As a leader, you can use AI to turn raw data into clear dashboards, spot patterns in customer feedback, and monitor performance. That’s powerful, and you should be using AI for that. But the next step – getting AI to make decisions for you – is a step too far. Use AI to inform your thinking, not to do your thinking for you.

https://swiy.co/go-dashboards-not-decisions

As a leader, how are you planning to use AI in 2026, and how are you planning not to use AI? AI is evolving so quickly you need to keep both of those things in mind.

My very first job was as a software developer, working for a small company in Perth that was subcontracted to large companies – for example, in the mining industry. Our software monitored their equipment, reported faults, and transferred data into their databases.

My first work trip was to a cable station in a remote corner of Hong Kong Island, overseeing the installation of our system to monitor the lasers in an undersea optical fibre cable between Hong Kong and Taiwan.

Every day at that cable station, I saw our customer do the same thing when he walked in: He greeted us, then walked over to the printout of the error report over the last 24 hours, hoping there would be zero errors from all the lasers.

Why? Because even a single error was a signal to alert his boss, and the team would then keep watch for a few days for more errors. The worst case scenario was a laser that was likely to fail, because it meant sending a ship out to sea, dragging for the cable from the ocean floor, pulling it up, breaking it, replacing the laser, and then dropping it down again. This could cost thousands – or hundreds of thousands – of dollars.

That error report was just a big matrix of errors from all the lasers over the last 24 hours (think of it like printing an Excel spreadsheet).

So what does this have to do with AI now?

That report, which thirty five years ago was just a paper printout on a dot matrix printer, is what we would now call a dashboard. And with AI, somebody without much technical skill can use AI to create a dashboard from their data.

They don’t have to hire a tech company.

They don’t have to ask their IT team.

They don’t have to wait weeks, months, or years for it to be ready.

They can do it NOW. With AI, they might even be able to do it themselves.

It could be equipment error data.

Or site safety data.

Or sentiment feedback on social media.

Or patient data from medical devices.

Or ongoing equipment monitoring.

Or workplace data showing employee engagement or office use.

I want to make two points about this use of AI.

First, AI has reached the point where even non-technical people (like you!) can create those sorts of reports and dashboards.

And you should.

Second, you’re using AI to create a dashboard to help you make better decisions. It’s NOT to use AI to make those decisions FOR you.

I’m emphasising this because AI HAS evolved to the point where an AI autonomous agent could make decisions for you based on data. It doesn’t need the dashboard; it just monitors the data, gathers the results, and makes the decision.

But this is not the way to use AI now. Just because you can doesn’t mean you should.

And in this case, you shouldn’t.

Thirty five years ago, when my customer and his bosses examined the report, THEY made the decision about when to bite the bullet and send out a ship. In theory, now AI could make that decision, but it shouldn’t.

Use AI for dashboards, not decisions.

For more about AI do’s and don’ts for 2026, join my online presentation about AI for leaders in 2026. It’s free, public, and open to all.

Register for the virtual masterclass:

https://swiy.co/go-dashboards-not-decisions

Hosted on Acast. See acast.com/privacy for more information.