The manifesto for Data Universe 2025 is coming! Like last year, it’s written by entrepreneur, best-selling author, and DU conference chair, Alistair Croll. To find out more, we spoke with Alistair about why he’s making it our mission to help you decide better, two business challenges he sees coming soon, and his big hope for our next event.

In the manifesto, you will say, ‘the antidote to coordination problems is common knowledge.’ What do you mean by common knowledge in this context?

Let me give you an example that I didn't put in the manifesto because it’s too politicized. We recently had a very good, concrete example of common knowledge in North America. Everybody said President Biden was still fit to run for a second term. And then, when everybody saw him in the debate, everybody knew that everybody else knew that he was losing it and could not run again. And within a few weeks, he stepped down.

Change happens when everyone knows that everyone knows something.

As I write in the manifesto, Data must not only help us decide better but also make us act better. It must be the child at the Emperor’s parade, willing to speak truths so self-evident in hindsight that suddenly, we know everyone knows.

What impact will AI have on businesses tomorrow that people should prepare for today?

I have two big predictions from AI in business: The rise of shadow software developers and the second-order effects of generative AI.

In the early 2000s, we saw the rise of something called shadow IT. I’ve spoken to CFOs who ran reports of all the Amazon Web Services (AWS) charges on all the company credit cards and found out they had fifty AWS accounts. They had no idea; they couldn’t unplug it because the business was running on it, and this was shadow IT.

The other day, I built something for a conference I run using code and Google’s app script. It extracted information from a spreadsheet and put it into a slide template. It was straightforward to do because that sort of thing is nowadays. Am I a developer? What if that app became part of our company’s processes?

Forget the rise of shadow IT—this is the rise of shadow software developers. QA does not test the code these people write; it doesn’t have version control, and it doesn’t have documentation. Some of it is tied to their Google enterprise account, so if they leave the company and their account is deleted, say goodbye to all the scripts they wrote attached to those spreadsheets that are now integral to how the company operates.

A new wave of shadow IT is about to break, cresting not because of the easy deployment cloud computing delivers but because of the easy development AI promises.

Prediction two. Most of our current business processes use something else as a proxy for value. For example, if I’m a government agency granting money, people must complete a 150-page grant application. And because writing a 150-page grant application is hard, I’m using that as a proxy for quality – if you have the time and the effort to do that, I should look at you.

But today, whenever there’s a process, we should check each step and ask: What is my proxy for value here? Then ask: Does technology change how easy or hard that proxy for value is? Writing a 150-page grant application just got easy, while identifying someone just got hard since we can no longer be sure we’re talking to a human.

Across every facet of a business, operators need to reconsider what is now easy and hard and adjust their processes accordingly to mitigate the second-order consequences of AI.

What is your hope for DU25?

We help organizations decide better and act better to be as efficient and effective as possible in achieving their goals. That’s why Data Universe 2025 will explore where the rapid pace of innovation is leading us and ask how best to harness that change in service of what truly matters: better outcomes.