BEGIN:VCALENDAR
VERSION:2.0
X-WR-CALNAME:BrightTALK Event
PRODID:-//BrightTALK//NONSGML BrightTALK Event Calendar//EN
CALSCALE:GREGORIAN
METHOD:REQUEST
BEGIN:VEVENT
UID:https://www.brighttalk.com/webcast/9273/664192
DTSTAMP:20260615T175940Z
ORGANIZER:MAILTO:noreply@brighttalk.com
LOCATION:BrightTALK
URL:https://www.brighttalk.com/webcast/9273/664192?utm_campaign=commun
 ication_reminder_starting_now_registrants&utm_medium=calendar&utm_sour
 ce=brighttalk-transact
DTSTART:20260616T170000Z
DTEND:20260616T180000Z
SUMMARY:Live webcast: Structured Prompting 2.0: What the Research Actu
 ally Says
DESCRIPTION:Click here to attend: https://www.brighttalk.com/webcast/9
 273/664192?utm_campaign=communication_reminder_starting_now_registrant
 s&utm_medium=calendar&utm_source=brighttalk-transact\n\nPresenter: Pre
 senter: Lance Cummings, PhD. > Host: Scott Abel\n\nPrompt engineering 
 advice is everywhere. Ask AI to act as an expert. Add chain-of-thought
  instructions. Include examples. Write longer prompts. But which techn
 iques consistently improve results, and which are supported more by an
 ecdotes than evidence?\n\nJoin Professor and trainer Lance Cummings of
  the University of North Carolina Wilmington for a research-based look
  at how prompting practices are evolving and what tech writers need to
  know as generative AI use expands.\n\nLance explores a shift already 
 underway: the move from crafting prompts to designing information envi
 ronments. You'll learn why content structure, information architecture
 , and content type increasingly matter more than prompt wording, and w
 hy tech writers are uniquely qualified to guide this transition.\n\nDr
 awing on current research and lessons learned from a real-world disast
 er communication chatbot project, Lance will examine the strengths and
  limitations of popular techniques, including role and chain-of-though
 t prompting, few-shot examples, and prompt length. He'll also explain 
 why using AI to evaluate AI-generated content creates reliability prob
 lems and what a defensible testing process looks like.\n\nYou'll learn
 :\n\n* What research reveals about prompting\n* Why telling AI to act 
 like an expert changes style more than accuracy\n* How information typ
 es such as task, concept, reference, principle, and process improve pr
 ompts, system instructions, and agent behavior\n* Why AI-generated eva
 luation is often unreliable and how to build a testing approach\n* How
  to apply a repeatable rubric for assessing AI output quality\n\nCan't
  make it? Register and we'll send you a link to watch the recording at
  your convenience.\n\nThis presentation is brought to you by The Conte
 nt Wrangler and sponsored by Heretto, a powerful component content man
 agement (CCMS) system and content operations platform to deploy help a
 nd API docs in a single portal designed to delight customers.
SEQUENCE:1781628371
END:VEVENT
END:VCALENDAR