Future of AI & Machine Learning

Machine Learning (ML) for the Future

Currently, Machine Learning relies on its algorithm to produce a usable model based on a specific dataset (Chen, Liu, & Brachman, 2018). The "generated model" is subsequently used to deal with real-life situations.

The central issue with ML model is that past learned knowledge does not inform future learning, which is contrary to how humans learn. Humans do not learn in isolation. In fact, we use past knowledge to help us understand future learning and for problem-solving. The implication is that, in order to create a robust model in the current paradigm, ML requires a significant amount of dataset and training examples; when at the same time, large quantity of data is expensive and difficult to acquire. Furthermore, any changes in the environment tend to create new situations for ML, which can render any previously created model useless.

The future of ML needs to adopt a better approach. This new concept of ML must focus on “learnware” (Zhou, 2016). It would be a specific, well-performing and pre-trained learning machine. That is to say, the learnware would be designed for specific and well-described tasks. Thus, those looking to acquire or to buy a learnware must first determine their own requirements (what it is they want the ML to do), and then search the market to identify and acquire the appropriate learnware that responds to those requirements (as shown on the Learware Market Illustration at the top right corner). One may have to use personal data to render the system more adaptable. The advantage is that it will be less expensive and will be “reusable (the learnware can be upgraded), evolvable (the learnware should respond to environmental change) and comprehensive (specifications can be added) (Zhou, 2016). But what does the future of Artificial Intelligence look like?

AI in the Future

Artificial intelligence (AI) is rather self-reliant and capable of mimicking humans on matters such as problem-solving and performing task independently. But what does the future of AI look like in a crucial area such as education?

Some of the most popular arguments made in favor of AI in education centers around time saving, cost-effectiveness and greater efficiency. But critics of AI in education especially in distance education sound the alarm against the "commodification" of learning which, in their views, will potentially harm both students and teachers and will “turn schools into digital diploma mills” (Noble 1998a, b). Daniel Shift however is quite optimistic about a future path for AI in education. It is a model built on what he calls a “robust differentiation” (Shift, 2020). That is, a system that will develop good knowledge of students from their cognitive abilities to emotions and level of engagement, and maybe even their future educational trajectories. And what will the teacher’s role be?

Teacher’s displacement has been one of the main concerns expressed against AI in education. But this futurist concept of AI in the field does not displace teachers. They will rather have new roles and still be in charge, but the AI assistant will provide updated and sound information on every student, which will enable unmatched levels of differentiation. Thus, teachers will move away from performing those simple but considerable rote task to focus more on their professional development.

Squirrel AI stands as the largest AI educational technology with millions of students in several countries (Shift, 2020). The main vehicle of learning here is a laptop. Teachers play the role that pilots have in the cockpit. The important thing is that there is a role for teachers, but the student-teacher ratios increase considerably. Even though AI in education has made considerable stride, the social-emotional roles that teachers play still remain challenging to automate.

As a language educator, I am excited about the opportunity to, one day, use AI, VR and AR as a way to effectively enhance students' cultural-awareness and their communication skills. One can make the same futurist argument for teachers' professional development where the same technology can be used to expose teachers to real life situations that could arise in the classroom.

Poll: Are you excited about how AI will transform how you do what you do?

Use this link or the QR code below to vote! https://www.menti.com/alt6eho1nqp1

Discussion Prompt: Do you have any apprehension as to how you might professionally be affected by AI in the near future?

Page updated

Google Sites

Report abuse