Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
How Generative AI Helps Predictive AI
 Originally published in Forbes, August 21, 2024 This is the...
4 Ways Machine Learning Can Perpetuate Injustice and What to Do About It
 Originally published in Built In, July 12, 2024 When ML...
The Great AI Myth: These 3 Misconceptions Fuel It
 Originally published in Forbes, July 29, 2024 The hottest thing...
Where FICO Gets Its Data for Screening Two-Thirds of All Card Transactions
 Originally published in The European Business Review, March 21,...
SHARE THIS:

10 years ago
Which Came First? The Chicken or the Egg… Data Science Can Help

 

Imagine that you are a professional observer with a magic telescope. Like ordinary telescopes, it lets you see things that are far away, but this one is attached to a magic screen, so I can use it to tell my story.

On your magic screen, you see a farm in a distant land. Each morning you observe the farmer feeding his chickens a measure of food, and each evening he returns to harvest about a dozen eggs from the coop. The behavior is repeated daily for quite some time.

One day you observe that the farmer has doubled the amount of food he is feeding his chickens. Somewhat surprisingly, the output of eggs virtually doubles. Since you have chosen to scientifically observe this farmer’s behavior, you and your team of experts conclude that somehow the farmer has figured out that doubling the amount of food will yield approximately double the amount of eggs. You go on and theorize, from what you have observed, that the marginal gain from egg production exceeds the marginal cost of chicken feed, and the farmer is now running a more efficient, more profitable operation. You decide to call this place, “The Egg Farm.”

What you failed to understand (because the magic screen does not let us hear conversations or communicate with what we are observing) is that doubling the amount of food a chicken eats also results in fat chickens. And since you didn’t know that eggs were in abundance in this farmer’s world, you could not know that market pressure had pushed the marginal cost of producing eggs past the point of profitability.

However, this particular farmer’s business advisers had told him that the price for chicken meat was on the rise, so he should fatten up his chickens, sell them, and become a reality TV star. He responded that he didn’t need to become a reality TV star because he had not taken any of the eggs to market since he started doubling egg production; he had put them in an incubator in another building and transitioned his business into a “Chicken Farm.”

The Age Old Question

It’s an age-old question: “Which came first, the chicken or the egg?” Or, better phrased for our example: “Which is more profitable, the chicken or the egg?” Is that the right question to ask? How about, “What business is this farmer actually in?” Or, “What is the proper product mix of chickens and eggs?”

Without data science, we can use one of two types of reasoning to craft a narrative about this farm.

  • Inductive reasoning with imputed values for missing data suggests that this business is an egg farm.
  • Deductive reasoning from deterministic data suggests that economic conditions forced a transition from egg farm to chicken farm.

Data Science

If only we had a data scientist available for consultation, we could get to work. Why? Data science allows us to combine hypothesis-based (deductive) reasoning and pattern-based (inductive) reasoning to extract actionable information from diverse data sources. You can think of it as a dynamic environment where static or empirically based models are iteratively transformed into new, better models. But data science doesn’t end there. Academically rigorous data science can propose questions you didn’t know you should ask, and use data to predict the future rather than report on the past.

Data Into Action

Data rich or data poor, companies with the greatest ability to turn data into action (those with the best data science teams) are going to win. Is your organization ready to build a data science department? Would your senior business leaders know what questions to ask your data scientists? Do you have a good idea what data is available for analysis? I’ve got several more questions to ask you about your data science readiness, and we can also discuss how to really use data science to determine which came first… the chicken or the egg?

Leave a Reply