Scribd
Upload
228 views1 page

Checklist ChatGPT

This document provides a checklist for determining whether a business process is suitable for using an AI language model like ChatGPT. The more criteria a process meets, the better the fit. The criteria include whether the process is repetitive, text-heavy, tolerant of minor errors, uses seed data, could benefit from drafts, has clear instructions, has a low cost of failure, and allows for human review. Meeting these guidelines helps ensure the model can effectively and safely assist with the process.

Uploaded by

Maynor Mijangos
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
228 views1 page

Checklist ChatGPT

This document provides a checklist for determining whether a business process is suitable for using an AI language model like ChatGPT. The more criteria a process meets, the better the fit. The criteria include whether the process is repetitive, text-heavy, tolerant of minor errors, uses seed data, could benefit from drafts, has clear instructions, has a low cost of failure, and allows for human review. Meeting these guidelines helps ensure the model can effectively and safely assist with the process.

Uploaded by

Maynor Mijangos
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Is this a good process for ChatGPT?

A check list for where AI language models like GPT and ChatGPT can help
Looking for more? Get practical advice, use cases and support on how to apply AI in your business or work.

Sign up for our free community at theaiexchange.com

How to use
Identify a process in your organizatio
Ask the following question
The more boxes checked, the more likely the process is a good fit for using an AI language model like ChatGPT.

Is the process repetitive?


The payoff for integrating AI into a process is higher when the process is more repetitive and frequent.

Is it a text heavy process?


Processes that require generating or altering text either with templates or custom writing done by people are good
candidates for AI language models

Are numbers and accuracy unimportant?


AI language models like ChatGPT are good at language, not math. If numeric equations and accuracy are important,
your process might not be a good candidate

Do you have seed information?


If you already have inputs or data fed into the process, and part of the process includes understanding, extracting
what's important, and synthesizing that information, your process would be a better candidate.

Is a draft valuable?
A good candidate is a process where getting a first draft would save a lot of time.

Does the process have specific instructions?


The more specific you can get about your ask of an AI model based on current instructions, the better prompts you
can write and ultimately the better results you can get.

Is the cost of failure low?


Language models are trained to be good at generating the most likely text, not to be factual. If your process has a
high cost of failure, or an “average” generation would be unideal, it might not be a good candidate.

Can you keep a human in the loop?


Adding a human review step at critical parts of your process will ensure you get higher quality results. A good
candidate should have this ability readily-available or be easily added.

Created by The AI Exchange, a community of business leaders and professionals actively applying AI to their work and
businesses. Join and get more tips like this straight to your inbox.

You might also like