Disrespecting mules risks more severe consequences than just angering Mother Nature.
What’s cooler? A mule, or a Ferrari?
That’s a loaded question, designed to make a point. The Ferrari is cooler. Ferrari has spent decades dominating a sport they invented and producing wave after wave of spaceships that look like cars that make you tingle in places you thought long dead. They worked really hard at beating mules in any coolness competition.
Even if you hate irresponsible driving and excessive speed, you’d have to have a pretty unique psychological profile to genuinely believe that mules are cooler than Ferraris.
Here’s the same question again, but from a slightly different point of view: what’s better? A tested technology that demonstrated its capacity to deliver at minimal cost, or an exclusive solution that breaks down if the road is not paved, requires high end expertise to operate and is eye-wateringly expensive to both acquire and maintain?
The Ferrari doesn’t look as cool in this second scenario, does it? Thankfully, we’re talking about Ferraris and mules, and not with the world’s recent obsession with generative AI as a spectator sport. If we were, you’d be feeling silly for trying to use a Ferrari at all costs, even to bring medication to those secluded mountain orphans up in those villages where there are no highways or petrol stations.
Reality rarely lends itself to such simple, silly parables. Ferraris have their uses, and so do mules. However, it’s time to recognize the common mistake of choosing flashy. Prestigious technologies over proven, practical solutions.
Fortunately, this isn’t a new problem, and your newfound love for generative AI hasn’t created it. The issue is the brittleness of some technologies, and just how little we bear it in mind when selecting them.
“Brittle” does not define bad technology. Exceptional, modern, buzzworthy technologies are quite often extremely brittle, in spite of being engines for change that transform entire industries.
Brittleness does not preclude a technology from being powerful or performant, but rather indicates that it will only operate well under an extremely specific set of circumstances.
Generative AI is in exceptional company when discussing brittle technologies and design patterns. Kubernetes can manage vast, complex infrastructures but they require meticulous setup and continual maintenance that many teams are not ready or qualified for. REACT is a powerful library for dynamic web pages, but its engineering approach can transform straightforward code into a hell march. Compute calculations are spectacularly effective when dealing with data that lends itself to parallel processing, but can actually harm performance when used on the wrong data, or just wrongly.
The more specialized a technology, the more impressive it is against its specific use case. And the more useless it is outside of it. There’s no magic there, no hidden knowledge.
Deploying brittle technologies in the wrong use case can become an inescapable, extravagantly expensive deal with the devil. In the worst of scenarios, it may bring your entire infrastructure down as you engage in sunken costs fallacies to the very end of your competence and budget.
That a technology is new, and powerful, and Netflix used it to become Netflix in no way means you should automatically adopt it. That OpenAI made a gajillion dollars in funding and everyone is turning themselves into Ghibli characters doesn’t mean you should use ChatGPT for everything.
A Ferrari is supremely cool. It’s just not always better than a mule. One may even make the case that it often, based on the circumstances and project requirements, Italian sports cars are not always the solution.
Writing those words hurts, no matter how true they are.
There is a reason shiny technologies are cool, and there is a reason we respond to them with excitement and enthusiasm. To however make ridiculous plans like firing all your support agents because you have AI now, that may be a step too far. I mean, imagine how silly that would be, right? Klarna?
There is no easy solution to this problem. No Networkati-level wisdom to impart that will guide you on the right way.
To pick the right technology that will serve your needs with great coolness but maximum efficiency is the challenge that has made or broken empires. You shouldn’t just listen to your technicians for this. Specialists are deep in the trenches of their own profession and rarely have the wide view required to help you through this maze.
The real answer isn’t technological, it’s human. True wisdom comes from experience, from overcoming and learning from setbacks. Hold on to your best engineers. They are weird and expensive, yes. They also carry insights deeper than which version of Reach to install.
All else failing, you can hire in a consultant. Not a Spring chicken, but one of the ancient guard, a keeper of cosmic wisdom acquired from the experience of delivering a couple of successes and navigating dozens of failures.
Reach out, we might help put you in touch with someone.
After all, you do care about those mountain orphans, right?