t’s been fascinating following Kent Beck’s exploration of ‘augmented coding’, and in particular his analysis that AI today is very good at creating features but not very good at preserving optionality.
Sooner or later, unless you’ve got a very capable programmer (like Beck) keeping a close eye, when writing code AI will default towards a dollar today.
In other words, it will naturally focus on shipping the feature, adding something new, increasing complexity to deliver results today, rather than thinking about how to keep it’s options open to possible futures.
This makes sense.
AI as a technology is about referencing and recombining prior human knowledge.
Creating 10 dollars tomorrow requires imagining a new tomorrow that might look different to yesterday, and building towards it.
Envisioning a future, creating a new reality by seeing it and then working towards it, has been and I believe will remain fundamentally human activity.
AI will be an incredibly powerful tool in helping us achieve these feats, unlocking our potential and unleashing our creativity.
But we must lead it.
It’s the steam train.
Only we can can look ahead and lay the tracks to get where we need to go.
