There has been a lot of buzz around technology and robots that can teach students about Artificial Intelligence or AI. We thought it would be a good idea to demystify some key aspects of AI and explore why it is important to understand the terminology and what it means for our students (and teachers) in the future.
The simple definition is that an algorithm is a set of well-defined instructions that the computer follows step by step in order to complete a task. For example, a cake recipe could be viewed as an algorithm for making a cake. This is the most important term to understand as it is the foundation on which robotics is framed.
Automation refers to the technique of making a process or system operate automatically with minimal (if any) human input. Examples of automation include: automatic garage opening doors, automatic pet doors, machinery production lines, etc.
Machine learning is where computers act (or learn to act) without being explicitly programmed to do so. They do this by using algorithms that tell the computer to receive and analyse incoming data to predict outcomes (within an acceptable range). As new data is fed into these algorithms, they learn to further develop their operations and improve their performance.
Machine learning is considered to be a form of AI as computers mimic the way that humans ‘learn’ from past (historical) data or behaviour. An example of machine learning is using a speech recognition tool (e.g. Dragon Speak, Alexa or Siri) that improves at picking up your words the more you use it so it seems to become ‘intelligent’ over time.
Artificial Intelligence (AI)
In its simplest form, AI refers to the ability of a computer (or computer controlled robot) to perform tasks that are generally associated with human intelligence, such as visual perception, decision-making, speech recognition, and language translation. AI machines or computer systems are programmed to think like humans and mimic their actions, and in doing so, ‘learn’ over time so that they become more ‘intelligent’.
As technology advances and AI evolves, our understanding of what constitutes artificial intelligence will change and the previous benchmarks that we used to define AI will become outdated. Already, machines that use optical character recognition to identify text are no longer considered to embody artificial intelligence as this function is now thought to be an inherent computer function.
What are the applications of AI?
AI is already being widely used in the transformation of a variety of fields. For instance, in healthcare machine learning is being applied to make better and faster diagnoses than a human is capable of doing. AI models are capable of indicating early lung cancer by correctly identifying tiny specks of cancer on CT scans at a higher accuracy rate (about 95% of the time), than human radiologists can typically achieve (approximately 65% accuracy rate) (Svoboda, 2020). This reduces the risk of a false positive (or false negative) diagnosis being made.
Researchers have also designed an AI tool called MammoScreen to identify regions on 2D digital mammograms that are suspicious for breast cancer and determine their likelihood of malignancy. This has the potential for improving radiologists’ performance in breast cancer detection (Slater, 2020). Some breast-screening AI tools can expedite tasks such as the automatic triage of patients and prediction of treatment outcomes.
Self-driving cars use a combination of sensors, computer vision, image recognition radar and AI ‘learning’ to travel between destinations without human input. We may still be a long way from perfecting an autonomous car that can be widely used, but autonomous rigid inflatable boats, underwater vehicles and aerial vehicles are already being deployed by some Navies around the world – including in Australia.
Apart from automatically grading students’ work, adaptive learning programs could be designed to adjust to the needs and abilities of the student. As the student navigates through learning modules set by the teacher, the AI enhanced programs would adjust the content and questions for the student based on their previous right or wrong answers. By learning about each student, the AI program is able to use that information to tailor the learning content specifically to that individual student. Taking into account the student’s strengths and weaknesses, the system could provide specific feedback about that student, identifying areas for improvement for both the teacher and the student.
Why is it important that we know these terms?
AI is continuously evolving to benefit many different industries from businesses to engineering, healthcare and education, banking and manufacturing, to name a very few. We are surrounded by technologies that are becoming paired more frequently with AI processes to make tasks easier, save time, reduce boredom and generally improve our way of life. AI technologies impact what we do on a daily basis – and will have an even bigger impact on what we do in the future.
Although we know it is important to prepare our students for an AI world that is rapidly evolving, we don’t really know what future workforce skills will be. We do know that as AI takes over more of the repetitive or dangerous work, humans are then released to undertake the tasks that require creativity and empathy, critical thinking, problem-solving and collaboration skills. By engaging our students with AI in our classrooms, we are teaching them these higher order cognitive and non-cognitive attributes. By teaching our students to code we are encouraging them to become creators, not just users, of AI in the future.
Interested to teach AI in your school?
Check out the Fable system. Their education team is putting the final touches on an educational program that uses the Known-Nearest Neighbour (K-NN) machine learning algorithm to teach students how to code their own AI programs that categorise tomatoes based on colour.
Slater, H., 2020. MammoScreen AI Tool Improves Diagnostic Performance of Radiologists in Detecting Breast Cancer. [online] Cancer Network. Available at: <https://www.cancernetwork.com/view/mammoscreen-ai-tool-improves-diagnostic-performance-of-radiologists-in-detecting-breast-cancer> [Accessed 11 August 2021].
Svoboda, E., 2020. Artificial intelligence is improving the detection of lung cancer. [online] Nature.com. Available at: <https://www.nature.com/articles/d41586-020-03157-9> [Accessed 11 August 2021].