This page looks best with JavaScript enabled

AI is Like Teenage Sex...

Newsletter Issue 13: Don't fall for the "AI-first" hype. Focus on being "data-first" and ML/AI will naturally follow.

 ·  ☕ 3 min read

I learned about this question in a panel discussion at a conference last week:

While most leaders acknowledge AI’s role in driving growth, companies are still figuring out how to change themselves to meet it. What is your advice on getting the scope of AI initiatives just right? How would you structure an AI-powered organization?

This question betrays a strange mix of two fears: FOMO and failure.

FOMO: Fear of Missing Out

“Most leaders acknowledge AI’s role in driving growth,” so most are doing AI! Why aren’t we? This instantly reminds me of Dan Ariely’s famous quote in 2013:

“Big Data is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it.” — Dan Ariely

In the last decade, most companies have figured out how to do Big Data. But just replace Big Data with AI, and it holds true today.

Many big companies want to be seen as using AI. Most startups want to claim to be “AI-first”, whatever that means. In fact, companies with .ai domain raise 3.5x more money.

FOMO is not just real, but it is incredibly irresistible.

Fear of Failure

Notice “getting the scope of AI initiatives just right.” Mind you, that restrain is despite most leaders acknowledging “AI’s role in driving growth.”

The addendum to that teenage sex quote is that nobody really believes everyone’s claim! Deep down in their heart, they know that only a very few are able to do it.

Remember all those studies show that 78–87% of AI projects fail? Maybe that is what is playing on people’s minds.

So, What Should You Do?

First and foremost, ask, do you really need AI/ML? Don’t try to somehow force-fit AI into your system.

If you do need AI, then you first have to start collecting usable data.

If you have been already collecting data, try to maximize the return on that investment. More on it in the next section.

Finally, start with small steps. Walk before you run. The AI Hierarchy of Needs by Monica Rogati is a good guide to follow. Don’t try to directly jump to the top of the pyramid.

The Machine Learning Hierarchy of Needs
The Machine Learning Hierarchy of Needs

Source: Monica Rogati

How to Maximize RoI on Data Investments

Most medium and big companies have been collecting and using data for about a decade. They have been using some Business Intelligence (BI) tools.

How to maximize your decade’s worth of data investments? Don’t wait till you hire and set up your ML/AI team. Evolve your Data Analysis team to start doing data science.

Modern Data Warehouses have capabilities to train ML models using SQL (e.g. BigQuery ML, RedShift ML)

SQL for Data Analysis is pretty good to start with. You should check out Chapter 9 of the BigQuery book covers building regression and classification models, k-mean clustering, recommender system, and doing the hyper-parameter tuning. (I am not being paid to recommend these books.)

You will be amazed at how much ML you can accomplish right away in your data warehouse.

Books on Data Analysis and Machine Learning with SQL

Share on

Satish Chandra Gupta
Satish Chandra Gupta
Data/ML Practitioner