Are We Ambitious Enough With AI?

I think one of the biggest limitations in how most people use AI today is not necessarily what AI can do, but what we imagine it can do.

Most of us are still using AI to do what we were already doing, only faster. Write an email, summarize a document, prepare a presentation, analyze some data.

Nothing wrong with that. But the driving question that is emerging is: are we being ambitious enough?

I came across an interesting article by Felix Rieseberg from Anthropic, titled How to Be More Ambitious: A Tutorial for Humans with Models.

Felix makes an interesting argument. As AI becomes more capable, we need to rethink not just how we do our work, but what we attempt in the first place.

He suggests four ways to think about being more ambitious with AI.

1. Speed: Move beyond doing things faster.
   Getting AI to prepare a presentation in five minutes instead of five hours is useful. But that’s just doing the same thing faster. The bigger opportunity lies elsewhere.

2. Altitude: Aim for a bigger outcome.
   Instead of asking AI to prepare a quarterly business review, ask it to identify the most important decisions the business needs to make. Go a step further and ask whether quarterly reviews are even the best way to manage performance.
   The ambition shifts from producing a report to improving how the business operates.

3. Depth: Go deeper into the work.
   Instead of simply asking AI why customers are leaving, ask it to investigate the problem as a top professional would. What original data should be examined? What assumptions need testing? Which methods would help uncover the real reasons? And how would we verify the findings?
   The ambition is not to get a longer answer, but a more rigorous one.

4. Range: Think beyond your own field.
   A business problem doesn’t always need just a business perspective. What if we brought in the methods of a forensic accountant, a data scientist, a journalist or a psychologist?
   AI makes it much easier to bring different kinds of expertise to the same problem. The ambition is to explore approaches we might never have considered ourselves.

Three things I take away from this:

Think bigger: Don’t just automate existing tasks. Question what you could accomplish.

Go deeper: Demand better evidence, methods and validation, not simply more polished answers.

Look wider: Explore perspectives and approaches beyond your own expertise.

We spend a lot of time learning what AI tools can do. Perhaps we should spend just as much time learning to imagine what we could do with them.

As AI becomes more capable, our ambition needs to grow with it.

Reference: How to Be More Ambitious: A Tutorial for Humans with Models — Felix Rieseberg