Project Management

Please login or join to subscribe to this thread

How can a PM easily identify the errancy of a GenAI tools?

linkedin twitter facebook   Artificial Intelligence  
avatar
ABIODUN SORINOLU Director/CEO| COMRAGEINC Edmonton, Alberta, Canada
GenAL for Project Managers.
Sort By:
avatar
Rami Kaibni
Community Champion
Senior Projects Manager | Field & Marten Associates New Westminster, British Columbia, Canada
When you verify the responses provided by the AI tool. The PM has to verify all answers because at the end of the day the final say is for the PM, not the AI.
avatar
Ahmed Hassan Ali Hassanin Mohamed Mahboula, Ah, Kuwait
I think in the future; PM will not able to identify because our brains will be stopped and our total accreditation to follow what AI informed us to do.
and our task only how to present the information we got it in proper way/good presentation.
avatar
Florence Ndum IT Project Manager and Consultant| Independent Biberach, Germany
A very insightful tip I got from the 'Generative AI Overview for Project Managers' course is that you can review the output with your team, especially those who are experts in the topic in question
avatar
Kiron Bondale Retired | Mentor| Retired Welland, Ontario, Canada
Ahmed, at that point, the machines might as well take over if we abdicate full responsibility to them!

Abiodun, to add to what Rami has said, aside from the PM ensuring they and/or the team has sufficient domain expertise to verify AI tool outputs, ways to reduce the likelihood of errors are to ensure quality data is used for training and inputs and an appropriate AI model is selected.

Kiron
avatar
Adam Russell Principal Consultant (Project Management)| Tiligent Pty Ltd Putney, New South Wales, Australia
with ChatGPT, if I want to rely in the info, I ask it to provide citations or references for the information it's generating, then I check the citations independently, e.g. general google search or scholar.google.com for academic citations. But all the other methods that have been around for ages are available: review by experts, delphi technique, asking multiple sources (e.g. ask Bard and ChatGPT, ask multiple times. Also, if you have access to the OpenAI API there are more parameters available that can control the response. I wrote some Python code to do this, but I was playing the other day with the Zapier GPT interface hooked into a Google Form and that looks like it works ok. Work in progress though.

Please login or join to reply

Content ID:
ADVERTISEMENTS

"Happiness is good health and a bad memory."

- Ingrid Bergman

ADVERTISEMENT

Sponsors