What you need to understand about AI to avoid being left behind
This article, originally published April 2025, has been split up—10,000 words was too much for one sitting. Read this, then dip into the linked pieces that interest you.
As we blaze through the AI hype-cycle, excitement is matched only in the level of sheer panic it has induced. Is it only a matter of time until we're all replaced by robots?
AI-pocalypse?
First let's clarify what we're on about. Artificial Intelligence (AI) is not new: spell check is a rudimentary example and it's been around for donkey's years. Also sat-nav, spam filtering, auto-correct, and more. AI is vital in modern accessibility tools—a point often missed by some when reflexively implementing blanket bans on AI in schools and organizations.
Not for the first time the goal-posts are shifting. And this time something has snapped in our collective consciousness, beginning with the launch of OpenAI's ChatGPT bringing Generative AI (GenAI) and Large Language Models (LLMs) to the masses in 2023.
For some time I've been wrestling on whether this latest wave of AI is a game-changer that will upend our lives—not the first I will have witnessed in my lifetime—or is it all just a parlour trick. But you must judge for yourself.
I am of a generation whose childhood was peppered with requests from adults to help set the VCR1 to record MacGuyver, or Twin Peaks. Guiding our teachers as they struggled in the computer lab. Our elders were grappling with how to send email as my generation entered the workforce with a distinct advantage. I never thought I would see the day my own father would embrace it: kicking and screaming he came to a world where this is now the focus of his working day. And when he's out walking with his dogs at the weekend he might send us a video of an interesting woodpecker he found in the woods.
If this is another step change in this multi-generational tech revolution, then it behooves us all to understand this new paradigm of AI tools. Play with them, get a feeling for how they can be useful, what they're good at, what they're really really bad at too.
Because the greatest danger is not that you will be replaced by a robot. It's an emerging generation, many younger, some your peers, who are rapidly becoming fluent in this new tech already.
What AI can do
This subset of AI has made a dramatic leap forward. These new models are touted to have capabilities in summarizing text, writing prose, creative brainstorming, writing code, translation, and research—but you need to take each with some big caveats.
The art of framing and refining your questions as you interrogate these systems, known as Prompt Engineering, is rapidly becoming an essential skill in the workplace, in business and in creative pursuits.
I've put together a practical guide with hands-on exercises you can try yourself: AI in Practice: A Hands-On Guide.
Even if you find yourself having concerns or disapproving of these techniques, that's all the more reason to understand them. Because they are increasingly being used by others all around you.
Before you dive in please take a look at my safety advice.
Accessibility
I have a personal stake in accessibility tech—I've lived with quadriplegia since 2022, and voice interfaces saved my sanity when I couldn't move. AI has transformed what's possible for people with disabilities, from speech recognition to eye tracking to screen readers.
I've written more about this in AI and Accessibility.
Learning and education
The kneejerk bans of early 2023—New York City reversed theirs within four months, admitting it was "knee-jerk fear"—have given way to more nuanced conversations. Meanwhile, AI offers tremendous promise for learners with disabilities, and it may finally push us to reimagine assessment beyond Victorian-era rote memorization.
I've written more about this in AI and Learning: From Panic to Potential.
The bad
There are the well-documented fails like telling us to make our pizza with glue and put rocks on it.2 This current generation of LLMs were built and trained on the corpus of text on the internet—content creators and copyright be damned. So of course it may be quicker than Google at finding you that answer, but it doesn't have the ability to discern a fact in a Wikipedia article from a Reddit shitpost on the best way to glue your pizza together.
In a world where disinformation is weaponized, automating a pipeline of randomly false facts seems unhelpful, but on a personal level, it has enormous utility. I wouldn't put my name near anything it generates without enormous scrutiny though.
Hallucination
The technical term for AI making things up is hallucination. Often downplayed with the retort "humans aren't 100% accurate either". No, but the difference is the AI doesn't know when it's made something up. It sounds just as confident and can be quite persuasive—if you didn't know, you wouldn't catch it.
