What are AI assistants and tutors? How are they being used? Are they coming for your job?! Your children?! Your soul?!
AI Assistants
AI assistants have been present in our homes and phones for some time now. Many consult Siri or Alexa on a daily basis and they are being used by teachers and students alike to do everything from defining words to playing games of Jeopardy. These AI assistants are also now being incorporated into many of the most common productivity apps used by learners. Micorsoft recently introduced their AI digital assistant called Copilot which is powered by ChatGPT and Google has their own version built into Google Workspace.
These AI assistants will soon be used as naturally and frequently as spell check.
Children are using smart speakers powered by AI assistants like Siri and Alexa at home and in school.
Defining words and acting as a talking encyclopaedia are some of the most basic ways these AI can be used. Teachers and students are finding news uses every day.
AI Tutors
AI tutors such as DuoLingo’s owl differ from Siri and Alexa in that they are specifically designed to aid students in learning by adapting content and pacing to individual needs and abilities. The AI is able to answer students’ questions as well as provide feedback, and as we can see with ChatGPT, do so in a conversational manner. The learning analytics built into the apps and sites that utilise AI tutors can provide valuable feedback to teachers and learning designers about the progress of students. Major online education platforms have introduced AI tutors and are seeking to address concerns about students using the technology to cheat. Students might ask for the answer but the Khan Academy AI - see the video below - won't give it to them but will offer tips and clues. AI is also helping teachers with the huge volume of administration they have to deal with and, theoretically, freeing them up to spend more time with students.
Learning Theory informing AI research and development
In the video below Dr. Burr Settles who is the research direct at DuoLingo explains how the zone of proximal development lies at the heart of the company's proprietary AI algorithm. The algorithm is trained on over five billion exercies each week and is therefore able to predict what exercises will be too easy or too difficult for any given student and offer exercises that will keep them within the zone of proximal development. What would Vygotsky think about this all if he were around today?
Where to from here?
The chart belows shows the huge sums of money that are being invested into AI in the education sector. Where there is this much money and interest, new developments are sure to follow. I think we will see far more interesting uses of AI when it is incorporated into wearable technology which allows us to interact with the AI and the world around us in new ways.
The speed at which the technology is developing I find to be simultaneoulsy exciting and terrifying. My first reaction to using ChatGPT was to promptly fall off my chair and say, "Oh my god, this is so cool!", with the later considered reaction being "Is this damn thing gonna steal my job as a learning designer before I've even started it?!"
I'm generally an optimist and technological evangelist but the disruption this technology will cause has caused me enough anxiety that getting this assignment done has been a challenge.
Where is AI funding going? From 2019 to 2020 there has been a huge increase in investment in education. Over four billion dollars was spend on AI with an education focus.
Source: State of AI in 10 charts. Stanford University Human-Centred Artificial Intelligence Blog.