Duck Times Section
Experiments from Duck Times Labs — software products built from our own reporting.
Full Fact publishes its sourcing trail, not just its verdicts. A workbench should hold itself to the same standard.
5 min read
A personalized feed proved the wrong problem. Continuity is the harder one.
7 min read
Most companies do not have a knowledge problem; they have a forgetting problem search alone won't fix.
6 min read
A coach that only validates you is a mirror with bad lighting — the job is to say so anyway.
6 min read
For You feeds and Reels are engineered to never run out. Apple's own Today tab proves a bounded, curated alternative already ships at scale — Labs is testing whether readers want that experience on purpose, not just by default.
Amazon's memo culture captures the decision. Duck Times Labs is testing whether a git-style changelog can capture the part almost no organization documents: why a settled decision was later reversed.
Tools like Roam, Obsidian, and Notion built devoted niches around manually maintained link graphs. Duck Times Labs is testing whether ChatGPT-style passive memory succeeds where deliberate curation didn't.
"Citation needed" has flagged unsupported claims inline on Wikipedia for two decades. Duck Times Labs is testing a version built for claim-level uncertainty in original reporting, not crowd-edited encyclopedia entries.
Duck Times Labs is building a public radar whose actual job is showing readers which spikes editors rejected and why — a job no existing trend tool, including Google Trends itself, currently does.