SMEs can challenge big beasts by snapping up AI benefits first
The debate over the effect of artificial intelligence and automation is not going to cool down any time soon. AI is a disruptive technology: because of it we face profound changes to our lives, our work and even, some say, our very existence. While some of the predictions are hype, there are huge implications for companies.
Historically, SMEs tend to hold back in the face of new technologies. First-mover advantage is usually seized by new entrants or those with the scale and resources to fully take advantage of emerging opportunities. The companies in the middle are left behind.
With AI, the rules are different, and holding back is a dangerous approach. Since the technologies are so accessible and the benefits so great, there is no reason why SMEs should not welcome these advantages with open arms. The winners in the automation race will be the fastest, nimblest and hungriest, not the big beasts.
There are three key benefits that SMEs can secure from AI and automation.
First, the potential for a low-cost, accessible way to get insight from data faster, enabling businesses to make better decisions. This is happening in many industries already — in agriculture, for example, machine-learning techniques help to profile soil fertility and then predict, based on weather and climate, when and where to fertilise a crop.
Second, in retail markets, predictive algorithms are being used to anticipate what a customer may buy next. The ability to anticipate customer behaviour is improving all the time.
Third, AI provides ways to improve and deepen relationships with customers through greater personalisation without having to employ large numbers of staff to provide a personal touch.
Companies can benefit from the growing use of robotic process automation, where AI-enabled systems take on complex but repetitive processes. For example, at PA Consulting Group we have worked with a life sciences company that has used robotic process to speed the development of drug formulations. In one case, the effectiveness of the drug was measured by counting cells in a biopsy, identifying whether or not each cell was affected by the disease. This laborious and repetitive task had previously required the expertise of a laboratory scientist.
By using a pattern-matching, machine-learning algorithm, we enabled AI to support the scientist by recommending which cells were affected and which were not. The more it was used, the more accurate the algorithm became. The time taken on the task was shortened considerably and accuracy improved.
We call this approach “assisted intelligence” because it does not replace the expertise of a particular person but enables them to do their job faster, easier and more effectively.
Tasks that require information from multiple sources to make timely decisions are great candidates for experimenting with AI
The SMEs that will benefit most from AI are doing three things right. They are engaging their workforce — training people and giving them the experiences and skills that will complement the capabilities that AI can bring. This will enable companies to use AI to support their workforce rather than replace it.
They are also identifying individual employees’ specific skills and experience, finding the people who know the short-cuts to getting things done and working out what they need to do to pool this knowledge, and enable computers to learn from it.
They are experimenting and following a “think big and start small” approach. They are identifying tasks in the business that demand skill and experience but are repetitive and data-intensive. Tasks that require information from multiple sources to make timely decisions are great candidates for experimenting with AI.
Once the SME knows the identity of the good candidates for automation, experiments with AI can yield results quickly.
The SMEs using AI are also being bold. The advantage SMEs have is that vast amounts of this technology are in the public domain and even the most powerful of AI systems are often underpinned by open-source components. This means that implementing AI projects does not have to be a capital- or human resource-intensive activity.
This shows clearly, that with the application of skill, planning and vision, SMEs can steal the advantages of AI from their bigger rivals and beat the traditional winners in the technology race.
The author is chief innovation officer at PA Consulting Group