Deft AI Lab Announced by Krishnan and Murphy
Cofounders release public beta of writing model from new AI lab.
Rohit Krishnan retweeted the launch announcement for Deft, an AI lab he cofounded with James Murphy. The post presents a small beta model built to produce higher-quality writing. Replies from researchers and writers note reactions to its ability to evade detectors, with some calling that focus a side quest and others welcoming tools that move past slop. One academic shared an example of output criticized for quality. The conversation among visible replies centers on the beta release itself and its stated goal of better writing.
Announcing Deft, a new AI lab for better writing, cofounded with @jmrphy See the picture for launch announcement the Deft model wrote for itself Currently, 86% of user queries are fully human according to pangram. This is still a small beta model and it might make mistakes.…
Reactions
15 postsRO 
Rosmine@rosmineAnnouncing Deft, a new AI lab for better writing, cofounded with @jmrphy See the picture for launch announcement the Deft model wrote for itself Currently, 86% of user queries are fully human according to pangram. This is still a small beta model and it might make mistakes.…
- likes: 1.4K
- replies: 138
- bookmarks: 1.5K
- reposts: 89
TC Tuhin Chakrabarty@TuhinChakrHate to be that person but researchers (including myself) have shown that its possible to write high quality text by fine-tuning an LLM that then fools Pangram. ALSO THIS IS GARBAGE WRITING. HOT MESS. Prompt in reply
- likes: 99
- replies: 6
- bookmarks: 92
- reposts: 2
NL Nathan Lambert@natolambertHappy to see this! The written word is extremely powerful and we’ll learn a lot about LLMs in making them master it.
- likes: 155
- replies: 7
- bookmarks: 68
- reposts: 4
NL Nathan Lambert@natolambertFooling AI detectors is silly to include tho, is a total side quest.
- likes: 24
- replies: 2
- bookmarks: 1
- reposts: 0
AC Andrew Curran@AndrewCurran_Detectors in ruins. This was always inevitable, one way or another. The moment models write at or above skilled human level they will be invisible to detectors, as the artifacts/patterns will no longer exist to detect. Good writing is just good writing.
- likes: 1.6K
- replies: 106
- bookmarks: 765
- reposts: 91
AV Alex Volkov@altryneLet the cat and mouse of AI writing begin
- likes: 6
- replies: 1
- bookmarks: 2
- reposts: 0
