For the past three years, since AI became available free (more or less) to everyone with a phone or a computer, it’s been hard to go an hour without it being mentioned in the news or everyday conversation. AI has been the fastest-developing technology I’ve known in 35 years of writing about the latest next big things.
So is it something we seniors/oldsters/Boomers should ignore, like getting your news from TikTok or putting your preferred pronouns at the bottom of an email?
Or should we get stuck in and use it multiple times a day as a kind of outboard brain, as I’ve come to do? How can we benefit from AI, even enjoy it, rather than just tolerate it and live in hope that the machines don’t rise up against us?
The best approach is to regard AI as a vastly cleverer, more flexible and intuitive form of Google, which you can ask complex, multi-part questions.
Or as an infinitely knowledgeable and patient colleague who never minds you asking them how to do this or that.
Let me explain how AI might help you by recounting four exceptionally useful or intriguing things it has done for me in the past few weeks – things for which traditional Google, immensely powerful as it is, would have been a bit useless. (As it happens, Google has just been updated to include an AI mode.)
We were on holiday in Canada and were looking to find some interesting things to do on our last day that, crucially, were less than an hour’s drive north from Calgary, the airport from which we were flying home.
Google was hopeless, suggesting places that were up to four hours north, south, east and west of Calgary. I suspect they were all advertisers because, as people are apt to forget, Google is really an advertising company. ChatGPT, however, took a fraction of a second to send my phone a set of attractions and a fantastic overnight-stay suggestion that were the making of an already terrific trip.
A couple of weeks earlier, we had been staying at a wonderful glamping site in the mountains. The only disappointment was that the view from our deluxe tent, of a distant blue lake and snowy peaks, was blocked by a stand of silver birch trees.
For fun I took a photograph of the obstructed view and asked AI to remove the trees.
It obediently did so, inventing a pretty but wholly fictional landscape it guessed might have been hidden. The simulated photo was undetectable as a fake. Creepy, I know, but it fascinated me.
On the plane home I (finally) wrote my speech for my daughter’s upcoming wedding. I had been avoiding this task all holiday. At one point, I used the word ‘tautology’ without really knowing what it was. I later asked AI to check the paragraph and learned that what I’d cited was definitely not tautology. Embarrassment averted.
Since then I’ve had a long late-night ‘conversation’ with AI about extremes of scale, from subatomic particles to the size of the universe. I had started by wondering how many atoms there are in a Rice Krispie, thinking that it was as many as a billion. It turned out it was a quintillion, which is a trillion trillion. ‘We’ also ‘discussed’ the vastness of the universe, and it was – honestly – like an unbelievably geeky conversation with an unimaginably brilliant superhuman mate down the pub. I opted to have the conversation by voice rather than text, as you can. It was my most intense, awe-inspiring, scary experience yet with AI.
How should you get into this absorbing/terrifying brave new world? Notwithstanding my friend-like science chat with AI, it’s unlikely that you will use it in the same way as the astonishing 70 per cent of American teenagers who rely on it for companionship, one-third of whom say that conversations with AI are ‘as satisfying [as] or more satisfying’ than talking to real friends.
But experimenting with AI is fascinating, fun and nothing to be uneasy about. What’s more, while you can pay for these AI services, they all offer a large amount of free use – most can be consulted by logging into their respective websites on a computer or through a free app that can be downloaded to your phone.
Ask AI a question – ChatGPT, Google Gemini and Microsoft Copilot are three of the most popular – then play around. It won’t bite. There’s much mystery about asking AI a question or ‘prompt’, but just go ahead and ask what you want. If AI misunderstands, you can always ask the same thing a different way – it is infinitely patient. When you’re more confident, get it to design a piece of art, compose a song, give financial and emotional advice, write an awkward email, or almost anything else.
Just make sure you double or treble fact-check any specific tips it shares before acting on them. AI does make mistakes. A French-language
AI agent, Lucie, was launched earlier in the year and retracted for getting things hilariously wrong. It told users that King Herod ‘played an important role in the development of the atomic bomb’ and that cows’ eggs are ‘considered to be a healthy and nourishing food source’.
How does a superhuman entity get something so wrong? It depends on what information the AI has absorbed. Remember the old computing acronym from 40 years ago, ‘Gigo’, which stood for ‘garbage in, garbage out’? Advanced these things may be, but they are still mindless robots. So if they’re told rubbish, they will spout it back.
AI chatbot Claude has the clearest explanation I’ve found of how AI works: ‘Think of it like learning a language by reading millions of books – I’ve absorbed patterns of how words relate to each other. When you ask a question, I predict what a helpful response would look like based on all the patterns I’ve learnt.’
One last word – this from ChatGPT.
Do we humans, I asked it, not have an insurmountable advantage over AI because we have bodies that can physically pull the plug and disconnect a computer if it goes rogue?
‘Yes, we have a plug,’ it replied. ‘But having a plug doesn’t matter if no one is willing or able to pull it – or if the plug becomes too entangled in everything else needed to survive.’