Come gather ’round people
Wherever you roam
And admit that the waters
Around you have grown
And accept it that soon
You’ll be drenched to the bone
If your time to you is worth savin’
And you better start swimmin’
Or you’ll sink like a stone
For the times they are a-changin’
Bob Dylan, The times they are a-changin’, 1964
How are you doing?
Much changed these last couple of months? Noticed anything unusual going on?
OK, so maybe I exist in a weird bubble where 87% of the working population uses Claude Code. And 68% of that 87% feel qualified to tell the other 32% how to use it. But there’s no denying something has shifted. Shifted in a tectonic kind of way.
Sticking with stats, developers in leading companies are boasting that LLMs now generate over 90% of their output. Code is essentially data. So if you work with data in some capacity, as an analyst or engineer, this is the way things are heading. As Bob says, better start swimmin’
Be the change
You’ve got a choice. You can sit back and let change happen to you. Or you can plunge into it, move with it, and join the dance (Alan Watts - there’s just too many good change quotes out there).
How’s about we have a go at naming the changes you’re dealing with and by doing so, attempt to better understand them. Psychologists call this labelling. Pull up a couch.
The moving target
Change number one is the fact that damned destination just won’t sit still. It’s very hard to define a Target Operating Model when the target keeps moving. It's meta, but the biggest change is the fact everything is changing - and at different speeds. AI advancing, agents incoming, architecture upended. Transformation can no longer be a fixed destination, it has to be a capability.
The eversion of expertise
If the inversion of expertise is junior people knowing more about certain tools than their senior managers, the eversion of expertise refers to the idea that technical expertise will be externalised. LLMs will soon have more expertise than all their human counterparts combined. Not yet maybe, but soon. The FOBO (fear of becoming obsolete) is real. The risk is the erosion of cognitive function and critical thinking.
The identity crisis
Right now you’re leading data analysts and data engineers, and maybe analytics engineers. Clear-ish roles defined by expertise in a set of tools and techniques. Roles that conferred status are now dissolving and blending - with a corresponding loss of identity and purpose. "Just do your best to use copilot" is not a clear career progression path.
The outward orientation
OK, so this one’s been hanging around like a bad smell. The sweet whiff of data democratisation. Only now it’s actually cooking. Because asking a data question of an LLM (whether via an agent, Slack or Teams) happens to be waaaay easier than doing a 3-day Tableau course. Internal expectations have changed - data teams have got to start thinking and acting like product teams.
Even the data itself
Ahhh... remember good old tables? SQL, schemas, rows, columns, and nice clean joins? Our raw materials are changing too. Time to welcome multimodal data into your warehouse (or lake, or lakehouse, whatever). Images, audio, video. Different ingestion patterns, different storage, different analysis. Just as well not much else is changing.
And that’s without even mentioning changing skills, tools, vendors, standards, systems, process, workflows, agents, documentation, governance, ethics... Oh, there you go, I mentioned them. That’s a complete list then.
What do all these changes have in common? No, not AI. Well, yes AI - but that’s for another time. It’s people.
Yes, those people. Your people
None of this change happens without people. None of it happens without people changing. And ideally people embracing the change and being energised by it, not feeling like hapless victims. Turns out people aren’t expendable, they’re invaluable. AI doesn’t make them redundant, it should make them extraordinary.
And leading people through this change? Telling them what they need to learn and do while involving them, getting them jazzed while keeping them calm, driving them faster while preventing burnout. That’s arguably the biggest challenge of all.
The data leaders that will thrive won’t be dealing with these changes sequentially - or even in parallel. They’ll be building their team’s resilience and capability to absorb them continuously.
Need any help?