Machine learning touches so many aspects of our everyday lives, from the way our TikTok For You pages are curated to Amazon Alexa’s ability to recognize the sound of our voices. But, as ubiquitous as the technology is, there are still some misconceptions about the machine learning field.
You might have wondered: How can a machine have thoughts? Can AI robots become sentient? And is machine learning going to replace our jobs? Ahead, Nitya Mandyam, Senior Curriculum Developer at Codecademy, shares the truth about some of the most common myths and misconceptions people have about machine learning and artificial intelligence (because, yes, those are different things).
We hope that demystifying the world of machine learning and artificial intelligence empowers you to learn some of the skills that go into the technologies. Considering a career in machine learning? Whether you want to become a Machine Learning/AI Engineer or a Data Scientist, we have lots of resources that will help you develop the skills you need to break into this exciting industry. Be sure to explore our whole catalog of machine learning courses to see what’s possible.
The misconception: All machine learning is artificial intelligence.
While you might see the term “AI” slapped on anything that involves machines automating tasks, there’s a difference between AI and machine learning. The simplest way to think about it is that machine learning is a subset of AI, Nitya explains.
The same way that robotics replicates human motor skills, “AI concerns itself with the broader task of replicating human cognitive intelligence,” Nitya says. Machine learning, on the other hand, uses algorithms to make decisions based on patterns found in data.
Distinguishing between AI and machine learning is more than just a nomenclature issue: “To ask the question, What does it mean for a machine to think?, is to ask the question, What does it mean for humans to think?” Nitya says. If you want to dive deeper into the various ways machine learning is impacting the world, check out our course Learn the Basics of Machine Learning.
The misconception: Machine learning will take away all our jobs.
People have been fearful that robots and machines will replace the human workforce since the Industrial Revolution. Nowadays, you hear a similar refrain about AI and machine learning. (Even tech bigwigs like Elon Musk have declared that AI will eventually make jobs “pointless.”)
One pressing concern to consider when talking about the job market in machine learning is the “ghost work” that goes unseen. Within AI systems, there are certain tasks that are not cheaper or more efficient to automate using machines, Nitya explains.
For example, there are people who are hired to do crucial — but non-technical — tasks, like flagging content or labeling data, so that machines can “learn.” This is a phenomenon that anthropologist Mary L. Gray dubbed, “ghost work,” which is essentially the idea that there’s a hidden labor force in tech performing the task of a machine. These jobs typically pay very little and do not provide benefits, creating a problematic wealth disparity within the tech industry.
The misconception: You need a PhD to get hired as a Machine Learning Engineer.
A Machine Learning Engineer is responsible for building the systems that computers use to learn and make predictions on their own. You do not need an advanced degree or years of academic research under your belt to get hired as a Machine Learning Engineer, but you do need to know how to code.
According to Nitya, working in machine learning requires familiarity with math and statistics. “Stats is at the heart of knowing whether or not you know a prediction is right or wrong and doing that type of analysis,” she says. (Need a refresher on statistics? Check out our path Master Statistics with Python or our beginner-friendly course Learn Statistics with Python.)
Beyond those technical skills, machine learning is very well-suited for people who “have the curiosity and inquisitiveness of a scientist, and the willingness to go down a rabbit hole,” Nitya says. People who like building things also thrive in the field, she adds.
The misconception: Machine learning can be used to predict the future and solve any problem.
Machine learning is more like a reproduction of the past than a crystal ball that predicts the future, Nitya says. Machine learning technology enables us to pick up on patterns that we would miss on our own, and then use that information to inform future decisions.
People often assume that machine learning is a magical panacea that can solve any problem, which speaks to the excitement around the technology. “When a technology shows promise, you have this knock-on effect where everybody wants to use that technology,” Nitya says. But machine learning is not always the right fit. For example, machines aren’t more objective than humans, and relying on them for everything can have negative consequences.
To use an analogy, different DIY projects require different types of tools. “Sometimes you need a pen knife, and sometimes a power tool — it all depends on the problem you’re trying to solve,” Nitya says. In this case, machine learning would be considered a power tool. “Sometimes you don’t need knives at all,” she adds.
Want to land a dream job in AI or machine learning? Here are the best job boards for finding positions in tech, a guide to common machine learning interview questions, and an overview of AI engineer salaries. Be sure to check out our career center for tips from tech recruiters, portfolio projects, and more.