Let People Like Things. (Even AI)
I always remember being confused that when a particular band got popular, the kids who were into them before anyone else would completely lose interest - in the music that they had spent the best part of a last term hyping up to everyone else.
I now understand this to be “hipster syndrome” or the “snob effect”, where the demand for a “product” decreases as more people adopt it, specifically because widespread popularity makes it feel less exclusive or special.
The key there is “makes it feel”. The guitar riffs of that now established band haven’t changed, they still contain the unique code to unlock something inside of us that we never knew was there. That 5 minutes in the cold plunge hasn’t diminished in terms of health benefits just because there’s now 1000s of TikTok videos of half-naked middle-aged men doing it too. But it can feel like it has - because now it’s not your secret cool new toy.
In a similar league - if not the same sport - is “that guy in the comments section”, the pre-emptive cynic who acts like they’re too cool to care… until it fails and they pile into show how they called it all along.
Just because something has been hyped doesn’t mean it’s over-hyped. The Apple TV show Slow Horses was raved about by many and when I finally watched it, I wasn’t underwhelmed – though in hindsight I shouldn’t have been surprised with Gary Oldman attached. Or the smartphone, which - granted, we are at the point where the yearly updates really need to be scrutinised to show tangible improvements. But! back in ’07 the smartphone was hyped to an all-time high for the iPhone, and it certainly delivered on that.
So what of the naysayers? Well, many will be familiar with the Gartner Hype Cycle, which tracks emerging technologies across 5 stages, peaks and troughs along the journey to success – or sometimes collapse. For those who aren’t, here’s the gist;
“Innovation Trigger” : a breakthrough idea starts a buzz.
“Peak of Inflated Expectations” : Hype explodes. It’s “the next big thing”
“Trough of Disillusionment” : Reality underwhelms. Critics circle.
“Slope of Enlightenment” : Quiet progress. Real use cases emerge.
“Plateau of Productivity” : The technology matures. It just works.
Those who work at the cutting edge of technology will be all too aware of this model and how these phases are reflected across the news and social media whenever a new technology comes on the scene. AI is no different of course, just hit F5 on LinkedIn - if your feed is anything like mine you’ll see at least half a dozen posts about how AI is either “revolutionising how [insert job of their target audience] will be done” / “going to destroy [insert their target audience’s industry] – are you ready? [Click here for spam]” / “just a set of overhyped prediction systems that will never be able to take over [insert person’s professional specialism here]”. What I’m struggling with is why, given we know the hype cycle and that we are on a journey with all of this, is it not enough just to be pretty cool?!
I’ve seen many takes on AI, specifically in the agentic age that seem to miss what I think is the main point. There are too many “in-the-weeds” looking at e.g. vibe coding and pointing out how there are “security concerns”, that the software it produces will be “endlessly buggy” and one that even suggested the inevitable failure of vibe coding will lead to them having a lot more work so sarcastically cheering it on.
I have listened to similar “in-the-weeds” takes before when discussing automation software, from those it stood to directly benefit - “but it doesn’t do this?” or “but we do this a particular way, and it cannot handle that”. For these instances, just as with AI and its capabilities today I say the same thing - “What is your way of solving that problem today?” [pause for response] “keep doing that then.” These systems shouldn’t be thrust upon us or marketed as doing 100% of everything - that will never be the case, especially given the unique approaches to business and life we all have - but if it can deliver 40%, 50% or even 80% of the output that would traditionally have been done manually - over the course of hours rather than days, surely we can agree it’s pretty awesome to have this new toolkit with which to help us streamline our day to day?
As with automation, AI is really good at providing solutions to repeatable, time-consuming tasks. And if a task is repeatable and time consuming it’s probably something that should be looking to automate, freeing up our time to concentrate on those tougher, more mentally stimulating tasks - whether it’s engineering a particularly tricky risk problem or crafting a perfectly worded press release, or planning a new product launch.
I think it’s pretty cool and I’ve been genuinely impressed by what I’ve already been able to build with it. We don’t have to pretend everything is revolutionary, but we also don’t need to downplay how genuinely useful, exciting, and empowering this new toolkit can be.
