Note
The Boundaries Around What We Can Do Have Moved
The most interesting strategic question about AI may not be how much AI an organisation should use. It may be what has become possible now that some old constraints are moving.
“What’s our AI strategy?”
It is a reasonable question, but I wonder whether it can lead organisations in slightly the wrong direction.
Once AI becomes the subject of the strategy, the next step is often to find AI use cases.
Then to create AI projects.
Then to measure AI adoption.
Before long, using the technology can start to look like the objective.
The more interesting thing about AI is not that we can now do AI.
It is that the boundaries around what an organisation can do have moved.
Things that were previously too expensive may now be affordable.
Software that would never have justified a development team may now be worth creating.
Experiments that needed significant investment before anybody could learn whether they worked may now be cheap enough to try.
Work that could not be personalised at scale may become much more adaptable.
Analysis that was impractical may become routine.
That feels like a different strategy conversation.
Instead of starting with “Where can we use AI?”, start with the constraints around the organisation.
What have we historically been unable to do because it was too slow, expensive, complex or labour-intensive?
Which of those constraints are changing?
And if they change, what would we choose to do differently?
The answer may be an AI-enabled service.
It may be a completely different operating model.
It may be a small piece of software that removes a longstanding irritation.
It may be an experiment that tells us an idea isn’t worth pursuing.
It may even be that the newly possible thing is still not valuable enough to do.
That is fine.
Technology creates options. Strategy still involves choosing between them.
This is why I increasingly prefer the question:
What can we now make possible that we couldn’t before?
AI will often be part of the answer.
But it doesn’t need to be the objective.
Originally published on Threads.