The Byron AI team is looking for 10-20 volunteers who are a) reasonably well-versed in a broad range of literary genres and b) are capable of quickly reading 500- to 1000-word chunks. No technical ability is required; what we’re looking for is people who are capable of reading an unidentified text sample and attempting to identify the author on the basis of the style. The performance is not important; we have a second-level test that is necessary for those readers who cannot identify the author on the first read, so the volunteers have to be able to resist the temptation to look anything up or resort to AI in order to identify the author.
That sort of artificial identification will pollute the data and we want to strictly avoid it. Byron AI backers are favored for this, so if you’re interested in volunteering, please email me with AI STYLE BACKER if you are a project backer and AI STYLE if you are not.
As is probably clear, the dev team did not wait for the campaign to successfully conclude before we got started on certain elements of the project. It’s still too soon to declare that we will succeed in the proof of concept, but the initial indications are positive.