Artificial intelligence and machine learning eat large datasets for breakfast and ask what’s for morning snack. They are powerful tools which can help us see insights we couldn’t before. These insights can be used in a predictive manner. (Of course they can be biased too, but that’s for another post.)
Predictive analytics is a fascinating field of study. And it is becoming part of the humanitarian conversation as well. But here’s the thing. It’s not new. Humans have been trying to predict things for millennia. In the humanitarian and development space we’ve called it ‘early warning’. But the warnings did not lead to action, so we added a couple or words, so it became ‘early warning and early action’.
But we still lack the action.
Predictive analytics are fascinating. Artificial intelligence is also fascinating. But they don’t solve the ‘action’ part. In many ways, they are just another shiny toy to add to the collection on the shelf.
The human side of digital remains the hard part. We tend to confuse the idea of being rationale beings with making decisions on data. We are rationale beings and the ‘data’ we use include a huge amount of emotional data.
Predictions are cool and fun. And we logically know spending $1 in prevention, can save us $10 in response. But that $1 can be used for other things right now that make me ‘feel’ better.
AI and predictive analytics projects should split their resources spending 10% on the tech challenge and 90% on the human challenge. If they would, their success rates would skyrocket.
The choice is up to us.