How to leverage A.I in your lifelong learning journey.
Self-directed learning is the process of taking responsibility for one's own learning. Although this concept sounds simple, we don't get as much practise as we should on how to be self-directed learners.
We generally come out of the senior phase in the school system after studying and being tested on a particular curriculum. In University, although a greater degree of autonomy is expected, subjects and exams are still set by someone else, and you are expected to learn about them.
We often work in someone else's frame of reference and there is often little time to explore how this articulates with your own deep interests.
However, the learning journey is more powerful and effective when it is self-directed. One can follow one's own interests and begin to learn about subjects at a suitable entry point. One can develop a frame of reference for a given subject in a holistic way. Passion for a subject is a great driver of knowledge.
We are still taught, to a large degree, in a system which was created during the First Industrial Age. As the late Sir Ken Robinson said:
"Schools were invented for the Industrial Revolution. They were designed to produce workers who could follow instructions and work in a factory setting. But the world has changed, and the skills that we need to succeed in the 21st century are very different."
Since Sir Ken Robinson's famous TED Talk, the ground has shifted beneath us. Large Language Models such as ChatGPT and Google Bard have arrived. Going through school, knowledge is taught as facts to learn, essays to write. But when an essay can now be written in a few seconds for free, where does that leave the school system? What kind of learning should we pivot to, now that the main currency of our education system can be counterfeited with such ease?
In a world where knowledge is free, widely available, instantly accessible and overwhelming in scale, how to curate this knowledge and synthesis it creatively is now the question. And this learning can only happen when it is directed by the person themselves - making connections, asking the right questions and developing a growth mindset.
One way of facilitating this is to embrace the new tools which are now at our disposal. They are easy to use and with the right strategies to contain any errors or hallucinations that these AI tools sometimes produce, they can be a great learning partner. Also, as the investor Warren Buffett says - they're not going to be uninvented. So it's time to learn more about them.
Self-directed learning, whilst it feels like a simple concept, requires a certain amount of unlearning. It can be realised much more effectively when the learner is given a framework and a set of techniques. One of these mental models is an awareness of how to develop your 'frame of reference' for a given subject.
"There is no such thing as absolute motion: motion is relative to something else." (Galileo Galilei).
In a school setting, learning is divided up into subjects, discreet packages of information. It is easy to be passive, learn what is needed, and then forget it. The connections between subjects are often not made.
To build up a frame of reference for a subject is to enter into it and to begin asking questions. These answers then spark more questions, and your knowledge of the subject starts to spread and grow. This growth, however, is interconnected and led by the learner, and so it means more to them. For a visual representation, it can be a like creating a mind map, with new nodes being added, connections becoming apparent.
One can begin learning about any subject, at any level - with one caveat - one needs to find learning materials which enable us to do this. These can be surprisingly hard to find at times. I believe that this is one of the main reasons adult learners sometimes don't fully embrace lifelong learning.
There is a strong emotional component to learning. We remember different models of teachers and teaching, which affected us both positively and negatively. As an adult, we can sometimes hold back from asking questions if we feel they are too simple, or that they are things that we feel we should know already. We don't want to look stupid by asking a question which we feel that everyone else in the room knows. (Even if quite often, other people don't know the answer, either.)
We might have certain self-limiting behaviours. A common one is 'I'm not good at technology'. People who say this often have had experiences with technology where they feel out of their depth. They have stopped looking for an entry point into the subject. They use the technique of proclaiming their ineptitude with technology as a way to get out of engaging with it.
There might be a new and large register of language that accompanies a subject. There will also be people who are already fluent in this register and who express this knowledge in different ways, some positively and some negatively. Knowledge is power, after all, and groups of people can use this knowledge in order to define themselves as a group, in order to show mastery of a subject, or in order to show their status i.e. I am an expert in this language and you are not. I am higher status than you. It is no surprise that learners don't want to show vulnerabilities in such a setting.
