favorites

Creators should not expect success, but keep creating and refining their craft in quiet times.

Larry Wall's "laziness" is a virtue that demands hard intellectual work — it drives programmers to create crisp abstractions to spare future toil. LLMs, unconstrained by time, do not seek simplicity but pile on garbage. Tools should serve our virtuous laziness, not let performative industriousness run rampant.

Based on Microsoft's Lee et al. (2025) study, the article argues that user confidence in AI correlates with declining critical thinking. When tools respond quickly and confidently, even experienced professionals stop verifying and questioning, becoming 'automation operators' rather than investigators.

Being too ambitious is a clever form of self-sabotage. Volume always compensates for quality — lower your standards, embrace the beginner's mind, focus on the process, and treat failure as information.

In the current era, experienced engineers are becoming increasingly devalued. The job market is no longer filled with opportunities as economic downturn meets the AI revolution.

Using historical cases like GMO food aid, Germany's nuclear exit, and the Philippines' unused nuclear plant, the article argues that techno-pessimism drives us from speculative future risks toward proven past dangers. Cites Jill Lepore's 'Frankenstein fallacy' to show reductive dystopian narratives are politically useful but constructively bankrupt.

Spotting new-school bullshit dressed in math and statistics. From ICE Score to Virginia Mason Quality Equation, the author dismantles formulas that look rigorous but mean nothing, offering three fixes: compare orders of magnitude, avoid numerical quantities, and assign scores to buckets consistently.

Nathan Marz shares the three-phase methodology from building Storm: first make it possible, then make it beautiful, then make it fast. Only invest in abstractions when you feel the pain; premature generalization breeds complexity.