The “New Normal” isn’t new anymore. The world of E-learning has changed forever, and it’s not going back. Even if students are returning to school, and eventually it happens - the model of a teacher, a whiteboard, and dozens of students sitting on chairs has changed forever. Instead, multitudes of new learning tools have entered the classroom. Technologies such as conversational AI, which were used mainly in textual interfaces and for commercial purposes, have entered this field.
Chatbots became a commodity only by the mid-2010s, but for lots of people they’re already falling out of favor. Honestly, I don’t blame them. Most of the conversations with them are robotic (no pun intended), mundane and limited in their scope. Most of their uses are commercial. Most of the bots are textual, and in a world turning to video, they are being left behind like educators who haven’t mastered the usage of online platforms.
 |
| Image by Gerd Altmann from Pixabay |
There’s an inexcusable gap between what conversational AI technology and virtual humans enable today to their actual prevalent usage. Virtual human, or Digital human, is a fairly new concept that combines advanced conversational AI technology with 3D animation technology to create an avatar that users can interact with, using natural language. There are many benefits to this new technology, mainly the ability to create a deeper connection with the users which is harder to create in a text-based interface.
Using conversational virtual humans, as we in cocohub.ai learned extensively in the past months, has enormous potential in the education sector. Even more fascinating is the fact that from every meeting and conference, we understand that there are more and more use cases in which this technology can be helpful. Therefore, we are working nowadays on making the technology available and accessible for teachers and educators. We are releasing new templates which can be used for easily creating quizzes and exams in a matter of minutes. In addition, we’re already working nowadays on integrating our videobots into popular learning management systems.
Being crowned a regional track winner, “voice assistants special track” runner-ups, and finishing among the best in the global finals, all in the Global EdTech Startup Awards - which led us to present in the 2022 Bett Show in London, only validated the need for our technology. And it turns out that there are lots of uses for this technology.
So, how can virtual humans be helpful and useful in the field of education? Here’s what we discovered.
1. Performing Quizzes
Our recently-created templates are being user-tested on several teachers and digital learning leaders. They resemble popular quiz apps, but instead of a graphic interface, you get a virtual human that performs the quiz in a conversational and engaging way. After the quiz it sends the results to your email (or directly updates the grade in the LMS).
Let’s say that you want to summarize your latest lesson in a quick 5-question quiz to make sure that your students got it right. So, when entering our platform, you type in the quiz or task’s questions, the correct answer(s), the bot’s response to correct and incorrect answers - all between opening and closing statements - and add the email address to get the results. Within several minutes, you can have a ready to deploy quiz presented by a conversational virtual human, and you can send a link to your students to try it out and practice.
However, the uses go much deeper than that. As this essay is being written, 3 researchers - one from a university, one from an education college, and one from a MOOC-creating company, are intending to use our product for research purposes to see its impact on students. Which leads us to what they’re trying to accomplish.
2. Self-Regulated Learning
As Dr. Danny Glick, Research Affiliate at the University of California of Irvine’s Online Learning Research Center shared recently on Linkedin, he’s creating a conversational video-bot that can help students manage their self-regulated learning process. “We plan to conduct a joint research study with cocohub, aimed at examining whether an AI-powered ‘learn to learn’ video bot intervention can impact students’ behavior and performance in a large-scale online course”.
3. Socio-Emotional Learning
Socio-emotional learning is another place where this solution can be useful. Since youngsters today are used to having virtual friends as part of their gaming experience - or even creating virtual companions. Using virtual humans to interact with them means that the whole process becomes gamified, hence much more appealing as a way of learning, and the virtual humans can also offer encouragement and empathy when the going gets tough.
4. Teaching computational thinking and conversation design certification
Then, there are our partners from IUPUI, where our technology was used to create a virtual teaching assistant sitting inside an LMS, and the course’s curriculum included two lectures about our product. The students used it to create “virtual best friends” or “virtual parents” that joined remote class sessions and answered questions about them, such as favorite hobbies, movies or food. This use case will be presented at this year’s important Educause conference.
5. Language Learning
Other researchers have expressed interest in seeing how these virtual humans are helping students improve their English speaking skills. MOOC creators, for example, are using one of the platform’s features - the ability to count the number of words the student uses in an answer to determine if the student’s answer is long enough. If it isn’t, they ask the student to present a longer sentence. Alternatively, they can count the student’s response as successful only if several keywords or phrases are being used.
Additional use cases include grammar checking, as “fill in the blanks” practices. This method is being used for example, by Solly Tamari (“Global Effective Education”), an EdTech innovator and entrepreneur for more than 25 years. We built a prototype he’s using for teaching children to use the correct word - Am, Is, Are - in present tense. Sending this bot to dozens of his students, here’s what some of them said: “Great tool, very useful for me”; “Cool idea! It gives immediate feedback”.
6. Personalized learning
A research that was shared recently by the Israeli ministry of education, conducted by Stacey Childress and Scott Benson in 2014 for Phi Delta Kappan, highlights the advantages of personalized learning as a method of narrowing the gap between the class’ fastest and slowest learners. “This attitude is more holistic, so it impacts the class schedule, the students’ relationships with teachers and among themselves, and the students’ attitude towards learning”, was one of the research’s conclusions. Meanwhile, they saw lots of promise in personalized learning as a way to address the current system’s flaws.
Enhancing, not replacing
Interestingly, personalization of content is also one of the key advantages of conversational AI. If you can create separate conversational flows for different customers in Fintech for example, why not use the same method when giving personalized tasks about the same learning material. Our partners in higher education already raised the idea of creating different paths within the same conversational bot, to address different levels of students’ knowledge regarding the same material.
One of the toughest obstacles to overcome is actually one that stands in the way of every technological revolution: the fear by the implementers that their deeds will lead to their own profession’s demise. This concern, most lately, was raised recently by a lecturer in an education college. “First we’ll write it, then those bots will replace us?”, he wondered, probably half-joking, during a workshop we conducted with them. The remark actually led to a somewhat heated discussion, where most attendees claimed that the opposite is true. “They won’t replace you” I answered, getting used to hearing that. “But they can enhance your capabilities, enrich the learning experience and bridge the gap between your availability and the students’ needs”.
The combination of conversational AI and virtual humans in education is very exciting. Just a few years after for example, smart speakers couldn’t identify kids’ voices. Improved NLP algorithms allow new possibilities, services and technologies used for purposes like lead generation or shortening queues, can now be used to benefit the next generation of students. Seeing educators embrace the change and start to use it, participating in user feedback sessions and helping us improve our product to create a user experience that will be tailored to the teachers’ needs and interests - make us, at cocohub, much more optimistic about the future of the education system.
Guest post by Eran Soroka, head of marketing at cocohub.
