Over 50 percent of businesses believe artificial intelligence (AI) will play a major role in their future projects, according to a recent O'Reilly survey. Yet despite such intentions, the research also revealed that 71 percent of respondents have not yet started with implementation. A McKinsey report found a similar disparity: while companies are optimistic about AI, two-thirds of respondents are yet to adopt any such technologies.

This contrast between expectation and implementation is not overly surprising. Any disruptive technology takes time to move from hype to reality. We have seen the same countless times in the past including with technologies as transformative as cloud computing. This same dynamic is playing out with AI, and while early adopters are pressing ahead in industries like healthcare, automotive, retail and financial services, many others are just beginning to experiment, or are simply watching and waiting at this point.

The impetus behind AI adoption will only grow as business leaders become increasingly aware of the value it can deliver to their company. McKinsey’s study revealed that 45 percent of executives who have yet to invest in AI fear falling behind competitively. As a result, the report said, “companies are twice as likely to embrace AI as they were to adopt new technologies in past technology cycles.”

Clearly this suggests an impasse of sorts – companies know they need to get involved with AI, but are struggling to understand how best to go about it. But with Gartner predicting the technology to generate $3.9 trillion in business value by 2022, the time for contemplating the ‘what?’, 'why?’ and ‘how?’ of AI is running out. Instead, here are the key steps organizations can take to begin taking advantage of it now.

ai machine learning thinkstock photos monsitj 645606376
– Thinkstock / monsitj

Do the things you would do with any new technology

Demonstrate AI’s business value through a low-risk proof-of-concept project. Such pilots, while typically led by senior or middle management within the IT department, are essential in achieving C-level buy-in for greater investment and a wider roll-out. So make sure someone with decision-making capacity has visibility.

Understand the software development lifecycle (SDLC) for AI applications – planning, systems analysis, systems design, development, testing, implementation and maintenance. As with any application, there needs to be a standardized, efficient and agile method which assures high-quality applications meeting defined objectives.

Similarly, understand the product development lifecycle (PDLC) for AI. This includes requirements, design, manufacturing for hardware and development for software, testing, distribution, use and maintenance, and disposal.

Stay in tune with the latest trends and key technologies. AI is moving at a rapid rate, so it’s crucial to maintain deep knowledge to remain ahead of the game.

Stay focused on the goal

It’s essential to always keep sight of the main objective that your new AI application is designed to address. AI for AI’s sake makes no sense – instead consider what the problems are that you’re trying to solve. It could be process automation, training datasets to detect fraud, or any number of things. Identify this, always keep it front of mind, and work backwards from there.

Identify the real people you need to bring AI to life

The data engineer, the data scientist, the machine learning engineer, the DevOps engineer – all of these and many others can be key people in turning AI from ambition to reality. Ensure you identify these people within the business – or bring in the talent from outside if needed – and, more importantly, give them the tools they need to do their jobs.

Avoid common mistakes

A number of pitfalls can derail an AI project’s success. These include: not mining your data sufficiently to know if it’s trustworthy and where there are anomalies; inadequate staff training; not understanding the models used and the underlying probabilities and their impact on the solution; lacking knowledge of how to evolve the models to improve underlying probabilities; and using machine learning where it is simply not needed.

As long as organizations know what to do – and what not to do – they can begin their AI journeys on firm ground. And fortunately, there’s also no need to fly solo, with a number of software providers and consultants standing ready to assist, from concept to creation and beyond.