Unlocking these things can be crucial in the journey towards self-directed learning and in developing one's skills. To help people say 'I don't know this subject just now, but I know how to go about learning something and I'll apply the same techniques here. I'll also leverage other peoples' knowledge, asking the right questions to people who have done this journey before.'
To borrow a phrase from businessperson and YouTuber Codie Sanchez - you can borrow someone else's 10,000 hours.
How to do these things? Use A.I.
A.I is considered to be a general technology as it has the potential to be used in a wide range of applications. Electricity is an example of an earlier general technology.
The reaction to A.I in an educational setting has generally been to consider it a disruptive force. It is indeed disruptive. A study by the software company 'Turnitin' stated that 35% of students used A.I to write essays.
It disrupts the type of learning that schools and Universities have specialised in for a long time - which is learning facts and writing essays about them.
The question for us as adult learners is, how do we leverage this tool to really fast-track our learning? Here I will give an example of how to do that, in the context of what I have talked about above - building a frame of reference, entering into a subject at the right point and keeping the emotional component of learning as a positive.
Let's take the subject of Machine Learning. Sounds complicated. This is what Wikipedia says about it. Spoiler - I don't really have a better understanding of Machine Learning after reading it.
"Machine learning (ML) is an umbrella term for solving problems for which development of algorithms by human programmers would be cost-prohibitive, and instead the problems are solved by helping machines "discover" their "own" algorithms,[1] without needing to be explicitly told what to do by any human-developed algorithms.[2] Recently, generative artificial neural networks have been able to surpass results of many previous approaches."
This is what goes through my head after reading this:
"This is pretty dense. I don't really understand what Machine Learning is after reading this. They mention the cost of it near the start, so that must be important. But why? What exactly are neural networks? What does it do? How does it work? I still don't understand it. Wikipedia is supposed to be easy. This is just too difficult for me. Maybe I'm not clever enough to understand it. I think I'll pass."
Let's look at how to use Google Bard in order to get a better handle on this. I will use the following prompt:
"Explain Machine Learning to a ten-year-old."
Google Bard's response:
"Imagine you have a box of toys, and you want to teach a computer to identify the different types of toys. You could start by showing the computer pictures of different toys, and telling it which type of toy each picture is. The computer would then learn to identify the different types of toys based on the features of the pictures.
Once the computer has learned to identify the different types of toys, you can start to test it. You can show it new pictures of toys, and see if it can correctly identify them. If the computer makes a mistake, you can correct it, and the computer will learn from its mistake.
Over time, the computer will become better and better at identifying toys.
Machine learning works in a similar way. Once the machine learning algorithm has learned from the data, it can start to make predictions. For example, you could show the machine learning algorithm a new picture of a toy, and it would be able to predict which type of toy it is."
So now we have a clearer picture of what Machine Learning actually is. Now we can start to develop our frame of reference by asking questions. (Google Bard has some very handy buttons in the 'modify response' menu including 'shorter' and 'simpler'.) Here are some example prompts:
Explain what an algorithm is in simple language.
Tell me the names of three magazine articles on the subject for beginners.
Who are the people writing online about this subject?
What kind of jobs can I get related to Machine Learning?
Explain what a neural network is to a ten-year-old.
How can I leverage Google Bard to learn more about this subject?
Using this tool helps us exponentially develop our knowledge of a subject, it points us to relevant and appropriate resources. It bypasses any negative emotional inheritance the learner may struggle with, and it can help with any subject.
Yes, A.I can throw up untrue and bad inaccurate answers. However, you're using it to find leads, to find resources. You're using it to build the framework and to discovers what the right questions to ask are. This is just the first step in the process for you. It's the map - it's not the actual landscape.
Thanks for reading.
P.S For an example of best-case use of A.I in an education setting, I would look at how Khan Academy is developing an A.I tutor to help learners get more from their teaching resources.
Here is a video of Sal Khan talking about his vision for how to leverage A.I in education on this website page.