Should AI or Humans Make the Final Decisions?Here is my “not-so-common” answer — my two cents.
Admond LeeBlockedUnblockFollowFollowingApr 14Long story short.
Few days ago I attended a sharing session by an eminent professor and one thing captured my attention (which has already been mentioned as the topic’s name of this article).
One interesting thing was that this topic during the sharing session turned out to be an endless (almost) question-answer-question kind of conversation between attendees and the professor.
This made me think deeper and subsequently consolidated my thoughts to answer this question here from the perspective of business and technical standpoint.
So let’s get started!First and foremost, the question itself is not complete to give a full picture for an answer…Should AI or Humans Make the Final Decisions?In my opinion, there is no right answer to this question due to the ambiguity and generality of the question itself.
And if you discern carefully, there are two elements lacking within the question — What and When.
Okay, let me explain.
First of all, the question doesn’t have any context or particular business problem to solve (which is the “WHAT”) that leads to an answer whether AI or humans should make the final decisions.
Second, even when we know the “WHAT” yet we still need to know in which particular situation (which is the “WHEN”) to be considered.
Therefore, to reframe the question, it’ll be this:“When” should AI or Humans Make the Final Decisions“to Solve a Business Problem”?This re-framed question makes it much easier for people to interpret and come up with a more conclusive answer instead of another answer begging another question.
Again, this is just my opinion and it’s indeed open for discussion.
Leave your comment below!1.
Why context or a business problem (the “WHAT”) is needed in the question?We all know that all companies have business problems to solve.
However, AI is unable to make decision for all business problems and there needs human intuition to make the final decisions.
In other words, to know if AI or humans should be the one to make the final decisions, we first have to know what exactly the business problems that we’re trying to solve.
For example, we delegate decisions to AI when it comes to choosing the best route to your destination using Google Maps.
It’s nearly impossible for humans to make accurate justification and selection without intensive computation within a short period of time.
In this context, AI should be the one that makes the final decisions for us.
On the contrary, we also have to understand that certain problems formulated have constraints and models (so called AI) can no longer be generalized to every single problem based the learning from historical data.
Take the recent case as an example, Amazon had been using their hiring algorithm to scan and review job applicants’ resumes since 2014.
After some time the Amazon’s machine learning team found out that the hiring algorithm showed bias against women.
This was caused by feeding the model with training data that mostly came from men — a reflection of male dominance across the tech industry.
Again, human control is still needed at certain cases and AI is meant to empower humans instead of replacing them.
2.
Why a situation (the “WHEN”) is needed in the question?Now that we’ve known what business problem to solve.
What’s next?Yes, we still need to consider a particular situation and when it occurs.
What do I mean by that?Let’s refer back to the Amazon’s example given earlier.
Say now the question becomes:When should AI or humans make the final decisions to decide which candidates to hire?One of the answers could be:AI should make the final decisions “to decide which candidates to hire” “when” the AI system shows no discrimination.
Or:Humans should make the final decisions “to decide which candidates to hire” “when” the AI system shows undesirable bias.
We have just answered the question based on the business problem and situation together.
Either one, the answers are clear and straightforward.
Zero ambiguity.
Final Thoughts(Source)Thank you for reading.
Personally, I find this question a thought-provoking one and therefore feel the need to share my ideas in this article.
I hope that by answering this question from a different angle gave you another perspective even before trying to answer a question.
This will make your answers more clearer to answer the real questions.
As always, if you have any questions or comments feel free to leave your feedback below or you can always reach me on LinkedIn.
Till then, see you in the next post!.????About the AuthorAdmond Lee is a Big Data Engineer at work, Data Scientist in action.
He is known as one of the highly sought-after data scientists and consultants in helping start-up founders and various companies tackle their problems through business and data strategy with deep data science and industry expertise.
He has been guiding aspiring data scientists from various background to learn data science skills effectively to ultimately land a job in data science through one-to-one mentorship and career coaching.
You can connect with him on LinkedIn, Medium, Twitter, and Facebook or book a call appointment with him here.
Admond Lee Kin Lim – Big Data Engineer – Micron Technology | LinkedInJoin LinkedIn WHAT I DO:‾‾‾‾‾‾‾‾‾‾‾‾‾Big Data Engineer at work, Data Scientist in action.
I help start-up founders and…www.
linkedin.
com.. More details