https://www.digital-science.com/blog/2018/02/robotics-takes-scientific-publishing/ Advancing the Research Ecosystem Mon, 14 Aug 2023 15:18:31 +0000 hourly 1 https://wordpress.org/?v=6.7.2 https://www.digital-science.com/blog/2018/02/robotics-takes-scientific-publishing/#comment-852 Tue, 20 Feb 2018 01:26:26 +0000 https://www.digital-science.com/?p=28372#comment-852 It is both expected and understandable that machine-generated writing would be applied to regular scientific writing. However, based on experience in medical communications, it also raise a number of interesting issues, including: 1.) For more complex material, there may be a point where specification of parameters to generate the writing piece and need for revisions might take more time and resources than required for the author/writer to generate the piece. What is the point of negative return?; 2.) In a climate of full disclosure, how would this be handled for an AI-generated draft? Would we have to specify the amount of the draft that was machine-generated? Would the AI be credited in the acknowledgments section?; 3.) It is becoming increasingly difficult to have manuscripts accepted and published by top-tier journals, eg, NEJM. Would a machine-generated assist for a manuscript be acceptable to such a journal? Would there be a test by another AI to score the piece and a limit be placed on the machine content?; 4.) For sponsored clinical trials, would a sponsor be willing to use AI assist for expediency with so much on the line for multiple stakeholders in communicating about the development of a new treatment?

]]>