Is using artificial intelligence cheating? If there is a genuine need at work to make something conceptual for a client, do you want your staff to struggle for hours, or to spend 45 seconds like I did?
On Saturday evening, we were reminding my daughter about doing her homework. She had to create a haiku in French.
While she and her mum were talking, I pulled out my phone, sneakily logged into the AI chatbot Chat GPT and put that phrase in. Within seconds I had the answer to the homework assignment. With my UGLY pronunciation I rattled it off.
Then I showed it to them in Google translate. My wife speaks decent French. It passed scrutiny.
Then I did the big reveal.
My wife’s comment to our daughter was, “See, that’s what your dad does. He solves the problem instead of doing the problem.”
She loves doing sudoku all the time, and I keep wanting to write a program that solves all sudoku forever – different sort of intellectual work. I’ve already partially designed it, much to her protests that, “that’s not the point!”
Needless to say, we both made sure our daughter did the work, rather than just submitted the ready made answers.
But think about that for a second. If that wasn’t a school and education context where you want to make the kid struggle so they learn, which do we prefer? If there is a genuine need at work to make something conceptual for a client, do you want your staff to struggle for hours, or to spend 45 seconds like I did?
Ok, a French haiku is unlikely to be asked for often, but that’s the subtle importance of what I’m saying. We have designers who design networking, computers, cloud and IT security. But if a project had a requirement for a side item that would be accentuated by a haiku in French – okay that’s straining it a bit – but we could then complete that part without asking around the office for a French speaker who can spend an hour doing poetry.
In other words, in business, it’s a tool that enables us to serve our customers better, faster or with less effort, ultimately driving the price down for them.
But what about education? Kids can just “cheat” like I did. I’ve long considered that the old school model of teaching facts, memorise them, and regurgitate them in an isolation situation is broken and false compared to how people work now. In the age of book knowledge, going to a good school was the only way to get knowledge transferred to you, and your memory was the only storage device. The internet changed that a bit, but the schooling model hasn’t fully changed from talk, memorise and regurgitate.
Remember in my Haiku request, I chose “about food”. That also is a stylistic variable for education – “ok kids, so you can have Chat GPT make anything. So let’s go through and find what topics it can and can’t translate well – food, exams, boyfriends, music. What else? You come up with themes and we then compare the French it makes, and see what are the best and most meaningful responses.”
I don’t know if what I did there is any good for pedagogy, but this has to be better than banning a tech that’s not going away.
I’m not alone in my views on this pedagogy theme. Many leading educators have stressed the need for change, and some forward-looking schools are doing that.
More valued now are:
Skills such as collaboration and team organisation vs doing it in isolation.
Information search, filter and evaluation vs memorise and regurgitate.
Build social standings on networked communities to show your worth, like how many projects you’ve competed on Kaggle, the AI coder site, vs. what university you went to.
When it’s possible to churn out content quickly, we need to change focus like we did with information search and filter. Sure, I can create French haikus, but the assignment should be to have Chat GPT make them, but for you to use your knowledge of French to work out if they are any good, or to make 10 of them and to filter the best, or to make a few and improve them. In other words, we should task children with using it as a tool in the learning, and adapt the French skill to how people will need to use it going forward.
Google translate can translate any page, and soon will be able to do live on-the-fly verbal translation, so for some jobs, why learn a language at all? That’s the question educators need to answer to stay relevant. Then to design education around that answer.
Is that easy? hell no. Educators need to redesign the whole paradigm of their work. It was easier before so the understandable reaction is to wish the genie back in the bottle. But it’s not going away. So we as a society have to do the work, for our kids’ sake.