Pandata Blog

AI design and development for high risk industries

Call it Machine Learning or Artificial Intelligence, the goal is to solve problems using data.  Solving problems using data is the driving force behind data science.  Everything else is just a weapon in our arsenal.  Yet, we see the buzzwords Artificial Intelligence (AI) and Machine Learning (ML) everywhere in the media.  What do they mean?  More to the point, which do we need to get to our goal of data-based problem-solving?

Machine Learning is the study of algorithms that learn automatically from exposure to data.  A lot of these algorithms come the area of applied statistics.  Statistics are a definite part of the data science arsenal. Machine Learning is one of the tools that we use to create Artificial Intelligence.   If ML is good for our problem-solving goal and ML is what makes AI happen, then is an AI solution a better one?

Not necessarily.  It depends on what you mean when you call something AI.   And here is where things get messy.   The dictionary definition of AI is that it is the theory and development of computer systems able to perform tasks that normally require human intelligence.  And that definition allows for a wide range of flexibility in what technologies count as AI.

Artificial General Intelligence (AGI) are those types of AI that fuel the imagination.  These are the AI applications that ponder the philosophical question of what makes a human different from a machine.  Movies and television are filled with examples, from 2001’s HAL to Star Trek’s character Data.

Narrow AI are the more common types of AI that we encounter in everyday life.  Narrow AI are AI solutions that are specific to a task.  Search engines are examples of Narrow AI.  So are self-driving cars and personal assistants like Alexa and Siri.   These useful tools make our lives easier by performing tasks that were once only the realm of human activity.  However, a wide range of technologies can be found in the space between a web search and a self-driving car.   How do we determine what is Narrow AI and what is just everyday technology?

No two people are going to agree on what makes a technology Narrow AI.  Is it the complexity of the algorithm that the technology is using that makes it Narrow AI?  It is the complexity of the task at hand?   Does AI always have to be built using Machine Learning or can you have a rules-based AI?  There are lots of questions and no real clear answers.   Does it really matter?

AI is transient in nature. Things that were once thought to be the sole purview of human intelligence eventually become common technological tasks that are taken for granted.  Take autocorrect.  Once upon a time, only humans could proofread and correct spelling mistakes.  Then it became a cutting-edge AI tool.  Now autocorrect is so commonplace, most people would not consider it to be AI.

What does matter is that we find solutions to data-driven problems using whatever tools are at our disposal.  Machine learning techniques help drive those solutions.  So do the technologies that fall under Narrow AI.  Maybe the solution can be found in applied statistics.   Data Science uses all the above to get at those solutions that make our lives and businesses better.   At Pandata, our data scientists and their arsenal, are at the ready to expand your competitive edge and impact your organization’s bottom line.  Contact us at to learn more.

Julie Novic was a Data Analyst at Pandata.