There’s a strange thing happening with AI.

Not long ago, businesses were being told they needed to adopt AI immediately or risk being left behind. Now, the pendulum seems to be swinging hard in the opposite direction.

We’re noticing it with clients. We’re hearing it from colleagues at networking events. There’s a growing distaste for businesses using AI in service delivery, marketing and content creation. In some circles, simply using AI is enough to cast doubt on the quality, credibility or even ethics of the person behind the work.

That’s quite a leap.

Perhaps a more useful question than “Did you use AI?” is “Did you use it intentionally?”

Boycotting the tool may be just as short-sighted as blindly embracing it, and we see too many businesses jumping from one to the other.

1. Has the AI pendulum swung too far?

There is plenty of terrible AI-generated content out there. We’ve all encountered the 800-word LinkedIn post that somehow manages to say absolutely nothing. The weirdly glossy, hyper-detailed images or repetitive, slightly uncanny video clips. AI slop is real.

Our reaction to it is far more interesting.

LinkedIn is experimenting with ways for users to flag content they believe is low-quality or AI-generated. Yet AI detection itself is notoriously imperfect. Human-written content can be identified as AI-generated simply because it contains patterns we’ve come to associate with machines. 

Shoutout to all the skilled writers who have been using the en dash in their writing for the longest time, and had to stop using it: these days it is considered the quintessential mark of AI-generated slop. Goodbye, respected punctuation mark that allowed us to take a well-deserved pause and breathe.

AI learned from human language. Now humans have spent months reading AI-generated language. Naturally, phrasing, idioms and patterns will travel in both directions.

We risk reaching a point where sounding vaguely like AI is enough to put a piece of writing on trial.

Three questions worth asking

Before dismissing something because AI may have touched it, ask:

1. Is the thinking original and valuable?
2. Does it genuinely represent the person or business behind it?
3. Has AI assisted the work, or replaced the thinking?

Here at Via, that distinction matters. Does it for you?

2. We don’t judge Photoshop the same way, do we?

Imagine dismissing a photographer’s work because you discovered they used Photoshop. *shock horror*

We generally don’t, right?

We understand Photoshop is part of the photographer’s toolkit. Used skilfully, it helps communicate their creative intent. Used excessively, well… we’ve all seen those photos where someone appears to have misplaced their pores. 

AI deserves a similarly nuanced conversation. It’s a tool, remember?

Using AI as a brainstorming partner to clarify your thinking, make language easier to understand, reduce repetitive work, or adapt an idea to a particular format does not automatically make the output less human.

The problems begin when we outsource the thinking itself to the machine.

A practical test

There’s a simple challenge doing the rounds that I rather like:

1. Put ChatGPT away for a moment.
2. Pick an ordinary object sitting near you. A paperclip. A coffee mug. A Post-it note. Now, without asking AI, name three alternative uses for it.
3. Then choose something you’re working on and describe it using three adjectives you came up with yourself.

It sounds almost laughably easy… and yet.

Was your first instinct is to open ChatGPT and type, “Give me three…”? That’s worth noticing.

This isn't a scientifically validated test for “AI addiction”. But as a thought experiment, it raises an important question: are we using AI to extend our thinking, or are we slowly outsourcing the muscle of thinking altogether?

3. What if AI taking jobs isn’t the whole story?

Recently, I spoke with someone I’d consider a highly confident AI user. He was actually giving talks about AI ten years ago. This is hardly someone hiding under the desk waiting for ChatGPT to go away.

And yet, he was worried.

If AI removes the entry-level tasks businesses currently pay people to perform, what happens to those roles? What happens to the business model built around them?

It’s a legitimate concern.

I see another possibility that's much more exciting. What if AI taking away menial work forces us to elevate the work humans do? 

We’ve spent decades building workplaces around productivity. And somewhere along the way, we lost our ability to be bored.

Boredom is only possible where there is empty space. It’s where two seemingly unrelated ideas collide and create something new. 

The alternative lens to see this through

As AI becomes better at producing outputs, the quality of the questions humans ask becomes more valuable than the volume of outputs we produce. 

That shifts the premium towards judgement, creativity, strategy, curiosity and decision-making. Inherently human skills, although technology is getting rather good at imitating some of them. Could a machine ever truly possess them? And if it could, at what point would we have to reconsider what we mean by human?  Science fiction has attempted to answer that particular question for us a few times. 

Time for another Blade Runner rewatch.

4. What system-literate leaders do differently

The court of public opinion often acts like a pendulum. We went through AI hype, and now we may be entering AI exhaustion, resistance and judgement. Eventually, the pendulum may find somewhere more… useful to settle on.

At Via, we don’t believe the defining question is whether a business uses AI. The same applies to any other software, machinery, automation or hiring another ten people (and will that be an offshore or local team? That’s a whole other blog post).

System-literate leaders don’t chase AI because everyone else is using it. And they also don’t reject AI because public opinion has swung against it.

They understand that the System comes first. They understand their people, processes, constraints and outcomes. Then they ask what technology belongs inside that system.

That’s a much less fashionable question. It’s also a much better one.

Don’t follow the pendulum. Build the system.

Some of the changes AI is bringing will be uncomfortable; others will be enormously valuable. Most businesses will experience a mixture of both. 

This week, choose one AI-enabled process in your business and examine the system underneath it. Ask what outcome it serves, what human thinking it preserves and whether the process would still make sense without the technology.

Want to see how mature your operating systems really are? Take Via’s 3-minute OS Systems Maturity Test.