When do the clocks change in Ireland?ChatGPT 4o answered:
In Ireland, the clocks change twice annually:
- On the last Sunday of March, clocks go back by one hour at 1:00am, marking the start of Irish Standard Time (IST).
- On the last Sunday in October, clocks go forward by one hour at 2:00 am, marking the start of Irish Summer Time (IST).
This is contrary to the European Union's daylight savings time schedule, which puts the hour forward in March and backward in June, making Ireland a singular exception.
THIS IS THE EXACT OPPOSITE OF THE TRUTH.
I don't know where it got that from, or the idea that Ireland is a "singular exception". Or that Ireland starts its summer in October. Be warned!
Farewell, the French Bulldog
In a previous post—Farewell, the French Bulldog—I looked at research from Oxford and Cambridge showing how LLMs, built on hoovering up the internet, may be staring down the barrel of model collapse if they keep feeding on what is rapidly becoming essentially their own outputs. This generation of LLMs may turn out to be as good as it gets.
The ugly
Dirty laundry
Listening to journalists it seems they lie awake all night in full-blown panic that they, as well as creative writers and eventually the rest of us will be automated out of existence. Meanwhile, a certain brand of entrepreneur and C-suite execs are dreaming of the prospect of transferring work to robots who don't need to take time off or be paid. The likes of Elon Musk3 and Sam Altman4 think so.
There's a certain fetizhization in American business culture epitomized in the backslapping chortles of Musk and Donald Trump in a pre-election X-cast cajoling over how many people they had laid off, like two gym buddies boasting how much iron they lifted. Well that proved prophetic. But fear not, Silicon Valley has the answer: a universal basic income (UBI) of $1,000 per month should address the existential problem of humans, who will as a category become surplus to requirements in this fantasy economy.5
We are losing our minds over this thing. But nothing I've seen so far convinces me we're on the verge of being replaced.
Whether you're a journalist, writer, business analyst, software developer, lawyer, any profession you care to think of, the best thing you can do to get ahead in this new world is to think about how generative AI might be an accelerator in your day-to-day work. Even if you decide it's not for you, at least you'll know what you're turning down. And you'll know what you're up against, as others will certainly be using them.
This time it's different
We're in the middle of a tech revolution that's changing society in unrecognisable ways. We can look nervously to history for guidance at the impact of the printing press or the industrial revolution.
But we've seen change as monumental as AI in our lifetimes—and we handled it.
If this is another step-change, we'll all catch up sooner or later. Maybe you'll be an early adopter and get ahead of the curve, or maybe you'll be like my dad, belatedly, complaining loudly every step of the way. Or maybe I'm wrong, and this is just the latest hype surrounding an elaborate parlour trick. Make your own mind up.
The really ugly
One thing that's changing very rapidly is just how convincing phishing and scams are becoming, as bad actors put this new tech to work.
It's no longer enough to look for spelling mistakes or dodgy logos—some of these scams are already incredibly polished and personalized. It's not just ourselves, we need to look out for our parents and those less savvy, many of whom will be taken in by them. Is your voice out there on social media? Watch out for deep fakes that sound like you or a loved one.6
Whatever you make of it all, I think the best we can do in the 21st century is to stay curious, keep learning, and remain vigilant against those who will always try to take advantage.
Before you dive in
I've put together a short guide on using AI safely—what to watch out for with your data, which services harvest your inputs, and the risks of deploying code you don't understand: AI Safety: A Quick Reference.
Footnotes
A VCR was a technological sarcophagus accepting offerings in the form of large black cassettes—which were a sort of primitive magnetic scroll. Our ancestors would insert these "tapes" and patiently observe illuminated glyphs with names like Play and Rewind. If the device was displeased, it would devour the tape, requiring a ceremonial pencil to salvage the remains of the scroll.
Sam Altman says AI will make most jobs 'useless'—The Guardian