Advanced computer programs influence and can even dictate important parts of our lives. Think streaming services, credit scores, facial recognition software.
As this technology becomes more sophisticated and ubiquitous, it’s important to understand the basic terminology.
People often use “algorithm”, “machine learning” and “artificial intelligence” interchangeably. There’s some overlap, but they’re not the same things.
We decided to bring in a few experts to help us get a solid grasp of these concepts, starting with a basic definition of “algorithm.” The following is an edited transcript of the episode.
Melanie Mitchell, Davis Professor of Complexity at the Santa Fe Institute, offered a simple explanation for a computer algorithm.
“An algorithm is a series of steps to solve a problem or achieve a goal,” she said.
The next step is machine learning that uses algorithms.
“Rather than having a person program the rules, the system learned by itself,” Mitchell said.
For example, speech recognition software that uses data to learn which sounds combine to form words and sentences. And this type of machine learning is a key component of artificial intelligence.
“Artificial intelligence is basically the ability of computers to mimic human cognitive functions,” said Anjana Susarla, who teaches lead AI at Michigan State University’s Broad College of Business.
She said we should think of “AI” as a generic term.
“AI is much more comprehensive and all-encompassing than just machine learning or algorithms,” Susarla said.
Because of this, you may hear “AI” as a loose description for a bunch of things that exhibit some level of “intelligence.” Like a software that examines the photos on your phone to sort the ones with cats to advanced caving robots.
Here’s another way to think of the differences between these tools: Cooking.
Bethany Edmunds, professor and director of computer programs at Northeastern University, likens it to cooking.
She says an algorithm is essentially a recipe — a step-by-step guide on how to cook something to solve the “hunger” problem.
If you take the machine learning approach, you would show a computer the ingredients you have and what you want for the end result. Let’s say a cake.
“So maybe it would take every combination of all kinds of foods and put them all together to try and replicate the cake that was provided for it,” she said.
The AI would leave the whole problem of starvation to the computer program, determining or even buying ingredients, choosing a recipe or creating a new one. Just like a human would.
So why are these distinctions important? Well, for one, these tools sometimes produce results with biased results.
“It’s really important to be able to articulate those concerns,” Edmunds said. “So you can really analyze where the problem is and how we’re going to fix it.”
Because algorithms, machine learning and AI are now pretty much burned into our lives.
The Columbia University School of Engineering has another statement on artificial intelligence and machine learning, listing other tools that can be part of AI alongside machine learning. Like deep learning, neural networks, computer vision and natural language processing.
Over at the Massachusetts Institute of Technology, they point out that machine learning and AI are often used interchangeably, since most AIs today include some level of machine learning. A piece from MIT’s Sloan School of Management also goes into the various subcategories of machine learning. Supervised, unsupervised and reinforcement, like trial and error with some kind of digital “reward”. For example, teaching an autonomous vehicle how to drive by telling the system when it made the right decision — like not running over a pedestrian.
This article also points to a 2020 Deloitte survey that found that 67% of companies are already using machine learning and 97% plan to do so in the future.
IBM has a helpful graphic to explain the relationship between AI, machine learning, neural networks and deep learning and presents them as Russian nesting dolls, with the broad category of AI being the largest.
Finally, with so many companies using these tools, the Federal Trade Commission has a blog detailing some of the consumer risks associated with AI and the agency’s expectations for how companies should use them.