A recent UK TV documentary, ‘Swiped – the school that banned smartphones’, prompted me to revisit a blog I wrote some years ago called ‘Will IT destroy us?’.
Memory said I’d penned it four years ago. It was actually six. How time flies. In the blog, I posited a number of just some of the downsides of our connected world, but highlighted two areas – Artificial Intelligence and teenage screen addiction. Now I’m re-examining them from the standpoint of their adverse effect on writing and literacy.
At that point, AI was still a nascent technology: now, it seems everything comes with built-in AI. Many of those uses are for the good of mankind, such as reducing animal testing in drug development; many aren’t. AI is capable of writing text such that even experts are unable to distinguish fake from fact. While the dangers of war from bomb and bullet are obvious, far more insidious are the ceaseless wars of disinformation from hostile nations. The elections in the former Soviet country of Georgia were recently annulled because of Russian interference by internet.
But what about us as writers? Publishers are driven to ask if any of the material in a manuscript is AI-generated. Taking matters to reductio ad absurdum, what is to stop an ‘author’ writing a novel entirely by AI, it being accepted by a publisher and becoming a bestseller? It could happen and if it does, what does that augur for writing as a creative art?
When I wrote my blog in 2018, some people and institutions were recognising the dangers of kids almost glued to their screens. French schools banned smartphones and one UK school did likewise. Governments worldwide have, as usual, been slow to react, but all of a sudden, there is a growing movement to protect the young from themselves. Australia is debating banning social media for under-16s and others are likely to follow suit.
‘Swiped – the school that banned smartphones’ was created by celebrities Emma and Matt Willis, who were concerned that they were losing their three children to their phones.
A class of 12-13 year olds, plus some of their teachers, agreed to lock their smartphones in a glass case for three weeks. Daily diaries were recorded on videocams and the children were monitored both physically and mentally by Cambridge University.
While usage varied, some of the class admitted to being phonebound for up 9 hours a day. In the beginning, several believed that life as they knew it would be destroyed.
One of those was Harry. But by the end of the 21-day experiment, he said he was ‘more interactive’ and it had made him realise ‘I’ve got to be careful what I’m looking at online’. Another was Scarlett. ‘When I had no phone I was completely fine – I concentrated in lessons, I understood, I was more sociable, I was kind to people, I came downstairs and helped my family… we were playing board games every night.’
Head teacher John Player said: ‘There was a 17% drop in anxiety symptoms in students. The impact was really quick’.
‘They slept on average an extra hour a day, which is going to have a huge impact on mental health.
‘There was an 18% reduction in symptoms relating to depression and a 3% improvement in working memory in just 21 days.’
These are both disturbing and encouraging figures, but I think there is an added and distinct bonus. Kids on phones use textspeak, a bastardised form of English, which has already resulted in dreadful grammar and atrocious spelling. Screen addiction leaves no time to read. If we can restrict both phone and screen usage and encourage reading habits, perhaps we can rise from the illiteracy doldrum that has plagued us for decades.
Free image courtesy of Pixabay
Training the Large Language Model which underpins that beast known, ironically, I beleive, as Artificial Intelligence, used the electricity supply of a small country to create. It’s deployment is equally energy-intensive. This situation, all other AI issues aside, is a little reminiscent of Easter Island, where the islanders cut down their last trees, needed to make boats to fish for their only diet, in order to make totems to their diety.
In the fight between Big Tech and the governments of the world, Big Tech is hardly losing, and have not been required to tag AI-generated content as such. Not only this does this mean that it will become hard to prove which work is AI and which isn’t, it resides alongside humans’ information in that grand repository of the internet, indistinguishable. Two adverse outcomes will follow. The first is that AI will re-scrape its own data, and beleive all the semi-truth it disgorged in the first place, reinforcing it. The second is that it will dilute, distort and corrupt the store of data which underpins civilisation.
Already on social media, one can no longer admire a photograph of natural beauty without wondering if it is real or AI. Poor language skills at present mean that instant journalism using AI to write erroneous ‘explanatory’ captions to old images can be spotted, but in future, will they be distinguishable?
I make a point of calling out any unacknowledged AI text or image on social media, but it is a drop in the ocean.
I’m considerably less worried about Artificial “Intelligence” than I am about the concentration of wealth and power in the hands of the few tech companies (and their owners) that are promoting it. AI is not intelligent and is certainly not sentient. When we achieve artificial sentience – assuming we can do so before we wipe humanity off the face of the earth – that will be something to worry about. (Although we probably won’t have the time to worry at that point.)
Current computing capacity allows for so many positive things. Anything that calls for counting or pattern spotting, which has made, for example, identifying cancerous cells in tissue samples so much quicker and more reliable. But it’s a tool, like a knife, and like a knife it can be used for ill as much as good.
(I’m stealing this comparison from Isaac Asimov who was writing about this issue in the 1960s. Then the fear was over robots. The same fear, just applied to a different, earlier technology.)
The gun lobby in the USA says “Guns don’t kill people, people do.” This is pretty stupid, since all guns are designed to kill, and the sort of guns the gun lobby wants to protect are *actually* designed to kill people. But there is a point to be made that it does take the will of a person – and usually a trigger finger – to shoot a gun to kill someone.
The argument works better with a knife, which isn’t designed specifically as a weapon.
Even more so with AI. It’s the benevolence, malevolence or utter stupidity of the people using the thing, writing the code, giving the instructions, that makes it bad or good. Which brings me back to the tech companies and their obscenely wealthy owners. I don’t want people like that in charge of such computing power. I’m not sure I want states in control of it either, but rather a democratically elected government than a few oligarchs.
Not that I see any democratically elected governments doing a great deal to control AI. Rather they are cheering it on thinking it will make them money and bring them power. And now China has entered the arena with its own DeepSeek. Whoopee!
(By the way, I keep wanting to call it DeepThought, which was the AI with a John-the-Baptist-complex in *The Hitchhiker’s Guide to the Galaxy*.)
Oh, my! Is this a timely post. The current situation in Russia, the US and China where tech. is being used to rule and delete opposition is truly terrifying.
BTW Nigel the schools that my sons went to in Wales, banned mobiles in school back in the early 80s onwards. My best cello pupils have come from households who didn’t have a TV or severely limited the time that children were allowed to watch it. The only redeeming thing about TV now is that the plethora of channels and economic concerns mean that most programmes now are repeats.Weirdly the exception is sports, although as my SO says, that’s all the same each time anyway. 😉