By Matt Jones, Founder & M.D. – Oxbridge
When the average person thinks about artificial intelligence (A.I.), they probably imagine something in between Amazon’s Alexa and the character; Data from Star Trek. If you’re not familiar, Data is the stereotypical personification of A.I. (a human form, has no feelings and is highly intelligent).
A.I. is abundant today, used for many tasks such as recommending what you should buy online, understanding what you are saying to virtual assistants, such as Alexa, to recognise who or what is in a photo, or detect credit card misuse.
What Is A.I.?
The current definitions are pretty broad, and there’s no widely accepted version, but the things described as A.I. aren’t necessarily as they seem.
A lot of systems labelled as A.I. at the moment are little more than a decision engine, e.g., if this happens, then do that, and so on, or if the live chat user types in “help”, then ask them what they need help with, and you know how frustrating that can be as a user. This is not A.I., it’s just programmed responses, and we’ve been able to do that for over 40 years; it just so happens it’s now become cool.
An A.I. researcher at Google, Francois Chollet, has said that; “intelligence is tied to a system’s ability to adapt and improvise in a new environment, to generalise its knowledge and apply it to unfamiliar scenarios.”
This is a fitting description of what I believe we’re striving for, but clearly, this is where the overwhelming majority of things we call A.I. currently fail to achieve. They are usually good (often superhuman) at completing a particular task. For example, ask an A.I. designed to recognise photos of chairs, to suggest your next purchase from your favourite online shop, and it’s going to have a problem.
How Does Technology Fit In?
We’ve seen a vast (albeit forced) leap in adopting learning technology during the COVID-19 pandemic, and it would have been difficult for all of you reading this to have avoided interactions with some level of A.I. during this time
Hopefully, some of this was online learning, but even more likely Zoom or Teams meetings; undoubtedly, the unsung heroes of lockdowns, everywhere. Zoom has the facility that can automatically transcribe meeting notes based on detection of natural language in a meeting, using A.I. If you’ve ever used a virtual background on one of these services, it uses A.I. for that too.
What’s clever about this is that it wasn’t all “programmed” by developers. The system was “trained” using lots of data. Essential, what this means is that we can teach the A.I. by feeding it with information rather than instruct it.
Use Cases In Learning
What’s exciting for me is what A.I. can mean for learning. I imagine a world where online learning (not so much in the classroom as I still believe that they are invaluable for youngsters) adapts to an individual’s ability, learning style and pace, serving up content ad-hoc as and when it is needed.
If you’re the kind of learner that needs lots of visuals, then great, let’s serve you that. If you’re a reader, then you should be served materials to read. These kinds of techniques mean that my organisation, Oxbridge, enjoys pass rates much higher than average.
I am developing A.I. within Oxbridge right now to help support and augment our learner experiences. I have no intention of replacing our highly skilled and capable tutors, teachers and faculty staff, but we will use this technology to improve and increase student interactions, not reduce them.
As a society, we’re producing and capturing an incredible amount of data about ourselves. Most of which has a story to tell. A.I. is a great tool to make sense of patterns that humans may either not spot or take too long to see. I’m excited about what the next ten years hold and expect some exciting accomplishments, both in the field of A.I. itself and its practical applications in other industries.
Should Teachers Be Worried?
The evolution of learning has been slow, even static, for many years, and technology will undoubtedly change that rapidly. However, the simple answer is no because the field isn’t advanced enough, especially not in the mainstream, and it’s probably good news for teachers, as I believe A.I. will be a force for good, by lightening the load for educators.
Let’s touch on how narrow the scope of an A.I. usually is. I have a Robot vacuum, and this little guy is impressive; every day, he goes out and keeps an entire floor of my house spotless. He can navigate in and around obstacles, empty himself and even vacuum beautiful lines in the carpet.
However, once in a while, my cats will leave me a little gift in the middle of the floor. My vacuum will spend the next few hours spreading this gift everywhere because his scope is very narrow; he’s excellent at vacuuming but is unaware of the existence of my cats and their perpetual torment.
Will this change? Probably, at some point, but we’re some way off from this yet.
There are also strong arguments to say that A.I. will create more jobs, not less. What’s more likely is that we may help teachers to regain some of their dwindling free time. So don’t worry, cyborgs won’t be looking after your children any time soon.