Humans are notoriously terrible at understanding patterns that are not linear, so it shouldn’t come as a surprise that the recent breakneck development of generative AI has left many of us unsure on how to use it, how much to use it and if we should even use it.
One thing is clear to me: genAI is here to stay and it will change how we interact with the world, with each other and with ourselves. This, however, brings us to multiple ethical questions, emerging from the devastating environmental impact that training LLMs is having and will have on the planet, the risk of losing human authenticity in interpersonal remote interactions and the devaluation of human intellectual, technical and artistic craftsmanship as a whole.
Personally, at the moment I do use an AI tool, namely Claude Code. It is a superb agent and it has sped up many development processes of non-critical applications I use it for (especially with CSS; sorry frontend people) and has proven incredibly useful when delegating menial tasks to it. However, I believe that using genAI in critical, professional applications is not only risky, but disrespectful to humans that will interact with it.
For this very reason, everything that is written on this site in the info, now, about, blog and ai pages is completely written by hand by me, without the intervention of any genAI tool. In general, most of the content written anywhere else on this site will have been written by hand by me and eventually reformatted, grammar-checked or summarized by an AI tool. If that’s the case, the tags will include #genai/formatting, #genai/content or something to the same effect to separate them from the rest. I will update this page when it will become necessary to.
I encourage everyone reading this to consider what you are giving up when delegating creativity to an AI tool: your unique ingenuity is precious and non-reproducible. Think carefully of the added value you can bring to the world and let that be a motivation to do things the old, hard way.