How Generative AI Will Change the Way We Learn

When a topic comes up that I’m unfamiliar with, I often turn to ChatGPT. And it’s not just me who prefers learning this way. A recent survey showed that students and parents increasingly favor studying with ChatGPT over human tutors. 95% of students who studied with both a tutor and ChatGPT report that their grades have improved since adding generative AI to their toolkit.¹ The ability to get questions answered immediately – and at all hours of the day – is one reason students are drawn to these tools. Additionally, the tool allows students to learn at their own pace and refer back to information as many times as they need.

The market has taken notice of this shift. Tutoring company Chegg’s shares dropped more than 40% after the company said ChatGPT is cutting into its business.² Duolingo and educational media company Pearson also saw their shares drop over similar concerns. There are good reasons to believe that these trends will continue as:

Plugins fix known issues. There are some tasks that ChatGPT is not good for. It is not designed for advanced mathematics, and its answers to even simple math questions can be wrong. But there are plugins created to solve for this, catering specifically to complex math problems.³ Other plugins let users learn new languages in a conversational way within the ChatGPT interface.⁴ To help him learn Chinese, a New York Times technology writer used two plugins, Link Reader and MetaMentor, to generate flashcards.⁵

New tools focus on learning, not speed. To boost learning outcomes, Khan Academy has developed a user experience that leverages ChatGPT not to simplify access to information, but to force students to work for an answer.⁶ In the same spirit, Duolingo‘s Max gives simple explanations for why an answer is right or wrong and “roleplays” to let you practice using a new language in different social scenarios.⁷ Byju’s MathGPT guides students on math problems using analogies and visual aids.⁸

Search engine tech improves verifiability. Microsoft Bing and Google have incorporated generative AI technology into their search engines, giving users the benefits of both these technologies. That means users can receive more current information and also get sources to help verify results (see Figure 1).

Classroom integration is underway. Three public schools in New Jersey began piloting Khan Academy’s automated teaching aid Khanmigo in May.⁹ A school in California plans to use ChatGPT to help students grasp simple social studies concepts, thus freeing up time for more analytical classroom discussions.¹⁰ Harvard University recently announced that it would develop an AI chatbot to be an instructor in its Introduction to Computer Science course (CS50).¹¹

Integrated multi-media is coming. We’re on the verge of having digital tools that can match not only the end users’ skill levels, but also their preferred learning styles, whether it’s text, video, audio, or a combination of all three. Launched in November 2022, text-to-video startup Prof Jim is turning traditional textbook content into multimedia experiences and courses using AI, with avatar instructors.¹² For example, Pythagoras can explain his famous mathematical theorem on the shores of Ancient Greece or instructors can create avatars that looks just like themselves.¹³ EdTech startup AcadeMe is launching AcadeMe+, an AI-driven visual learning platform where teachers can search for any subject or keyword to find a matching movie, clip, or interactive lesson.¹⁴ In addition, the platform's AI Smart Lesson Creator can help them develop their own film-based lessons.

What’s Next

As we grow more accustomed to learning with these new tools, we should expect expectations to shift, with higher demand for flexible, real-time learning. The rise of generative AI in education means that the future workforce will likely be composed of individuals accustomed to self-directed, interactive, and customized learning experiences. Management styles will likely need to be adjusted to reflect this change. Individual autonomy, for instance, will be critical for employees accustomed to this kind of learning. AI-driven learning also involves real-time, immediate feedback. Managers, or the apps they use for their employees’ professional development, may need to recognize accomplishments – and provide necessary constructive criticism – immediately, rather than waiting yearly or even quarterly to give that feedback.

Questions to Consider

What benefits will learning with generative AI tools bring compared to traditional learning? This technology doesn’t always return the same answer, even if the exact same question is asked. What does that do to our learning abilities? Can we use this to our advantage as we try and increase our knowledge on new, unfamiliar topics?

How can we best measure the effectiveness and accuracy of AI-driven corporate training and customer education programs? How can we correct – or at least limit – misinformation? Can we incorporate some level of verifiability into these tools?

How might AI-driven education change the skillsets that companies look for when hiring new employees? Will companies need to look for different skills when hiring managers or can we train them to work with this new generation of AI-driven learners? What about customer representatives? Should AI literacy be a criterion we consider for all knowledge worker roles?