How do I get certified as an expert in AI skills?
In our previous article in this series on Getting Started with AI, we discussed how to practice your AI skills.
In this article, we’ll discuss how to get certified in your AI skills and what certifications are available.
A certification quickly provides an employer with evidence that you are proficient with your AI skills.
This is why a certificate is so valuable when you are first getting started with a career in AI.
There are several types of certifications available; however, some are more valuable than others.
So, in this article, we’ll work through each type of certification ordered (roughly) by their cost, difficulty, and perceived value to potential employers.
The first option for certification is a certificate for completing an online course. There are two main kinds of online courses: non-accredited courses from an online learning platform and accredited courses from a university. Courses from online learning platforms are typically lower cost, but they are perceived as lower value to employers. Accredited courses are more expensive, but employers perceive them as more valuable.
For free or low-cost non-accredited online courses, I recommend DataCamp for beginners, Coursera for foundational skills, and Pluralsight for experienced IT professionals. Unfortunately, these courses can be pretty hit or miss in terms of quality, so be sure to check their ratings, reviews, author credentials, and learning outcomes before committing to a course, and then exit a course if it’s not meeting your needs.
For accredited courses, I recommend online courses from Stanford, MIT, and Johns Hopkins. They have excellent courses, instructors, and degree programs in AI, ML, and DS. They are also some of the most well-respected universities in the world, so these certificates are quite valuable. However, these courses will be much more difficult, require greater investment in time, and cost significantly more money.
The next opinion is an AI learning-path certificate. This certificate is given after you’ve completed a series of courses that teach a set of related skills. There are typically 3 to 12 courses in a learning path. So, it takes more time and money to complete a learning path than a single course. However, this certificate is more valuable to employers. Again, there are both accredited and non-accredited learning paths.
For non-accredited learning paths, DataCamp offers learning paths by career or skill. Pluralsight has learning paths by role, skill, or industry certification. Udacity offers non-accredited mini-degrees. Coursera offers non-accredited specializations. As an example of a non-accredited learning path, I recommend looking into the Data Science Specialization from Johns Hopkins University through Coursera.
For accredited courses, many colleges and universities offer graduate certificates or graduate specializations. They’re less valuable than a full Master’s degree or a Ph.D. However, they are more valuable than taking a single course. They typically involve taking four or more courses within a Master’s or Ph.D. program. I recommend looking into graduate certificates in AI/ML/DS from Stanford, MIT, and Johns Hopkins.
Industry certification is a medium-cost and medium-value option. It comes in two main styles: vendor-specific certification and vendor-agnostic certification. The vendor-specific certifications focus on a single vendor’s unique platforms, tools, and technology stack. The vendor-agnostic certifications focus on the industry patterns, practices, and tools without relying on a single vendor’s platforms or technologies.
For vendor-specific certifications, check out AI/ML certificates from Microsoft, Google, and Amazon. These are the most valuable because they are the most widely used AI/ML platforms. For vendor-agnostic certificates in the US, check out the certification programs from the Artificial Intelligence Board of America (ARTiBA) and the United States Artificial Intelligence Institute (USAII)
A degree in AI or a related field is the most expensive and time-consuming certification option but also the most valuable. An undergraduate Bachelor of Science degree will typically take four years to complete. A Master’s degree is another two years, and a Ph.D. is typically another 2-4 years. This is a significant commitment of time and money. However, it’s the best way to ensure you have all the right skills.
For undergraduate degrees, I recommend looking into a Bachelor of Science in Computer Science (CS), Data Science (DS), or Machine Learning (ML). For your Master’s degree, you can choose CS, DS, or AI. However, the AI Master’s degree is considered a terminal degree, so there isn’t a path to a Ph.D. in AI (yet). For your doctorate, getting a Ph.D. in CS, DS, or Cognitive Neuroscience (CNS) is (currently) the only option.
Be sure to get certificates for each course you complete, and include all your certifications on your resume, LinkedIn profile, and website. Don’t include certificates for each course if you’ve completed a series of courses (e.g. a learning path, vendor certificate, or degree); just include the final certificate for completing the entire series of courses.
To learn more, please check out my next article in this series: Getting a Job in AI.