Channel 02 / 06

Skeptic's Toolbox

03 entries

The cases against default narratives; what doesn't hold up under scrutiny.

The trace

TIME REVERSE-CHRON

  1. [03]
    START HERE

    Your harness is half the model

    I ran four coder models through their officially-recommended harnesses and one came out 100x slower. The cause was my shell's rm -i alias leaking into Claude Code's tool subprocess. Your model comparison will credit that slowdown to the wrong variable.

  2. [02]

    Why My Local 80B Model Beats Claude at Coding Tasks, and It's Not About the Model

    A local 80B on two 3090 Tis finished the benchmark in 1:02 against Opus's 10:28, and the only thing I changed was the client. Claude Code re-serializes the whole prompt every turn, which kills prefix reuse on hybrid SSM models and costs 20 seconds of re-prefill per tool call.

  3. [01]

    When 84% Token Reduction Means Zero Accuracy

    Semantic code search cut my token spend by 84% on a real Laravel debugging task and found zero of the four files that mattered. I had been evangelizing the tool for weeks before I finally benchmarked it.

Other channels