I had cause to go to an Apple Store today. As my 10-year old iDevice could no longer be fixed, I decided to look at a potential new purchase. A genial Genius was helping me. While noting that I was likely not a gamer, he added that as a tech savvy person, I may want to use the new AI tools, and for that one model here may be better than the other. He then told me he was using a Generative AI tool to build an advertising pitch for an agency he wanted to go work for. “What tool is that?”, I asked. “Uh. I cannot recall. It is saved on my iPad at home”, he said.
My surprise might have shown from behind my FFP2 mask because he then went on to say, “I feel we have lost all ability to reflect, connect, remember as there is so much going on around us all the time”. I nodded in empathy, having just told him in excruciating detail about graphics controllers that we used to sell in the 1990s. That was how we got talking about gaming in the first place, and discussing how Nvidia’s winning role in Generative AI was all owed to gamers.
The whole exchange reminded me of a discussion from many years ago, when commenting on blogs and writing point and counterpoint on our own blogs was how we expanded our community and sharpened our thinking. Ben Casnocha, founder and brilliant writer, wrote how an unusual memory had a limited role in overall success, citing tools and devices available to recall things.
As the possessor of a reasonably good memory, sometimes inconvenient as it recalls verbatim things from years ago, I obviously disagreed with him. The example I used was Bill Clinton’s though. I have learnt a tough lesson in life. A famous person’s ability to recall things about someone makes that someone a swooning fan of that famous person; but a less famous person being able to recall personal crap someone may have said aloud in a business meeting is seen as an oddball. So now I keep my counsel and do not often let people know I remember all the information they oozed the last time we met. But I digress.
It is indeed useful to be able to recall things. It is not even necessary to have made a concerted effort to commit things to memory. That is not how memory works. We have mnemonic hooks on which things may hang; or we may have associations with certain words and names; and then there is the Baader-Meinhof phenomenon which can serve as a reinforcer for a new word or a new concept we may have come across recently.
Over the years, my view has not changed. If anything, it has become stronger.
In the 1990s when Altavista was the height of sophistication in search, I was a super competent “prompt engineer”, a term that has returned into common use. So much so my then-boss gave me a moniker he still uses for me. Whatever he asked for, I could find it for him from a poorly indexed, not-yet-crowded web. The skill had helped me find a community of expats in the country where I was sent as a country manager. And it continued to help us as we sought business partners, looked up new products, and found people we wanted to reach.
With many search tools available now, it is perhaps simpler to undertake an exercise in finding things. But a starting point is still needed. That intention, that starting block, can shape the course of the search — including when hyperlinks take you down a rabbit warren, it can help you get back to the original search intent.
In the 16 years since writing that post I have also had much experience of teaching undergraduate and graduate students. A rough estimate says I may have taught about 800-900 students as a visiting professor.
Through this experience I observed a phenomenon.
With the ability to search so much online, things can be found — and forgotten — rapidly. I call this transient appearance of knowledge “Google Intelligence”. I found its prevalence rampant and growing.
I teach inductively. My classes are noisy and engage in Socratic discussions and arguments. With case studies, I find students furiously Googling to stay ahead of the discussion, forgetting that we are discussing something that happened in the past not something that is happening now. But with stories, data, and phenomena from the news or history alike that I bring to the classroom, I also find students asking me, “But how do you remember all this?”
I really have no answer beyond saying, “I am interested in the world and how it works”.
That is only a partial answer.
More terrifyingly I find that factual gunk — we used to call it general knowledge when we were kids — is now not common as people do not bother with knowing or recalling facts when things can be trivially Googled. I teach decision making and critical thinking, and inductive teaching using facts and data seems to be better than teaching with dry, theoretical pathways meandering through what is and should be a practical discipline. Absent facts and data, that can be a tough ask.
I recall one morning when I was visiting a young friend. She said she was intending to write today. I said I would look forward to reading when she was ready to share, and went to do something. When I emerged 2 hours later, my young friend was still sitting in front of her laptop, with a blank Word document, and looking in the distance. “What did you write?”, I asked. With a sigh, she said that she realised she has to look up everything she wanted to write about. “Not like you. I cannot remember what I read in the Financial Times today.”, she said.
Now in my peer group, it is a minor joke, growing both in frequency and in terror, that we find in the midst of a conversation that we cannot recall words or names or entire episodes of life. But we are at least 15-20 years older than my young friend and just lamenting the ravages of life.
My young friend’s comment took me right back to my classrooms.
Something has happened in the last 2-3 decades and it is not comforting.
It seems we are unable to remember much, process anything except in the real time when it was looked up on Google (other search engines are available but stay with me here).
My young friend cannot remember or recall stuff not because she is not interested. Au contraire, she is curious and interested in everything. But somehow she just cannot recall anything she may have read just two hours prior.
Has having the world on our fingertips really destroyed our ability to appreciate prior art?
If there is no appreciation of what-is, how do we teach and learn why-it-is or how-to-change-it, or address any number of related curiosities?
As I look forward to teaching again in a few weeks, I am staring at Generative AI tools. And pondering how to design assessments that can truly assess what the students have learnt rather than how well some can use ChatGPT or any other tools. Going back to paper and pen, and to proctored exams, as has been suggested to me, seems like a regressive step, and unnecessary just because I, as their teacher, have not solved the teaching and assessment puzzle to my satisfaction.
Some like Ethan Mollick, a far more experienced teacher and researcher than I, are brilliantly experimenting with embedding Generative AI in their teaching and their classrooms. I am not sure yet of the experimentation possibilities in my class. The questions are many: Redesign the course? Rejig the teaching? Flip the classroom? Grade-as-we-go? Make my course pass/ fail? All of these? And many others.
Within my lifetime I have witnessed the arc go from memory, to why-memorise, to prompt engineering, to “Google intelligence”, to “Generative AI will serve up your assignment if you can only stretch yourself to typing your assignment question into that prompt box”.
Back to the Apple Store.
The Genius, who was helping me, could not remember the name of the tool he was using to make real something he said was his life’s dream.
I came away consoling myself that at least he remembers his dream.
That is a start.
That is what prompt engineering needs: a start.
And to start we need to know something.
So we can recall it and prompt the Generative AI tool whose name we may or may not remember.
(Disclaimer: These are my own views and do not reflect the views of the boards of JP Morgan US Smaller Co.s Investment Trust or Temple Bar Investment Trust or London Metropolitan University or Harmony Energy Income Trust, or Witan, where I serve as a non-exec director, and chair various committees at the time of writing.)