William Cope & Mary Kalantzis | University of Illinois, USA
This video is part of the session »Artificial intelligence and its application in information and media literacy education« in the conference on »Information Literacy and Democracy« 19th and 20thJune 2020.
The live session takes places on June 19th at 04.15 p.m. Central European Summer Time (CEST). If you want to join the live session, please send us an e-mail to infodem (at) uni-hildesheim.de.
You are cordially invited to discuss here. In this way, it is possible to provide feedback and pose statements and questions even before the live event.
First of all, I would like to thank you for this extensive presentation about Artificial Intelligence and its application in pedagogic field. As someone who is interested in Artificial Intelligence and machine learning it was fascinating to see the use of AI in education. I really like the concept of Scholar and think it really helps to avoid the unnecessary loss of valuable information/data that gets lost in a conventional classroom. Not only does the system benefit from all the input provided by teachers and learners, everyone using it is benefits from it. To be able to benefit from such a system you emphasized that the learners have to engage and be part of all the processes. What I would like to know is how it can be ensured that everyone is motivated enough to be active?
First of all, I want to say thank you for the very interesting presentation about Artificial Intelligence. The presentation showed me that Artificial Intelligence is also becoming increasingly important in the educational field. It also becomes clear that supervised and unsupervised machine learning cannot be seen separately. On the one hand, the machines are dependent on us humans, because the more people use a system, the more test material the system gets to work with, learn from it and constantly improves. On the other hand, as the user we supply material from which the machine recognizes patterns. The combination of both probably results in maximum success. The future project sounds extremely exciting and I wonder how one can manage to train the system to the point where it can understand the underlying thinking in texts at some point. This still seems unimaginable to me because the system has to receive many different data and you cannot expect every user to give complex feedback. One question that has arisen is if the learning success curve is saturated at some point? Does the system stop learning at some point when the input from the users becomes similar?
I support your point that currently human interactions are to be performed digitally which means that right now is an evolving time for artificial intelligence (AI). It became clear that for successfully using AI, huge calculations have to be made which can’t be done only by humans – machines and humans rather have to work in a synergistic way. The huge potential was shown with the program “Common Ground Scholar”. This program provides a platform that creates and spreads knowledge. Here, AI has the capability to use information literacy concepts to guide for example students through search and critical evaluation processes which is a point that I would say should be more prevalent in education (e.g. starting already in school). For generating enough data points, there has to be a lot of interaction and many people to join programs like these. Also, I think it is still hard to move away from current familiar learning management systems. Even though I think the inhibition threshold for being a part in developing AI is getting lower, I wonder if for example students are really motivated enough to engage. Right now, I’m rather experiencing the opposite in “digital classrooms”. There has to be willingness from both sides learners and teachers to use such programs as well as transparent goals and data in order for AI to eventually perform human-like.
Dear Mr. Cope and Mrs. Kalantzis,
Thank you for your very interesting presentation. I think that it is unavoidable to know that the digital world is growing and expanding into the human world. But the information you shared made me once again aware of the rapid development of artificial intelligence.
AI is now present all over the internet for example in search engines and on social media platforms. It opens up possibilities like never before in various areas of life. Even in the education field AI can be a help for both students and teachers. Your example of the project “Scholar” showed that AI should be considered for the future in (digital) classrooms. The machine is able to learn from human interaction on learning platforms and is subsequently able to give feedback. In this setting learners themselves are helping the machine to learn by providing information. They are now not only consuming new information but become teachers in this sense as well. When the machine then uses what it has learnt and applies it on e.g. future courses I find it interesting to know if there could be a danger in fully trusting AI and its self-generated feedbacks.
In this interesting session I learned that artificial intelligence is everywhere and that it is around us all the time. Gathering data happens seasonlessly and endlessly. The medium is the message for learning artificial intelligence. Computers can calculate faster than humans but there are also aspects computers can’t do. Machine learning is central in AI because human are training machines. Machine learning is separated in supervised machine learning (human, label meaning) vs. unsupervised machine learning (calculate numbers, clusters). The naming of things happens for example with barcodes or IDs. The binary world makes it possible to calculate numbers. The four main aspects of artificial intelligence are namability, calculability, measurability and renderability. Artificial intelligence comes from machines, which are equipped with comprehensive analytical capabilities. This power results in various – typically human – abilities such as perception, learning, reasoning, planning and decision making. Machine learning means computers are enabled to learn without being explicitly programmed for certain details. The computer recognizes patterns and regularities in the provided data. Just as humans learn to communicate, to recognize certain patterns (for example in the form of grammar) or to follow the rules when driving a car, machines can be trained to do this and then take over the associated tasks independently. Since AI systems are usually based on ML, the two terms are often used synonymously. Here we understand ML primarily as a necessary prerequisite for AI. For machine learning a system needs to „know“ sets and classifiers (cf. Kubat 2017: 1). Software needs a lot classes from letters and digits to identify characters (cf. Kubat 2017: 2). Attribute vectors need to be described precisely. Attributes are factors like size, shape, and filling colours (cf. Kubat 2017: 3). The system gets trained the classifier to identify if a attribute is true or false and to combine attributes (cf. Kubat 2017: 3). This example shows that humans are „behind“ artificial intelligence of machines and train the systems as a helping-tool to make calculations and the other named aspects easier in theirs lives. The examples from Kubat (2017) also show, that systems have to learn every particulars very accurate to make use of it. Machine learning would not exist without artificial intelligence.
Kubat, M. (2017). An Introduction to Machine Learning (2nd ed. 2017.). Cham: Springer.
o.A. (2020). Artificial Intelligence (AI) / Machine Learning (ML).
https://www.fostec.com/de/kompetenzen/digitalisierungsstrategie/artificial-intelligence-machine-learning/. Last retrieved: 07.07.2020.