Not too long ago, it was the stuff of scary science fiction, but today artificial intelligence (AI) is fast becoming a useful and accessible technology. It's still in its early stages, so larger organizations with significant IT budgets have been the early adopters. But affordable "pre-trained" generic services that use third-party data and are suitable for small and medium-sized businesses are already emerging.

AI in plain English

"Artificial intelligence is about computers behaving in a way that we would call intelligent in humans. Today, machine learning is the leading route towards that goal," says Markus Noga, head of machine learning at SAP. Machine learning refers to the use of algorithms that can learn from experience without having to be explicitly programmed by human beings. Experience, for a computer, comes from data.

The availability of Big Data, improved processing power and better algorithms are the foundation of AI. They are the reasons why machine learning will take enterprise software to a new level, Noga says.

Deep learning leads to better performance

Deep learning is another term for the core process behind AI. The AI algorithm changes as it is exposed to more and more data, and its performance improves continuously. "These algorithms have the ability to alter themselves and how they work," explains Chris Nicholson, CEO of Skymind, the developer of a widely used deep learning tool for Java. "You show them data, and they learn how to classify or group or make predictions about that data."

The arc of continuous improvement is important. It results in better outcomes and higher accuracy, which expands the range of tasks AI can perform reliably. For example, some types of sensory data, such as raw images and raw sound, are very disorganized and organic. But AI has advanced to the point where it can analyze such data and make predictions and decisions about it, Nicholson says.

Capabilities will grow as interconnectedness increases

AI shows particular promise in areas such as supply chain management, says Gary Neights, product manager at Elemica, a supply chain operating network. Its usefulness will increase as businesses' IT networks become more interconnected and provide access to a wider breadth of supply chain data. That same dynamic will drive AI solutions in other business areas, from energy management to sales forecasting to customer service.

For example, when AI is used to link transportation data with order data, it can improve delivery prediction accuracy and boost customer satisfaction. "A carrier might say it can deliver a product in two days, but AI might determine that, based on the carrier's past performance, there is a 30 percent chance it will take five days, not two," Neights says. That gives customer service and supply chain professionals proactive alerts about potential fulfillment challenges.

AI in action for a small business

One small business that is ahead of the curve with AI is Medisafe, which has developed a platform and app to improve medication adherence. It passively collects patient data (medications prescribed, conditions being managed, time and amount of doses taken, engagement levels with the app), and it uses that data and self-learning algorithms to personalize the experience for users. Medisafe is developing AI solutions in-house as well as leveraging third-party sources.

"In working with partners, including other health technology companies, U.S. payers, etc., we bring in other patient behavioral, demographic and psychographic data," says Jon Michaeli, executive vice president of marketing and business development at Medisafe. "That helps us complete patients' profiles and uniquely address their specific impediments to adherence."

Willing to pay for better adherence

The app is free for patients and health care providers. The company's revenue stream comes from pharmaceutical companies and insurers, which pay license fees to use the technology. Better adherence means more prescriptions filled for pharmaceutical companies and lower health system costs for insurers, because healthier patients require fewer hospital admissions and procedures.

Michaeli expects AI to play an important role in Medisafe's future. "It fits into the 'surprise and delight' category," he says. "When you can deliver exemplary user experiences to serve needs above and beyond all expectations, solving for people's unique barriers to success, you have the potential to win them over - and, in our case, indefinitely improve their health and well-being."

This story originally appeared on Inc.com.