Building An Unbiased AI Algorithm
The call to develop and include AI functionality in all aspects of life is growing at lightning pace. I find it interesting that with all the noise in the calls for AI, there remains so many skeptics. Is it fear of the unknown? Or could good old science fiction movies be responsible in part for these misgivings? Ultimately, as humans, our biggest #fear is of that which we do not know or cannot see.
So, what does AI offer us to help overcome our fear?
In conversations as well as many presentations I have witnessed, it seems the collective concern is that AI, by inference 'they' mean 'robots/machines' will take over our jobs. Question: If my writing this article is a job, can a robot/machine take over?
Many articles are now available, and more are being released daily to help improve our understanding of AI. I have decided to also start publishing articles on the subject. So far, I have given lectures and presentations on AI in two forms:
How to build an 'unbiased' algorithm for Machine Learning, and
To teach high school students how to use AI responsibly when completing school assignments.
Unfortunately, many #educators are still of the opinion that #students must not use #AI tools. I disagree with this notion, hence I have developed:
How to present your work/assignment for which AI was used.
A model that advises the elements to consider when designing an algorithm for data
Artificial Intelligence is our future. Understanding the difference between asking the right question, programming a machine to carry out a set of tasks, and the ultimate goal, the machine being able to act on collected data to better support our lives, I believe form part of our immediate challenges.
Do you really know how to ask a question? Will the data that we request, seek and collect provide accurate and balanced, verifiable answers to the questions asked?
In my model, I argue that to build an algorithm that works, we must include:
Accessibility, Accuracy, Culture, Data Validation, Ethics, Evidence, Privacy, & Responsibility.
Ultimately, it is time for governance. I will expand on this on a separate post.
- By Claudius Thomas
Click the below link to read the article published on LinkedIn:
https://www.linkedin.com/feed/update/urn:li:activity:7213936156239872001/
