Notes
Slide Show
Outline
1
Standards for ChatGPT-Enhanced Technical Writing:
Objectives, Obstacles, Options, and Opportunities
  • Dan M. Davis, Catholic Polytechnic University, USA
  • Co-Authors: Dr. Jennifer Nolan, Judith Jacobus & Dr. Mark Davis, USA
  • 2024-SIW-Presentation-008
2
 
3
Slide based on “Simple List”
  • Writing takes up a significant part of STEM professional’s day
    • Most of it is straight forward, but some is very stressful
    • Some is very formal (e.g. Proposals)
    • Some is very informal (e.g. eMails)
    • Very little is hand-written anymore
  • Little of Tertiary Education Focuses on Writing
    • One English Class as an incoming freshman
    • Maybe one tech writing course in grad school
    • <4% of college
  • Most did not get much help outside the classroom
    • There are classes offered at Univ of Chicago for newly hired PhD
    • A CEO of a major text firm says he has to give out Strunk and White


4
 
5
Is this Really a Problem?
  • Is writing an out-dated skill?
    • Quill pens and broad swords
    • Star Trek Computer
    • Abilities on the command line and writing program code
  • Writing and Language are drivers of analysis?
    • G-Factor testing relies on verbal abilities
  • Will needed skills Atrophy?
    • SciFi examples (Wall-E)
  • May skills be required in other environments?
    • If computer is not available
    • If societal infrastructure fails



6
 
7
 
8
Skills Needed for Communications
  • Verbal Skills
    • Grammar
    • Organization
    • Articulate
  • Physical Skills
    • Typing
    • Hand writing
  • Audience Awareness
    • Tone
    • Knowledge level
    • Biases
  • Types of written products


9
 
10
Analogy of Cinema Production and Technical Writing
  • In Cinema Creators
    • Concept creator
    • Writer of the dialogue
    • Cinematographer
    • Director
    • Music composer
    • Actor
    • …
11
Computer Skill sets for Better Writing
  • Computer editing
    • No longer need precise typing skills
    • Can now make corrections without leaving record of correction
      • MS Theses and PhD dissertations used to require perfect copy
      • No white-out, no correction tape
  • Computer Spelling Checkers
    • Requires a less time-consuming of work
  • Computer ;Grammar Checkers
    • Pick up most blatant errors such as use of wrong verb
  • Simple phrase Detection and Insertion of Appropriate Word
    • Word completion
    • Word choice offered



12
Intelligent Text Generation
  • Caveat
    • This presentation in NOT about ChatGPT
    • The Authors are experts therein
    • They are not even experienced  users
  • It is about the impacts
    • What might be in the future
    • What responses may be needed
  • Disciplines of interest
    • Ethics
    • Communication efficacy
    • Evaluative reliability
    • Danger of creation of false impression
13
ChatGPT Overview Intro
  • Provides readable text
    • Generates written material
    • Uses AI Techniques
    • Evolved over last 30 years
    • Available on-line
  • Input
    • Via verbal prompts from user
    • Typing, oral input feasible
    • The entire web
  • Series of Prompt
    • Do re-writes
    • Change tone
    • Change length


14
 
15
Potential Problems with Using ChatGPT
  • Ethics of Misrepresentation
  • Plagiarism misleads potential Users, Funders or Collaborator
  • Careful scrutiny of concepts may be discouraged
  • Cognitive stimulation nature of drafting lost or reduce
  • Fails to give credit to creators of original authors
  • May not apply human analysis of reliability of the source
  • Pressure to stay on schedule diminished
  • Potential violation of copyright laws amplified




16
Professional Standards
  • Need to mitigate harmful impacts


  • Disclosure
    • In acknowledgement Section?
    • Before oral presentations?
    • Who should promulgate rules?


  • Restricted uses
    • Are spell-checkers a significant issue?
    • Which uses really do have significant impacts/
    • Are some applications of this technology the future?
    • Will techies be tempted to give up skills they have developed?


  • Sanctions
    • Current plagiarism insights
    • Data manipulation
    • Enforcement issues



17
Conclusions
  • Computer-Generated Text will NOT go away


  • The unfettered use thereof may have deleterious effects


  • A Reasoned Review of Options may be in order
    • Disclosure rules
    • Levels of use (e.g. spell-checker vs.cut and paste “authorship”
    • Prohibited uses
    • Sanctions


  • Standards community has skills and experience others don’t
    • Cross platform studies to insure consitency
    • Unbiased approach to all stake-holders
    • Willingness to rely on others more versed in individual skill sets
    • Manifest “no dog in the fight”



18
 
19