Artificial Intelligence: What it is and What it’s Not

Artificial Intelligence

There are a lot of companies that claim they offer solutions that are powered by artificial intelligence (AI) and/or machine learning. But what exactly does that mean?

AI and machine learning

Artificial intelligence is a system that learns. It acquires knowledge and learns how to apply it. AI, as defined by Qualcomm, is an umbrella term representing a range of techniques that allow machines to mimic or exceed human intelligence.

If you’re thinking that AI has to simulate human intelligence, that is not necessarily true, says John McCarthy, a pioneer in AI research. Some of the branches of AI are:

  • Search: examining large numbers of possibilities
  • Pattern recognition
  • Learning from experience

According to MIT Technology Review, machine learning is a subfield of AI that looks for patterns in massive amounts of data. Machine learning algorithms have many applications, from powering Netflix recommendations to predicting disease risks. What AI still does not understand is causation. “They see that some events are associated with other events, but they don’t ascertain which things directly make other things happen,” says technology journalist Brian Bergstein. “It’s as if you knew that the presence of clouds made rain likelier, but you didn’t know clouds caused rain.

AI or not? How to find out

There are many solutions out there that are labeled AI or machine learning – but aren’t, according to Functionize, a software platform testing company. These things include coupon generators, chatbots, and car diagnostics. Most of them work from limited sets of data and are not powered by learning algorithms. Ask most chatbots about something that is not in their script, and they usually will not be able to respond. These are called rules-based chatbots.

The best way to determine if a solution is AI-powered is to ask:

  • Is it trained to analyze a pool of data and make the correct choices?
  • Can it be further trained by using an expanded data pool?
  • Can it recognize and account for changes in data patterns?

About AllazoHealth’s AI engine

AllazoHealth utilizes artificial intelligence to make a positive impact on healthcare outcomes. We help pharmaceutical companies, payers and pharmacies improve both patient health and return on program investment.

Our AI engine learns from a comprehensive data set of over 14 million lives, including payer, provider and retail claims, social determinants of health, and patient interventions. These diverse data sets complement each other by providing multiple views of patient behavior. We can access broader and deeper data on individual patients than any individual organization.

AI gets smarter over time because it continually learns from new data. So with AllazoHealth, your patient engagement program will take immediate advantage of our years of learnings.