Workshop Over Workslop

I recently had a good conversation with a head of data & AI. He had created the AI strategy for the company with ChatGPT. The result: an “80% AI strategy” as he termed it.

Because his contacts didn’t have time for workshops, he simply gathered all the existing documents, held a few one-on-one meetings, and brought it all together with ChatGPT.

On the one hand, I find it remarkable that this is possible thanks to LLM. And sometimes 80% is enough to get started.

On the other hand:

1) Is 80% enough to be sure to that you’re doing the right thing? A strategy determines whether you win or lose. Whether your company rises or falls.

2) If AI increases productivity so much, where has the time gained gone? Why didn’t the colleagues have time for workshops?

I would expect that because of AI automation employees spend less time chatting with computers and have more time collaborating with colleagues. But the opposite seems to be the case: we spend even more time alone with the computer than with colleagues.

However, innovation does not arise in isolation, but in interaction with humans. Your business, your products, and your services are for people, not technology.

Transformation is not just a change in technological infrastructure, but a change in organizational structure and personnel culture. Changing requires wanting to change.

“Structure follows strategy.”—Alfred Chandler

That’s why strategy must be human-centered. Doing the right thing is doing what your stakeholders want.

AI is great if you know what you want. What you prompt is what you get. WYPIWYG.

But to find out what your customers, colleagues, and investors want, you have to talk to them. Who you talk with is what you learn.

You won’t learn anything new from an LLM: LLM samples from the average. Consequently, an LLM-generated strategy will be an average strategy.

Having an average strategy is the same as having no strategy at all. The whole point of a strategy is to find an outstanding way to win the business.

Don’t fall into the workslop trap: doing things because they are easy.

I know from experience: workshops are hard work. As one participant recently said: “Martin, your workshops are exhausting. You have to think so much.”

We choose to let stakeholders design their AI strategy, not because it’s easy, but because it’s hard.