You may have come across our announcement of Full Edit, our newly added language feedback mode. Full Edit offers edits of unprecedented quality, leaving other language tools far behind. Here’s how.
New language models
Full Edit uses a new set of Artificial Intelligence-based language models that are incredibly powerful. While more language tools now use AI, their tech doesn’t get to this level. Errors are commonly missed and incorrect suggestions are often given, especially when sentences get highly technical or scientific. Full Edit’s language models are simply wired differently, offering the highest performance to date in both coverage (how many errors are caught) and accuracy (how well these are fixed).
Pure Artificial Intelligence
Full Edit doesn’t use hard-coded rules to give language feedback: all edits are offered by Writefull’s algorithms. This is exceptional, as, to our knowledge, no other language tools are AI-based end to end - all use grammar rules to some extent. Those are triggered when a writer makes a common error, and a hard-coded replacement is then offered (for instance, a rule could be: ‘replace researches by studies’). Rules work to some extent, but fail to correct errors that are less frequent or that involve language usage rather than grammar. As Full Edit uses AI only, it corrects anything that doesn’t flow or is unclear, redundant, or ungrammatical.
Going beyond the traditional language check
As Full Edit screens language as a whole, it corrects much more than grammar, vocabulary, punctuation, spelling, and style - the categories that Writefull covered to date along with other language tools. Here are some of the new types of feedback it gives:
1. Ungrammatical or awkward phrasing
2. Word order
5. Fluency / language use
Convinced to give it a try?
About the author
Hilde is Chief Applied Linguist at Writefull.