but effective rollout of AI will require vital cultural change and a completely altered attitude to data.