My AI adoption was immediate.
As soon as ChatGPT landed in the UK, I had it. Started playing with it straight away.
Images. Editing. Spell check, grammar, making things actually readable. Even using it like a better version of Google.
If there's ever a proper "Google 2.0", it's going to be AI. Probably something like Claude, if I'm honest.
But here's the weird bit.
The place I've been slowest to adopt it? Power BI. My actual job.
I've let it help here and there. Wireframes. SVGs. Code reviews for DAX, Power Query, even a bit of Vega-Lite in Deneb.
All useful.
But the thing I've dragged my feet on is AI-driven reporting.
And there's a reason for that.
AI can read your data. It can build a dashboard. It can tell you your monthly gross profit is £10,000,000.
What it can't do is tell you why that number is wrong.
It won't know Dave in accounting messed up the tax last month. It won't catch that Janet in products listed toilet rolls at £100 instead of £1. It won't question why Steve in sales somehow created three invoices for the same deal because the button lagged.
And good luck getting it to spot that Abdul switched all the dates from UK to US format.
I'll catch those in minutes.
Not because I'm smarter than AI. But because I understand the business behind the data.
AI might flag "outliers". It might say something doesn't look right.
But it doesn't sit in meetings. It doesn't listen to stakeholders. It doesn't understand context.
That part still belongs to us.
So yeah, I've been slow to let AI "take my job".
Because it won't.
What it will do is strip out the boring, repetitive bits.
And that's perfect.
Because the part I actually enjoy? Taking messy questions, messy data, messy businesses... and turning that into answers people can use.
That part isn't going anywhere.
#PowerBI