Time has certainly flown by since my last blog in December last year — nogal.
If anyone recalls, I did something rather unexpected during that December holiday: I immersed myself in a project based on a dormant business idea. My thinking at the time was simple — with AI now abundant and pervasive in the technology world, surely it should be possible to redevelop the product with the help of modern AI tools in far less time than before.
And it certainly proved to be the case.
In this blog I want to share what I am discovering in my own domain — software development — where AI is already demonstrating that significantly greater productivity is possible, often with less manpower. However, let me add quickly that im not going to bore you with techie-stuff, so do please hang in there……
We hear a lot of anecdotal claims about this, but allow me to confirm from direct experience that it is indeed becoming a reality. Software businesses – and others – can now accomplish much more with smaller teams and less staff capacity.
Of course, in a country like ours where unemployment is already rife, this raises something of a moral dilemma. Increased productivity through AI is economically attractive for businesses, but it also forces us to think carefully about what this means for employment in the years ahead.
So the question I’ve been asking myself is this: if one person, assisted by AI, can now do the work that previously required several developers — what does that mean for the future of small businesses?
Looking at my own “shop”, the answer has been quite revealing. I have not done hands-on coding for literally decades. My role has long been directing the business rather than writing software, so the idea of personally producing code again never really crossed my mind.
Yet today, with AI tools assisting me, I find myself generating code at a rate roughly four times faster than some of the developers I employed in the past….believe me.
On the one hand, this is clearly good news for small businesses. It means a company can become more competitive, spend less on software engineering resources, and still move projects forward faster than before.
And it helps enormously if your development manager happens to be called Jean-Claude — a little nickname I’ve given to my AI assistant, inspired by Anthropic’s Claude model (a competitor to OpenAI’s ChatGPT, which I’m sure you have all come across). Trust me, Jean-Claude knows more than 99% of the software engineering guys I have encountered over the years. And, to labour the point I can engage in an exchange of ideas and, yes, debate with Jean-Claude…. and you’re now thinking these software guys really need to get out more, … he he
It certainly makes you think about what a “team” will look like in the small business of the future.
Perhaps the real change is this: instead of teams growing larger, they may simply become smarter — a few people working alongside powerful AI tools.
And if that turns out to be the case, the skill of the future may not be writing the code itself, but knowing how to think clearly, ask the right questions, and guide the machines helping us build things.
Perhaps the role shifts from programmer to architect — as I always used to tell my guys: “once the conceptual design is done, the rest is just code.” Designing the structure while AI does much of the heavy lifting.
Something tells me Jean-Claude and I are only just getting started.
Leave a Reply