There is a great deal of noise about AI in the media — and most of it is distracting.

Part of the confusion is due to luminaries like Elon Musk with his dire warnings about the future, along with the popular and fantastical depictions of AI common to science fiction.

In reality, AI has barely begun. Most AI today is simply some form of machine learning that’s dependent on instructions and other input provided by a programmer.

“We can trace the origins of these algorithms back over the last forty years,” says Patrick Conrad, a software engineer in the aerospace industry. “It’s just recently, by combining the latest research and computing power, that we've begun to deploy them broadly.”

This evolution in computing has also created a frenzy among companies and their marketing teams, allowing many to claim AI as core to their business. Yet, more often than not, these companies are simply improving their analytics, insights and efficiency. These are not true AI companies.

With “true” AI still in the future, companies could still use help evaluating the space. Investors and business leaders should train their focus on core technologies, suitability and practicability to guide them through this often perplexing landscape.

Core Technology

The concept of core technology speaks to companies that are intent on improving the core AI tech stack. Companies with this particular focus have seen exorbitant growth over the past couple years.

Yet, core technology development in AI is a massive and capital-intensive undertaking. Betting on the right company will be difficult but also immensely rewarding for the lucky few. More and more, companies ranging from behemoths like Google and Amazon to start-ups like Zadara Storage and Ayasdi are pouring millions into the storage, compression and application of data. Their advances with neural networks, NLP (natural language processing), decision management, hardware development and other key technologies are needed to create the plumbing for “true” AI tomorrow.

Suitability

If your business is not about improving the fundamentals of AI, you need to be sure AI is fundamental to you. “Business leaders need to ask themselves: Does AI solve a big, important problem holding them back — or could it massively disrupt my business?” says Indraneel Mukherjee, CEO of LiftIgniter, “personalization tech” firm. (Full disclosure: my company has helped fund LiftIgniter.)

If the answer to Mukherjee’s questions is, “yes,” companies must be prepared to invest in data, infrastructure, software and people—they must also be prepared to change in significant ways.

“There are opportunities for order of magnitude improvements in business,“ says Conrad, “but that requires the will to ask, invest and act.”

Practicability

Questions of suitability naturally lead to questions of practicability, particularly around the problems of access, efficiency and sustainability. Affordable, effective platforms that make AI available to anyone are emerging. Thirty eight percent of enterprises are utilizing these technologies today and that’s forecasted to grow over 60% in the next year. These companies will democratize AI and advance it to a point of ubiquity. As an investor or business leader, that’s a place you want to be.

“Developers must utilize machine learning to be relevant,” says Nexosis’ CEO Ryan Sevey. “Having ML crunch numbers and begin to turn data into useable information is where we are. This can’t be a bad thing for businesses to do.” (Full disclosure: my company has helped fund Nexosis.)

In practice, a democratized approach that supports future needs will lead to real business transformation and position companies to benefit when the big breakthroughs in AI tech arrive. For investors and business leaders looking to the power of AI to solve big business problems and drive big returns, tuning out the noise and focusing on these core elements will serve them well as the AI hype train rolls on.