Curated public example
Writing & design · Workflow
Published by Robin Sloan
This credits the original publisher. Better Loop membership or a shared assessment is not implied.
Sloan built an Atom editor plugin offering inline completions from a recurrent neural network trained on old magazine text. His public account describes the working interaction and concludes that its creative effects did not yet justify the effort.
Evaluate suggestions against your creative purpose and retain the freedom to reject them; novelty is different from better writing.
This is a 2016 author experiment using a custom RNN, not a modern assistant benchmark. The author reports insufficient benefit at that stage; no finished publication or productivity gain is established.
AI use: Documented in the source.
Sloan documents a custom recurrent neural network served through torch-rnn-server and used in Atom with rnn-writer. No exact model identifier is supplied.
A useful public example is not an assessment of a reader, a publisher or a Better Loop member.
Authored practice suggestion
Use material you are allowed to work with. This suggestion is preparation; it does not record a completed task or an improvement.
The retained continuation serves the stated purpose and voice; explain the rejection or revision of generated suggestions.
Private practice needs no account. Your configured model provider may process the evidence you select.