AI For Business: How Are Companies Making Data Work For Them?
Every news cycle seems to unearth a new buzzy story about the transformative powers of Artificial Intelligence. It was reported recently that scientists had discovered a new type of antibiotic using a powerful algorithm, an achievement heralded as a breakthrough in the fight against the growing problem of drug resistance. Industries ranging from journalism to retail, HR and aerospace have all seen fascinating examples of AI technology change the ways they do business in recent times. But how do these eye-catching stories translate to real businesses adapting to AI on the ground?
AI is a huge opportunity for businesses to capitalize on the relatively recent and massive proliferation of data. According to McKinsey, AI has the potential to deliver additional global economic activity of around $13 trillion by 2030, or about 16% higher cumulative GDP compared with today. AI and machine learning are examples of tools that businesses can utilize to make sense of what is known as Big Data (data sets too large or complex to be dealt with by traditional data-processing software). But many companies, particularly small and medium-sized ones, have faced challenges adapting to the technology wholesale. These include a deficiency in technical skills required to utilize AI, a lack of understanding of its benefits or uses, or difficulty in defining a strategy and finding an appropriate use-case for the technology. It’s vital to have a clear picture of what you want to achieve with an AI project; much like a jigsaw puzzle, you need to know what the result is supposed to look like before you start putting pieces together.
Once these barriers have been overcome, however, there is a virtually endless variety of applications for the technology. AI can improve automation, augmentation and scaling of many human-like processes. These can include:
Automating expert decisions: using algorithms to calculate optimal maintenance schedules in a factory, for example. Ticket resale company StubHub reported a 90% reduction in online fraud after implementing a predictive analytics and statistical analysis package.
Automating unskilled work: e.g. handwriting recognition to speed-up form-processing.
Relieving skilled workers of unskilled tasks, freeing them to perform their core duties. This has implications in all sorts of industries, from cutting down on the amount of paperwork that nurses have to complete, allowing them more time with patients, to artists automating repetitive, time-consuming parts of the creative process.
Machine learning tools can analyze large datasets to identify patterns or make predictions. Examples include Natural Language Processing applications that can analyze social media posts to see what customers are saying about a brand or tailored sales promotions for customers. Spanish financial services company CaixaBank recently reported a 50% increase in ‘real sells’ through personalized, targeted messages.
A healthcare agency, for example, can use data from claims and biometric measures to produce machine learning models that accurately predict the likelihood of patients developing a disease. The model could also predict interventions most likely to improve the patients’ health outlook.