AI Programming on Lynx Tech Blog

274 Commits in a Month: I Shut Down More Projects Than I Started

Mon, 10 Aug 2026 10:00:00 +0800

First, Let’s Put the Numbers on the Table Over the past 31 days, I sampled a dozen or so repositories I have on hand: 274 commits. The most active one, lynxhot, had commits on 21 out of those 31 days. A blog that was only set up on August 3 racked up 100 commits in 8 days. On top of that, there are still more than a dozen research reports sitting on my hard drive—from Tesla’s open-source approach to car building to layered trash bags, from a content analysis of 371 Zhihu answers to the whereabou

Prompts, Context, Harness: How the Focus in AI Has Shifted Three Times Over the Past Four Years

Mon, 10 Aug 2026 04:30:00 +0800

First, an Experiment In early 2026, developer Can Bölük ran an experiment: same model, same set of tasks, nothing changed except the engineering layer around the model—specifically, the format used by the harness to handle code patches. The task success rate jumped from 6.7% to 68.3%. Tenfold. Not a single line of the model changed. That number has been making the rounds in AI circles lately because it turns something many people had vaguely sensed into a conclusion that is impossible to ignore:

The Development Curve of AI: Steep Early On, Explosive Later

Fri, 07 Aug 2026 12:00:00 +0800

Two Completely Different Curves The difficulty curve of traditional software development is roughly a gentle start followed by a steady climb: setting up the environment, learning the framework, writing the scaffolding—one step at a time. Later, as the codebase grows, the difficulty does increase, but you know exactly where every bit of that difficulty is coming from. AI-assisted development follows an entirely different curve. The early phase is terrifyingly steep—a demo in a day, a prototype i