Small, focused diffs have almost always served me better. They have several advantages:
making small changes, basically a continuation of the previous point.
Small, focused diffs have almost always served me better. They have several advantages:
making small changes, basically a continuation of the previous point.
Small, focused diffs have almost always served me better. They have several advantages:
making small changes, basically a continuation of the previous point.
Avoid wandering through the code
work on specific parts of your code only. I recognise this wandering even in my small projects.
Generally, doing less is faster and easier! Depending on the task, you may be able to soften the requirements.
try and reduce reqs. For my personal tools this is often achieved by not having to deal with edge cases and my own behaviour being predictable, or that I can prescribe myself a specific way of working.
Software, like writing, can benefit from a rough draft
work from a rough draft and iterate
meaning everything is work in progress and marked as such
How good should this be?
in software dev you need a sense of how good the code needs to be in comparison to use case / context / impact of failure time available
Some tips on how to build software quickly. Via Alper, who says it seemd obvious to him, but then said he also recognised it bears repeating. It seems useful to use as cheat sheet even of tiny personal tools I make, and for home vibe coding projects too, wrt how to initiate interaction w LLMs
The thing is, software is not an asset, it's a liability. The capabilities that running software delivers – automation, production, analysis and administration – those are assets. But the software itself? That's a liability. Brittle, fragile, forever breaking down as the software upstream of it, downstream of it, and adjacent to it is updated or swapped out, revealing defects and deficiencies in systems that may have performed well for years.
software is a liability. Dutch equiv of this phrase? The assets are its impact : automation, production, analysis, admin
Kommer den artificiella intelligensen att bli bättre på att tänka än den mänskliga? Kognitionsvetaren Peter Gärdenfors förklarar varför så inte är fallet. Den mänskliga intelligensen består av en rad olika färdigheter och specialiteter som har förfinats under tusentals år. Mycket återstår innan den artificiella intelligensen kan mäta sig med det tänkande som inte bara människor utan även djur har. När vi förstår att vår intelligens är en bred palett av många olika förmågor ter sig tanken på att AI-tekniken trumfar oss i schack och kan skriva avancerade texter inte lika skrämmande. Utifrån ett brett forskningsunderlag förklarar Gärdenfors varför AI-tekniken inte kan och inte kommer att kunna tänka på samma sätt som människor och djur gör. »Peter Gärdenfors tilldelas Natur & Kulturs debattbokspris 2025 för att han fördjupar AI-debattens centrala begrepp och utmanar dess utgångspunkter. Med lätt språk och stabil lärdom blottlägger han tänkandets evolutionärt slipade mekanismer, och skärper bilden av vad intelligens är och vilken plats tekniken intar i vår digitala värld.« – Juryns motivering
[[Kan AI tänka by Peter Gärdenfors]] via Sven Dahlstrand, dahlstrand.net Publ okt 2024 Seeks to define what thinking actually is, and how that plays out in other animals and humans. The 2nd part goes into sofrware systems and AI and how they work in comparison.
Lisi Hocke on NewCrafts 2024 conf in Paris. Great images of handwritten notes too! #pkm