Applications

Enter the predictive era with AI!

In this article, we will attempt to define what Artificial Intelligence is, the role data plays in it, understand why Machine Learning is the future of AI, and discover the use cases that are already revolutionizing the daily lives of our schools, hospitals, local authorities, and businesses.

“Artificial intelligence is when a machine tries to do what a human does, but not as well.”

This definition, which might seem counterintuitive or even crudely iconoclastic, is actually one of the most precise ones available. It makes it easy to distinguish what is and isn't AI in its goal of replicating human cognitive processes. For example, a super-calculator is not AI; a computer will always be better than a human at performing a mathematical calculation. Conversely, software capable of systematically recognizing a person in photos contains AI, because this is a typical case where a human would be better at the task. If an AI is systematically less efficient at its task than a human could be, what is the point? Scaling! What a human does perfectly, an AI will do almost as well, but most importantly, it will do it millions of times more in the same amount of time. A super-calculator whose calculations allow for photo recognition is one of the countless AI use cases that are part of our daily lives.

The deployment of AI solutions is made possible by the multiplication of the data we produce and the explosion in the computing power of our equipment. This datafication digitalization of society, along with a recent algorithmic maturity, allows for a decisive turning point for businesses: moving from the "rear-view mirror" management that has prevailed until now (drawing lessons from the past) to a predictive vision. 

To better understand how our data allows us to make predictions, we must first distinguish the four levels of data analysis: observe, understand, predict, and prescribe.

  • Observe: the famous rear-view mirror that allows us to identify what happened through a detailed analysis of historical data.
  • Understand: the higher level of analysis that allows us to draw concrete lessons from data in order to understand what happened. 
  • Predict: the first level of analysis turned not toward the past, but toward the future: being able to anticipate what will happen in a probabilistic way. 
  • Prescribe: the final level of data analysis, which allows businesses to make the best decision by leveraging what they have learned.
Take your data out of the fridge, Mick Levy, Dunod Publishing, 2021

The predictive and prescriptive levels are made possible by the advent of Machine Learning tools. The concept of Machine Learning relies on three phases: learning (the machine recognizes patterns and correlations in historical data), the inference phase (the machine applies what it has learned), and the supervision phase (managing the lifecycle, model performance over time, and potential retraining).  

These Machine Learning models now allow businesses to enter the predictive era and deploy three types of operational AI solutions: forecasting & optimization, hyper-personalization, and automation.

Forecasting & optimization

Accurately anticipating the needs, risks, and failures of an industrial process in order to implement appropriate actions in the right place at the right time. 

Example: Predictive maintenance (Industry)

Detect structural defects, premature wear, and performance loss in machinery to prioritize maintenance operations and maximize production.

Hyper-personalization

Anticipate a user's specific needs based on their data and provide tailored suggestions to assist them in their daily life.

Example: Personalized learning (Education)

The goal of this application is to support students like a virtual tutor and recommend content based on their progress, cognitive profile, and learning gaps.

Automation

Relieve employees of tedious tasks that require numerous iterations without human added value in the process.

Example: Energy management for building portfolios (Energy)

The objective is to develop an energy manager that automates the supervision of a building portfolio, tracks any consumption drift, and forecasts usage to improve overall energy efficiency. 

At Craft AI, we are deploying over a dozen use cases in key sectors of our society, such as energy, education, and healthcare.

This is possible because we are approaching maturity on the algorithmic side and are able to collect sufficient data. However, a third ingredient is missing to truly speak of a revolution: usage.

There is a deficit in the use and adoption of AI in companies, as noted by Gartner in a study : 85% of AI projects do not make it to production and generate no ROI. Two main reasons can explain this lack of adoption:

  • The difficulty of scaling AI projects (the infamous "production wall") 
  • Societal distrust of AI and the fear of machines making decisions for us.

If addressed correctly—the former through MLOps, the latter through trustworthy AI—you will be able to easily develop multiple use cases and fully enter the predictive era.