Artificial Intelligence (AI) has evolved over the years. The scope of what AI covers has changed. AI is used in robotics, various computer systems, but especially in bots. Virtual Assistants are simple until AI is injected. AI has been defined different ways. One common categorization is 6 branches of machine learning. First is the physical branch, where censors and complex algorithms allow the machine learning to control hardware or hardware managed by augmented systems.
The second trend is less semantic. This is where the AI system is provided with a domain, method of how system works, and the AI system learns to make it more proficient. Instruction can be devised with scenarios, and the machines used by social systems will learn and potentially expand beyond original goals. The forth goes beyond the social trend.
It gives the idea that there is less design needed. Like the social impacts, if you build instructional maps, even concept maps, the instruction guided by AI can add to the map and expand the instruction for the students. Openness trend is easy to understand. It means the search for information and use of AI to expand the knowledge of the information is unlimited (Dillenbourg, 2016).
Most AI until early 2000’s was simple mapping, but machine learning has changed the dynamics. Siri was introduced, simple and crude, but triggered the evolution of AI towards cognitive learning. Teachers can leverage AI for students who are confused from the lectures. The technology can help students with homework, and assist in grading. This is shifting the role of the traditional teacher into a facilitators. Assisting the pedagogy is important, but won't replace humans. AI should automate tasks in the learning experience, but not replace the central resource, the facilitator (Fahimirad, 2018).
AI provides students with feedback on deliverables. AI in most educational applications have algorithms that perform simple tasks, such as a chatbot answering simple questions. There are times when AI apps have intelligence and learn, even if there is no prior knowledge. Students learn at different rates. Instead of using rule-based AI, machine learning is being deployed to drive cognitive learning (Murphy, 2019).
Trust is a problem when deploying AI into virtual assistants. Access to big data without violating security regulations can be a challenging. Virtual Assistants are gaining trust in every day use and they have access to the best search engines on the planet.
AI systems, like Watson from IBM, build cognitive systems that can learn and instruct. Virtual classrooms with AI were developed, but not for MOOC’s in mind. The article used Watson AI system, which they called Jill Watson, to bridge AI with wit MOOC’s (Goel and Polepeddi, 2019). The technology was used in three experiments. In each experiment, there were successful improvements. There were varying degrees, in which there shows there is still room to improve. The big value in the research and the systems built were that we are closing the gap. If the instruction is concentrated and specific, AI systems can be used as tutors and function similar to TA’s. Like instructor and ta models, there will always be a need for instructor and the virtual assistant to work together in the instruction.
Students and teacher relationships are important and need to continue to develop. Voice Activated Virtual Assistants help bridge the communication between students and the teachers. Virtual classrooms with AI have been created, but there are more opportunities to help teachers in instructional design models.
The technology was used in three experiments. In each experiment, there were successful improvements. The big value in the research and the systems built is we're making great symbiotic relationships between students and teachers. If the instruction is concentrated and specific, AI systems can be used as tutors and function similar to TA’s. Like instructor and TA models, there will always be a need for instructor and the virtual assistant to work together in the instruction.
Digital Virtual Assistants leverage AI to expand the congitive value for application in education. AI machine learning makes these systems even better. DVA's can struggle to fully inject the personal touch. To take conversations and voice conversations to the level of trust for meanigful conversations, natural language needs to be added to complete the experience.