How AI will fully rework training
Alistair Berg / Getty Images
Nothing seemed strange about the new teaching assistant, Jill Watson, who briefed students on assignments and due dates in Professor Ashok Goel’s Artificial Intelligence class at the Georgia Institute of Technology. Her answers were brief but informative, and it wasn’t until the end of the semester that students learned that Jill wasn’t a “she” at all, let alone a human. Jill was a chatbot built by Goel to offload his eight other human TAs.
“We thought that if an AI TA would automatically answer routine questions that normally have clear answers, the (human) teaching staff could win the students over to the more open questions,” Goel told Digital Trends. “It was only later that we were motivated by the goal of building human-like KI-TAs so that students couldn’t easily tell the difference between human and KI-TAs. Now we’re interested in developing KI-TAs that will improve student engagement, retention, performance and learning. “
AI will change both the face and the function of education.
AI is quickly integrating into every aspect of our lives and like the students in Goel’s class, we are not always aware when we are dealing with it. However, the impact of AI on education will become apparent in the years to come as these systems make their way into the classroom.
Like computers and the internet, AI will transform both the face and function – the what, why and how – of education. Many students are taught by bots instead of teachers. Intelligent systems advise, teach and grade tasks. In the meantime, the courses themselves will change radically as educators prepare students for a job market where millions of roles have been machine-automated.
AI-powered education may sound like something from the distant future, but it’s already a topic of interest to both academics and businesses. AI-powered educational toys have flooded the market in recent years, many of them through crowdfunding platforms like Kickstarter and Indiegogo, where they often exceed their financial goals.
BSIP / Universal Images Group / Getty Images
Professor Einstein, for example, teaches kids about science with goofy facial expressions and a German robotic lilt. The company’s Kickstarter campaign, developed by Hanson Robotics and supported by Watson from IBM, raised nearly $ 113,000. Meanwhile, the startup Elemental Path is offering CogniToys, a range of educational intelligent dinosaurs designed to play games, have conversations, and help kids learn to spell. Their Kickstarter campaign raised a whopping $ 275,000 from backers in 2015.
“Big things are on the agenda for AI-powered educational toys,” Danny Friedman, director of curriculum and experience at Elemental Path, told Digital Trends. “I use them in every classroom as a complementary learning tool that is not only built into a teacher’s curriculum, but also tied to a student’s personalized data such as preferred learning methods and areas of interest. I also put them in every home, not just to answer questions, but also to encourage pro-social interactions. AI-powered toys will be just as ubiquitous in households as the cell phone. “
“Big things are on the agenda for AI-powered educational toys.”
A student’s engagement with AI only increases as they progress through the school system. AI learning toys are being replaced by tutors, whose job it is to identify weak points and enable additional training.
Teachers are relieved of the simple task of grading papers in subjects ranging from natural sciences to social sciences. Systems like Wolfram Alpha can already answer complex math equations and questions in a language that is informative and accessible. Integrating such an engine into an automated rating system – especially with quantitative problems – would be a breeze. The educators will be happy to be able to focus on the more personal aspects of the education.
“When it comes to AI in teaching and learning, many of the more routine academic tasks (and those that are least rewarding for faculty) such as B. the grading of tasks, can be automated, “write the researchers Mark Dodgson, director of the technology and innovation management center. University of Queensland Business School and David Gann, Vice President of Imperial College, in a report on AI and higher education for the World Economic Forum.
Pepper, a multipurpose human robot design from Softbank, is the first humanoid robot to be used in Japanese households.
Once a student reaches high school, she may well be entering freshman next to what AI education experts in one Pearson report refer to as the “Companion of Lifelong Learning”. For the past nine years, this digital companion would have accompanied her in class, helped her with her homework and learned with her.
The learning partner – who may manifest as a robotic T-Rex, or more likely something more subtle like a smartphone application – even acted as a student on occasion, so the human student could teach them what they learned and deepen their knowledge.
“This companion would be accessible to the entire student [his or her education]”Wayne Holmes, co-author of the Pearson report and professor at the Open University’s Institute of Educational Technology, told Digital Trends.” It could always indicate that they are doing or helping them with work they find difficult It will also provide information for the teacher to engage with. The idea is that over time the tutor can create this profile of the individual that can be used to assist them in their further development. “
These digital learning partners should support rather than replace teachers, emphasized Holmes. In fact, he expects educators to have their own AI assistants to make their jobs easier and more effective.
These digital learning partners are designed to support rather than replace teachers.
“Teachers would have their own companion, they would be their own AI teaching assistant,” he said, adding that a student’s companion and a teacher’s AI assistant would “communicate with each other so that the teaching assistant knows what is going on with the profile of the individual student is going on and would be able to interact with it. “
When a student enrolls in a university, if you will, she will be the product of two “heads”: the AI in her brain and the AI she developed as a learning partner. And at the university itself, AI will be everywhere – as TAs in the classroom, as support in the enrollment office, and even as academic advisors. This year the Technische Universität Berlin used a chatbot called Alex to help students plan their course calendar.
TPG / Getty
“I think the advantages of the chatbot system lie in the completeness and availability of the information,” said Thilo Michael, currently a doctoral student at TU Berlin, who designed the system as part of his master’s degree. “The chatbot tries to translate students’ questions into searchable queries just like a human advisor would, but has all the information available at once. Human advisors would have to search various online systems and might even provide incomplete information. “
Michael emphasized that the system is not intended to replace people. “The system is able to answer pragmatic questions about the courses and majors available, but it cannot answer questions on a broader level,” he said. “I think the system could very well be used in combination with advice to have the best of both worlds.”
Outside of traditional learning institutions, AI has the potential to make education accessible to more people. In developing regions where there are few teachers, a robust AI system can be used to teach students with minimal or no engagement from a human educator.
The XPrize Foundation, which develops Moonshot competitions to promote “Radical Breakthroughs for the Good of Humanity,” provides the team that will create the best basic learning application that can replace a teacher for children with access to a tablet but not a human currently US $ 10 million to educators. In June, XPrize selected eleven semi-finalists from nearly 200 teams to compete in the Global Learning competition. It is likely that the winning system will be assisted by the AI to offer more personalized and dynamic lessons.
Still, there is no shortage of ethical issues that need to be addressed before AI is fully implemented in education, something Holmes and his colleagues are quick to recognize. For one thing, educators need to take into account the privacy and confidentiality of the data collected, especially when that data relates to children. For example, who does the information belong to? And who will have access to it?
“There is no obvious answer to this problem, but it is an issue that needs to be addressed,” said Holmes.
And before psychologists educate a generation with AI tutors, they should understand its developmental implications. Do students become addicted to the technology? And what happens if the system works incorrectly or fails? Similar difficult questions, but those that are worth answering for the future of our greatest resource – the minds of the next generation of humanity.