Keeping our tools sharp

Recently, in a conference talk, I referenced a quote from the late Apple Computer cofounder, Steve Jobs, where he suggested that as tool builders, perhaps the most remarkable tool humans have made is the computer. That was a quote from 40 years ago; I’ve suggested that perhaps AI is the most remarkable tool humans have made.

The use of the word tool to describe AI works as a description, but it also provides a nice analogy to physical items, saws, chisels, lawnmowers and so on.

Anyone who has experienced the frustration of trying to cut with a blunt saw or the slow progress one makes cutting grass with dull blades will tell you it’s frustrating. We have the same negative experience with kitchen knives and scissors, that feeling of "I’m sure it was easier before", then if you don’t act, you can end up frantically trying to cut, making little progress.

We know software typically receives updates, even if this is only to fix problems, ensure compatibility with updated operating systems or to reduce the threat of security issues. As we adopt AI tools, it’s important to remember they are software tools. The code needs to be kept up to date. But it’s more than just that – just like your sat nav needs updating to include new roads and junctions, the data in your AI may need to be refreshed. Failure to do so will lead to frustration and poor results.

The physical tool analogy fails a little - after all, most of us don’t use our limited selection of home tools enough to wear them out regularly. But with AI in the workplace, we’re not mowing our small lawn once a month; we’re more like a carpenter, stonemason or groundsperson working with our tools all day, every day.

In reality, that means making sure that we have sufficient budget and assignment of time from technical partners to maintain and retrain our AI. We need to consider how to keep our tools sharp.