The consequence of getting it right

Madeleine Albright, AI and the consequences that begin after a decision succeeds

There was an article in the New York Times last week about an economist who had created a John Maynard Keynes bot and asked what Keynes might make of artificial intelligence and the economy today.

It made me wonder what would happen if I did something similar with Madeleine Albright. I wanted to know what she might make of current world events and, more importantly, where she thought they were leading. I asked AI first, drawing on her public record, and then tuned into her myself.

The AI version was to be expected: weakening international institutions, pressure on alliances, growing nationalism, authoritarian power and the difficulty of maintaining international cooperation when national interests increasingly dominate. It was consistent with the Albright we know from her writing and public life.

What I received intuitively was different. We have gone further than we realise. We are looking at conflicts as separate events when they may be interconnected. Some of what has changed cannot simply be put back.

What did she mean? We tend to separate what is happening in the world into individual events: a war, an alliance, a region, a political movement. It makes them easier to understand and analyse, but the world doesn’t operate in those neat compartments. Everything is connected. Pull one thread and something shifts somewhere else. What happens in one conflict changes the assumptions being made in another. A boundary crossed in one place changes how seriously a boundary is taken somewhere else. What looks like a series of separate events may actually be a much larger shift that we haven’t yet recognised.

What happens if we get it right?

We devote considerable intelligence to anticipating failure because failure is understood as risk. When an outcome is successful, we congratulate ourselves and move on to the next. Once the intended result has been achieved, it tends to validate the reasoning that produced it, even though the wider environment may have changed precisely because the decision worked.

This becomes more important when the decision is being made by someone with significant power or influence. The consequences don’t stop with whether they achieved what they set out to achieve. Their decision can change the conditions around everyone else. A government crosses a boundary and other governments reconsider where the boundaries now are. A company successfully deploys a technology and competitors feel pressure to follow. An investor backs something that succeeds and capital moves towards it. The original decision may have been entirely rational, but its success has changed the environment in which the next decisions will be made.

Framing the analysis

Was the frame around the analysis large enough? AI could synthesise an extraordinary body of recorded information and construct a credible view from it. My intuitive perception didn’t give me more information about the individual conflicts I was considering but it certainly disturbed the boundaries I had placed around world events.

Those making decisions with the greatest downstream effects already have access to extraordinary intelligence in the conventional sense: data, modelling, advisers, experience, specialist expertise and increasingly powerful AI. Yet none of those can compensate for a question framed too narrowly, because every sophisticated analysis still begins with a judgement about what belongs inside it.

We can have excellent information, excellent analysis and even achieve exactly what we intended and still have asked too small a question.

A note on my practice: My intuitive work includes an Advisory Council of historical leaders and thinkers, including Madeleine Albright. I engage with them as a source of perspective and insight, which I then question, discern and bring into my own thinking.