The faster teammate is the worse teacher. a september 16 study had novices pair with copilot and…
the faster teammate is the worse teacher. a september 16 study had novices pair with copilot and with a human: copilot won the task and cut workload, but retest scores dropped.
speed and learning now point opposite ways.
Context
The paper Fast and Forgettable (arXiv 2604.18538, accepted at the ACM ICER conference, August 2026) is a within-subjects study of 22 university students who wrote Python HumanEval tasks for 20 minutes each in human pairs and individually with GitHub Copilot, and were retested individually a week later. Participants performed significantly better with Copilot than with a human teammate and several workload dimensions were significantly lower, while the human teammate's emotional effect was significantly more positive. On retest the paper reports a nonsignificant absolute retest performance reduction in the AI condition, with a larger decrement in that condition.
The study is small, with 22 participants in a convenience sample, one task set, one Copilot setup and a one-week retest, so nonsignificance means the drop cannot be generalized beyond this sample; it does not show the drop was absent. The observed descriptive retest drop in the Copilot condition is as reported, and the paper's introduction says participants learned comparably. Among the 11 stronger teammates the lower retest was marginally significant after adjustment, and among the 11 weaker ones it was not significant. The worse teacher framing is a generalization that the results do not support as a finding. The preprint dates to April 2026, and a 16 September study was not found.
Watch next
- The ACM version and replications with larger samples.
Sources
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 21 September 2026 at 18:03 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →