AI Has A Massive Identity Problem - A Reframe

May 5, 2026

Hi Humans.

Let me start with numbers that should give all of us pause.

Despite $30–40 billion in enterprise investment into GenAI, 95% of organisations are getting zero return. Just 5% of integrated AI pilots are extracting real value. The vast majority: no measurable P&L impact. This according to an MIT Media Labs report at the end of 2025.

The National Bureau of Economic Research in March 2026, published a study of 6,000 executives from across markets, with 9-out-of-10 reporting little or no impact on employment or productivity. They reported 69% usage, but an average of 1.5 hours per week.

That's not a technology failure - the models work.

That's a narrative failure and it's one that needs to be talked about.

The most transformational technology of the last century has done a terrible job of selling itself.

At an enterprise level, AI has been breathlessly sold as a productivity tool — one of human replacement, trumpeted by media as the harbinger of a job apocalypse. Goldman Sachs famously predicted 300 million jobs disrupted. Some in the tech world have gone further, suggesting that within a decade (many also say sooner), AI-driven automation will effectively eliminate the need for human labour altogether.

So inspiring....Give me strength.

Tristan Harris of the Centre for Humane Technology made an observation that has stayed with me: with over $1 trillion invested or committed, the fundamental economic incentive of the AI industrial complex is inherently anti-human. There is no payback if humans are still in the loop.

And then the layoffs came. Tens of thousands in the last six months alone — ironically, many from the technology companies most loudly evangelising AI at scale.

The narrative is doing serious damage. Not just to adoption, but to the humans being asked to adopt - their very sense of identity.

Because here's what this narrative feels like at ground level.

I try to keep a reasonable rhythm of learning — following practical AI voices on YouTube and LinkedIn, staying across the tools - using them daily.

Fact is I can't keep up.

Even in this space, of self-driven learning, the narrative falls over again.

Every single day: 'Insane New Updates. You Won't Believe!' — 'Tool XYZ Just Killed Tool PDQ!'

The thing is, I only just started to get a handle on 'Tool PDQ'...

This creates a spectrum of human response that I suspect most of us recognise somewhere in ourselves. At one end: the AI Ninja, determined to master every new development.

In the middle: people genuinely trying, using it daily, but quietly feeling left behind.

At the other end: the growing number of people who have concluded and been made to feel that it's simply beyond them, so why bother or care.

That last group is not a talent problem. It's a narrative problem.

And it translates catastrophically when those same humans are told by their leadership to "use more AI."

Anxiety and fatigue. An impending sense of replacement or displacement. A suspicion and wariness of the intent and motives of the organisation's reason for deployment. Their very sense of worth and identity being challenged.

Organisational cultures are already under huge pressure, as engagement globally has tanked - poorly thought out and executed deployment of AI could crater them - but that's for another day.

Here's where I want to offer a genuine reframe.

The dominant narrative of productivity is obscuring something far more important: AI's true potential as the most powerful tool of possibility we have ever seen.

Productivity is an economic obsession with a terrible identity of its own. It's loved by economists, reductive in practice, and largely about cost reduction against flat or declining inflation-adjusted revenues. We have spent the best part of a decade focused on this approach to growth in the volatile economic environment and we are reaching the point of seriously diminishing returns.

What if we asked a different question?

Instead of: how do we get our people to use AI and how many can we replace?

What if we asked: how do we use AI to eliminate the dross and drudge — the work that adds little value anyway — AND free human imagination and ingenuity to explore and create something genuinely new?

The author David Graeber coined the term "bullshit jobs" to describe the material proportion of corporate roles that add little or no real value. An honest interrogation of this, before deploying AI, would be a radical and productive place to start.

If the work itself doesn't create meaningful social or economic value, it doesn't matter how efficiently you do it. You're just doing the wrong things faster.

Which brings me to a word I've been sitting with.

Impactivity.

Not productivity. Impactivity.

The measure of what happens when humans use AI to amplify their distinctly human advantage — their judgment, their creativity, their relationships, their ingenuity — to create ideas, business models, products, and experiences that couldn't have existed without both working together.

one + one = at least 2.5. Ideally more...

It's a tragic misuse of this technology to just use it to try and do what we already do, but 'better'. We don't need incrementality as a world, we need big shifts - new hope.

Impactivity forces us to ask the hard question: are we even working on the right things? If not, then what could and should we be co-creating and striving for? AI can help us get there.

This is ultimately a leadership challenge — not a technology one.

AI strategy and deployment cannot sit with IT to manage like an ERP or Cloud rollout. It has to be human-centric.

Where are the opportunities embedded within the organisation - the ones that only those in the business at the critical interfaces can see.

Change led by humans who understand human appetite for and resistance to it, empathy for the human identity, and the nuanced difference between compliance and genuine adoption.

The enterprises getting up to 90% initial adoption — and they exist — are taking a slower, more human-centred approach.

They're co-creating with their people, not deploying at them.

Also - slow down - think.

To take more time is counter-intuitive in what feels like an arms race, but that might be what is needed to truly understand the possibilities; the nuances, the human quotient and how to create an aspirational, but practical roadmap for use and deployment, rather than drown your people in a fire hose of tools and learning modules.

Use smaller, considered pilots to demonstrate viability and collect stories from teams using the tools, who are deploying AI to eliminate what drains their time and energy and amplifies their value-creating human advantage.

The narrative needs to shift. From replacement to augmentation. From productivity to Impactivity. From how many humans can we remove to how much more human can we become.

That's the identity reset AI desperately needs.

And it starts with us.

Remember: HI + AI = Amplified Human Impact.

Lets Keep Being More Human. 🟡

What's your experience of the AI narrative in your organisation? Are you seeing genuine adoption — or compliance with anxiety underneath it? I'd love to know in the comments.