“Ask Gemini. Or ChatGPT.” That is a frequent response from at least one of my children when I admit that – shock, horror – I don’t know the answer to something.
If the all-knowing parent is found wanting, ask chatbots such as Gemini. It has all the knowledge of Google, but it can also hold a conversation that you can follow.
The youngest members of the family have grown up with the belief that technology can do almost anything. Turn on the TV? Sorted. Set a timer to count down their games time? Absolutely.
Turn on and off lights at a voice command, send the robot vacuum to deal with the crumbs, or play endless KPop Demon Hunters tracks? You get the idea. If you can’t ask Google, Siri or Alexa to do it, can it even be done?
READ MORE
Well yes, actually.
Because as they have learned over the past few years, digital helpers might not always get it right. Or even do what you ask at all. Like the time Google simply refused to turn on the smart plug requested – to the disbelief of the eight-year-old.
Or when Siri decided that instead of KPop Demon Hunters, it would only hear the “demon” part of that request, with unpredictable results. And that is just the bog-standard assistants, not the generative AI type of helpers.
They have their own issues. Believe half of what you see, Marvin Gaye once sang, and none of what you hear. You wonder what he would make of the new generation of artificial intelligence (AI).
AI gives you overviews of search results and email summaries that occasionally need to be taken with a pinch of salt. It features an eagerness to please that sometimes results in it simply making things up. It is a work in progress, but despite these efforts, AI still needs to have its work checked on a regular basis.
It was sobering, therefore, when research from CyberSafeKids revealed that more than one in four under-12s were interacting with chatbots regularly – and that I was, inadvertently, contributing to that statistic.
That figure was higher for 12- to 15-year-olds, at 36 per cent. Both of these figures were significantly higher than the 5 per cent and 12 per cent respectively that were recorded last year.
According to the latest report, some 9 per cent of under-12s are even using tools such as ChatGPT for their schoolwork, while 8 per cent just felt like a chat. I can’t be the only parent who would be fine with Google or OpenAI listening to the minute details of a Minecraft build.
The report didn’t specify whether parents were part of the equation; in our house, I am the sole conduit to ChatGPT and Gemini, so there is no sneaky access. But that may change in the coming years as the technology is integrated into more and more devices. Gemini has already replaced Google Assistant on Android phones; other devices must surely be next.
Why are some independent Irish breweries turning off their taps and closing?
This leaves parents with a dilemma. How do we help our children to navigate these new technologies without leaving them exposed to the worst of it?
Banning children from using them – a one-size-fits-all solution that seems to be the go-to answer when the ugly side of technology raises it head – isn’t realistic as a long-term fix, however tempting that may be.
Then there is the impact on our creativity, with AI able to produce everything from poetry to music with just a few prompts. Why bother to make content of your own when AI can do it for you?
Far better to teach our children about evaluating and checking information, rather than taking everything at face value. At some point, children will be exposed to these technologies, and teaching them about the potential flaws is just as important as pointing out their advantages.
CyberSafeKids espouses the idea that “same rules apply” as regards online and offline safety, and it is a good one to follow. In the real world, if your friend insisted that men had not, in fact, landed on the moon, but instead filmed it on a Hollywood studio, you might at least try to verify that claim.
You wouldn’t expect your child to ride a bike without first teaching them how to use it safely. You wouldn’t hand them a pair of skates and expect them to get on with it. Nor would you allow them to use so much as a toaster without at least explaining the risks.
So why did 18 per cent of children under 12 say they have never had any discussion about online safety with their parents? And 48 per cent of 12- to 15-year-olds haven’t had that conversation in a year – at a time when technology is changing at breakneck speed.
[ Children increasingly interacting with chatbots, survey findsOpens in new window ]
There are many reasons to be cautious of generative AI. Consider the amount of data it is hoovering up, in search of more intelligence, better responses, more life-like interactions.
Then there is the impact on our creativity, with AI able to produce everything from poetry to music with just a few prompts. Why bother to make content of your own when AI can do it for you? (Plenty of reasons, including that AI needs a steady supply of new data to keep training and improving models, or it will be left with its own output, and that leads to model collapse.)
Personally, I am wary about raising children who become too used to ordering a digital assistant around, without at least saying please and thank you.
AI is carving out a place for itself, and it has its uses. Just remember: it doesn’t always get it right. And regardless of whether it costs a few more seconds of processing time, manners are still a good thing, even in an era of AI.
