Artificial Intelligence (AI) is probably the most widely misconstrued tech term ever coined. Its long-term association with science fiction has positioned it in the minds of many people as something sinister; malevolent, almost. And even when we know its real worth there's still a reluctance amongst businesses of all sizes to adopt a set of tools that promise so much. So what's going wrong?
AI has long been hailed as a revolutionary technology, capable of changing the way we work, but very few enterprises have been willing or able to take full advantage of this potential. Recent research shows that just over a quarter of IT professionals have actually implemented the technology and there are a few key reasons for this hesitancy. The apparent complexity of integrating AI into existing business systems is seemingly an issue, as is a lack of expertise or knowledge.
The doubters and disbelievers are in for a shock, though. According to a 2020 Forrester Consulting study, organizations that adopt and scale AI are seven times more likely to be the fastest-growing businesses in their industry. That one fact alone should convince even the most ardent sceptic to put aside their doubts – the bottom line is that your bottom line is in the firing line. If your business doesn't adopt AI it will lose out to those that are.
Lufthansa, for example, recognized early on that with the right data and AI strategy it could improve customer services, empower employees and improve operational efficiency. The airline started by developing a computer platform for its data scientists to experiment and test AI projects prior to rolling them out around the company.
For any technology to gain widespread acceptance there first needs to be trust and for AI this can be a stumbling block. The idea of machine intelligence making decisions that can take precedence over human insights has the potential to be destabilizing. So how can companies build confidence and overcome any lingering reservations?
First, the vendors of AI tools need to ensure that their systems deliver decisions and recommendations that are explicable, comprehendible and fully traceable. There should also be a focus on privacy, safety and fairness, supported by a framework that ensures a human review of every AI decision and action. There's a potential pitfall here, though; the introduction of involuntary human bias, which has the possibility to inadvertently influence outcomes.
That's why we advise companies to pick an "ethics official," a person responsible for overseeing the day-to-day governance of each AI implementation and for communicating its ongoing objectives. Building trust and then maintaining it are the foundations for building a successful AI implementation.
However, to fully leverage that implementation requires more than confidence; businesses of all sizes also need access to a range of specialist skills – expertise that is currently in short supply. Finding deep learning, natural language processing and robotic process automation specialists is a real issue but without access to these sorts of competencies companies risk not optimizing the benefits AI can deliver.
Unfortunately, there is no quick and easy solution to the AI skills gap, although a sensible first step is to train staff in data and AI applications. Intriguingly, AI can also help to break the logjam by automating mundane tasks, freeing up staff to retrain and reorientate their careers. In some circumstances AI might prove useful for streamlining the process of hiring staff, too.
The final hurdle is bringing it all together while avoiding the pitfalls. That's why enterprises should choose a technology partner with a range of tools and approaches for AI implementation specifically designed to help even the most hesitant business bridge the AI gap. This can help lower the barriers to entry and make AI more accessible in practical, powerful and performance-enhancing ways. Think of it as a ladder, if you like, that's expressly intended to help businesses scale the hurdles between your business and AI adoption.
You can visualize this ladder as having four rungs: Collect, Organize, Analyse and Infuse. Each step helps create a painless progression towards the deployment of a scalable, secure and transparent AI implementation that will unleash the technology's advanced analytical power, the next gen driver of business growth.
AI is fast becoming an integral part of modern business operations. Companies that fail to understand that will lose ground to their competition: it's already starting to happen. Perhaps it's time to reach for the AI ladder before that gap becomes an unbridgeable gulf.
(Daniel de la Fuente is vice president of Data, AI and Automation at IBM EMEA)
This article first appeared in IBTimes.com. Read it here.