Tech for Business podcast

Prompting 101

0:00
12:00
Recuar 15 segundos
Avançar 15 segundos

Cole, CIT’s Power Platform Developer, to recap a recent Prompting 101 webinar focused on CRIT prompting—Context, Role, Interrogation, and Task—and why being specific in each area leads to better AI outputs. Cole explains each CRIT component, highlights common prompting mistakes like over-instructing the AI on how to reach an answer, and discusses how to address bias by explicitly naming it in prompts so the AI can avoid reinforcing it. He also answers how much context is too much, noting that while you can’t give too much in a single prompt, long multi-phase threads can introduce irrelevant context due to model limits. Looking ahead, Cole expects prompting to remain important while evolving toward reusable AI “skills” for consistent outputs such as CIT’s readiness assessment reports.

00:00 Webinar Recap Intro

00:24 CRIT Prompting Framework

02:28 Common Prompting Mistakes

03:37 Bias and Output Steering

05:12 How Much Context Is Too Much

06:21 Future of Prompting Skills

07:40 From Prompts to Skills

11:23 Wrap Up and Next Steps

11:40 Outro

Check out the full webinar here: https://www.citsolutions.net/events/webinars/ai-starter-series/

Mais episódios de "Tech for Business"