Knowledgebase

Training People to Use AI Responsibly Print

  • aiethicsresponsibleuse, troubleshooting, guide, howto, solution, zillionkinghost, hosting, support
  • 0

What they need to know.

THE FOUR THINGS THAT PREVENT MOST PROBLEMS

It produces plausible text, not verified facts It knows nothing about your situation unless told The first output is a draft Everything factual needs checking

THE MOST EFFECTIVE DEMONSTRATION

Ask it about something the audience knows well.

Let them find the errors themselves. That teaches scepticism faster than any warning.

WHAT TO TEACH ABOUT CONFIDENTIALITY

Specifically what may and may not be sent.

General caution produces inconsistent behaviour.

WHAT TO TEACH ABOUT ACCOUNTABILITY

That it is unchanged.

WHAT TO PROVIDE

The policy Approved tools Verification standards stated concretely Someone to ask

WHAT NOT TO DO

Mandate use without training Assume people will work it out Leave it to whoever is interested

FOLLOWING UP

After a month, ask what people are actually doing.

The answers usually differ from the policy.


Was this answer helpful?
Back

Are you happy with your experience? Leave us a review on Trustpilot.


Trustpilot