## The "Good Enough" Trap I wrote an article a few weeks ago about creativity. This article was focused on AI and how it impacts creativity. Here it is: [[What is Creativity]]. Before I begin I wanted to talk about how this blog specifically was an attempt to get better at writing, for me. I grew up never really loving to write but as I grew in my industry I found I was writing more documentation and architecture analysis than actual code! So a blog is formed where I can collect me thoughts and practice. Enter AI, I thought this would be a great way for me to get into the grove! I started "taking notes" in chats with Gemini providing article ideas and topics, I would start with a title and then throw more ideas in as I was out for a walk, watching TV... Etc etc Ex: I may start with "write a tech slice blog post structure about the impact of pineapple on pizza". And as I found articles or ideas I would then prompt "add a note about how they can add unnecessary moisture if not prepared properly before application" This would then go on, and once I felt I had enough in my head to structure something I would go back to my generated notes and be off to the races! All I would need to do is structure my current ideas and add some more and it would be perfect!!! I couldn't have been more wrong. What I ended up with what a sloppy mess of AI generated stuff mixed in with my own nuggets of knowledge (ie. The stuff I actually wanted to say). The result lost my voice! (Although close friends said they could explicitly see what I wrote vs what I didn't) And if I'm being honest I didn't even realise that I had left so much generated stuff, or that there was really even generated stuff in there! I did edit my work and write giant chunks of text, but the AI still slipped through, the long-winded explanations were "good enough" and seemed to fit my voice so I let them slip in. What I ended up with were definitely my ideas, but not my voice. I started to contrast this with generating code, an area where I am very proficient, due to my experience in the field I can very easily tell what code I want and what code I don't from a prompt. I know where the quality bar needs to set, and I am very aware of how to keep it there. Since I am a novice writer, I couldn't identify these things, AI became a cheat code for me to be something I wasn't. AI got to set the bar instead of me. And even worse, it robbed me of the experience of learning how to write! Which was my entire goal in the first place. This is from experience. Yet even in this field the AI slips in. I see something and think it's "good enough". It may not have been what I would have done, I could have even done it better! But alas, it's done now and it works, so now I can get onto other things! It's such an easy and slippery slope. Regardless of intent. Let me show you a trifoce ## Triforce of Goals (Quality, Speed, Cost) ![[Pasted image 20251017094842.png]] This brings up the "Triforce of Goals" or whatever you want to call it. Its something I bring up with managers or other developers at least a few times a year! The idea that there is a triangle of "Quality, Speed, and Cost" and in nearly any project. You can "pick only 2" out of the three. You can build something high quality and quickly at a high cost. You can build something fast for cheap at low quality, or you can build something expensive and high quality, but it will take a long time. By using something like AI this kind of mindfulness to a project all but goes right out the window. While quality may be top of mind (as it was for me) it's the constant barrage of generated "good enough"a that inevitably reduce the quality of what you are trying to do. Cost is already astronomically high for AI tools, in terms of monetary cost, environmental impacts, and its "reduction" of labour costs (ie. Effort). Which leaves us with speed, we wanted it fast and we wanted it now. And well... That's what I got! When using AI the wizard at the end of the yellow brick road seems all so much more attainable, and extremely achievable. Instead of taking the path, smelling the flowers and meeting some friends along the way, we jumped in a helicopter and rushed over without a second thought. We skipped the journey and had the finished product before we even realised what we wanted! And with that. I will be uninstalling all Source editing AI tools. Goodbye Warp, so long Codex, see you next never, Claude. The temptation to "just get it done" is way too strong. I know I will give in, and I know the result will be "good enough" for me to ignore it and continue. It's not only hurting me as a developer by failing to home the skills I've gained over 10 years. But it also robs me of the growth I love and adore that this career gives me. I will still use these tools for supplementary help (ex: throw this article in and ask it for any meandering thoughts or points I may not have driven home enough... Ie... 2nd, 3rd or 4th opinions). But not it's "core" So. If you want to learn to write, then write. If you want to learn to code, then code. Don't allow these tools to rob you of your growth and your skills! This article brought to you by my brain and fingers. Evan.