Thoughts on the Changing Development Landscape
A story I heard during a recent external meeting > We don't build with MSA anymore. We build monolithically. To be precise, we build structures that AI can understand well, all at once. > SaaS might disappear entirely too. Since you can build things internally to fit your needs, there's no real need to build and subscribe to SaaS anymore.
- MicroService Architecture and SaaS were the biggest keywords when I was building IT products. Now they're out? Suddenly I miss the bosses I used to work with. (Gotta go see them!)
- I probably won't return to building products directly. I'll likely do more overall management work, and that direction is in the social-tech field. What should I focus on examining more at this point?
A line I highlighted while reading today's geeknews newsletter: > There are clearly domains where the very fact of human intervention holds value, regardless of efficiency. Products involving elements like curation, creativity, presence, and approval play a different game than simple feature competition. That's why the article separately explains this as a constant concept called the "Human Coefficient". It's also possible to interpret this as **two distinct survival paths: tools that conserve tokens versus experiences where humanity holds value. > Rather than seeing these two trends as conflicting, I feel they represent different facets of the same shift. Code was originally a "language created to converse with computers," and if that interface is shifting to natural language, it's only natural that development methodologies evolve alongside it. However, whether this change leads to the 'demise of craftsmanship' or becomes a process of shifting 'what we consider craftsmanship (design, verification, operations, security, product sense)' depends on perspective. We are currently standing right on that boundary, making the act of posing and pondering the question "What must we create to survive?" increasingly crucial.
-
It was interesting to note that while code was once 'a language created to converse with computers,' now natural language—more specifically, 'English'—has become 'a language capable of conversing with computers.' In that context, what matters more than the method of creation itself is the thought behind 'what to create.' Hmm, wasn't this also the reason I switched from computer science to philosophy? Interesting.
- Software Survival 3.0 – What Must We Build to Survive?
-
Original: Software Survival 3.0
- AI Code and Software Craftsmanship
- Original: AI code and software craft