How to Get Ahead of the AI Curve

AI. It’s everywhere. It’s all people seem to talk about. In most every industry. AI seems to be the buzzword of the day, the year — the decade. Some corporate leaders are skeptical of all the hype and others can’t wait to get started. Even the doubters have to respond to the customers, directors and employees constantly asking for their stance on AI. And many of the ones who are sold on artificial intelligence don’t know where to start.

>> LISTEN: Implementing AI Is Simpler Than You Think

Regardless of which camp you fall into, here are some golden rules for how you can get going with AI and set yourself up for success.

Focus on one-second problems.

Paraphrasing a quote from Andrew Ng, if a human being can make a decision in one second or less, it’s a great candidate for AI. Look for places in the organization where humans are repeatedly making lots of single-second decisions — these are potential areas where AI can learn and repeat, creating significant financial savings. At athenahealth, we were using hundreds of humans to go one by one through faxes, determining what was important and what was junk — like human spam filters. The AI we built can recognize these patterns and perform the tasks faster and better, freeing humans up to make the complex decisions only humans can make.

Look for where scaling is stretched.

Even one-second problems, if only done by a few people, may not be enough to justify an AI investment. But there are some tasks that are simply too large and complex to be handled by humans. For example, a tourist and driver outside of a foreign landmark can easily assess the supply and demand of taxis and negotiate an appropriate fare. But only AI can scale that transaction by the billions to support an operation like Uber or Lyft. These large-scale tasks are perfect candidates for AI investment.

Create a decision framework.

As soon as you start talking about AI at your organization, people will inundate you with ideas: good, bad, realistic and unrealistic. It is important to create a decision framework that reflects what is vital to the company and its leadership. It might be ROI, it might be client satisfaction, it might be Net Promoter Score, but the answer will determine what you say yes to and what you don’t. The key is to tie the decision framework closely to the organization’s strategic objectives.

With these three elements aligned, the scaffolding is in place to start planning and realizing specific AI projects.

^{Girish Venkatachaliah is vice president of athena.intelligence at athenahealth.}

Get the best insights in healthcare analytics directly to your inbox.

Related

Who Is Responsible When AI Fails?

Mission Bio Raises $30M for Cancer-Focused Precision Medicine Tech

Precision Medicine, Sure. But What About Precision Payments?