25 Comments
User's avatar
Williams Hope's avatar

AI doesn’t remove ambiguity. It increases the speed at which ambiguity spreads.

Williams Hope's avatar

That is kind of true

Tam Nguyen's avatar

"We’ve been doing technology wrong for years. AI is simply making it visible."

This is the crux of it all. It's just painfully obvious now.

Pawel Jozefiak's avatar

"AI does not transform organisations. It amplifies them." That framing cuts through a lot of noise.

The procurement theatre pattern is so accurate - sign the Copilot contract, run three webinars, report progress to the board. What I find missing from most org-level AI audits is any honest look at the underlying process maturity.

Companies that couldn't measure outcomes before AI still can't. The technology just makes the gap more expensive. The 1% 'mature AI' figure tracks with what I see. Adoption is cosmetic. The hard work is the organizational redesign that nobody wants to pay for.

Peter Ashby Smith's avatar

Strong observation!

You’re right about process maturity. AI exposes whether outcomes were ever measurable in the first place. If you couldn’t define value before automation, you won’t define it after.

What I’m seeing is that AI doesn’t just amplify capability. It amplifies governance quality.

Loose metrics become looser. Clear systems become sharper.

The issue is that most organisations don’t have an AI problem, but a clarity problem. AI just removes the hiding places.

The redesign work is going to be expensive. That said, avoiding it will compound the bill.

Tsetsy's avatar

Procurement theatre is the perfect phrase. So many orgs think buying licenses equals transformation.

Jasmin Author's avatar

If incentives don’t change, behaviour won’t change no matter how advanced the model is.

Peter Ashby Smith's avatar

Indeed. Change is difficult, but what makes us tick is surprisingly simple.

Happiness's avatar

This reframes AI from “efficiency tool” to “organisational cognition.” That’s a deeper shift than most boards realise.

Peter Ashby Smith's avatar

100%, Happiness.

I read articles every other day about how AI hasn’t delivered on productivity gains.

And yet, very few leaders are asking the real question as to “why” this amazing new technology isn’t have the impact it should in their organisation.

The reality is that we’re not dealing with “just a tool” any more. These are different ways of working. That’s where transformation has to start, with how we work together.

Carina Ashur's avatar

The Cognitive Infrastructure Model is the real upgrade workflow redesign over feature training.

Peter Ashby Smith's avatar

That’s right. I’d love your thoughts on how we can persuade senior leaders to do the hard work around transformation rather than the easy “win” of buying a tool?

Grook's avatar

The deskilling risk is under-discussed. Augmentation requires intentional capability protection.

Peter Ashby Smith's avatar

It’s why the thinking needs to be inverted. Start with people, then find tools to better augment their processes. No AI architect job title can change an organisation by themselves. Capability need to be there.

Camella Andrew's avatar

Most companies digitised bad processes for 20 years. Now they’re automating bad thinking.

Peter Ashby Smith's avatar

Spot on. Nothing much has changed. The real slop isn’t AI, it’s “busy work”.

Samuel Cassy's avatar

Shadow AI becoming shadow infrastructure is the quiet risk no one is talking about enough.

Peter Ashby Smith's avatar

Oh yeah. Everyone is using LLMs and other automations, but the organisation has no idea they’re doing it. Huge risk, and it’s just a matter of time before a really big cautionary tale arises.

Blessed Manson's avatar

The line “AI amplifies organisations” is the real takeaway. It doesn’t fix weak culture it exposes it.

Peter Ashby Smith's avatar

Thanks, Blessed Manson.

Sweeping dysfunction under the rug is no longer an option in a world where automation of the mundane is possible.

Matthew Ace's avatar

The pre mortem idea is powerful. Most AI failures are predictable we just avoid naming them.

Will Walker's avatar

I think the root of this issue is that most of the people at the exec level are not tinkerers. They aren’t seeing how AI changes workflows or providing autonomy to their workforce to tinker in an internal sandbox

Misletoe Jerry's avatar

The strongest organisations aren’t asking “Which tool?” They’re asking “Which decisions?”

User's avatar
Comment removed
Feb 28
Comment removed
Peter Ashby Smith's avatar

Right. But the why is important.

Authority has to stay human because authority is about responsibility.

AI can analyse.

It can optimise.

It can draft.

It cannot own consequences.

The moment no human is clearly accountable, you don’t have intelligence at scale. You have risk at scale.

AI should sharpen judgement, not replace it.

User's avatar
Comment removed
Feb 27
Comment removed
Peter Ashby Smith's avatar

Yes. It’s way easier to buy an IT system, roll out a press release and declare your organisation “AI-capable”. Actually getting people and culture to shift is the harder game to play.