Artificial Intelligence and Machine Learning IRL

October 18, 2021
By BYTEWRTHY
In Colloquy

AI Encounters

Machine Learning

Machine Learning, Deep Learning, Neural Networks, Artificial Intelligence: these are simply the names of predictive toolsets developed to answer questions such as

Is this review more likely to be left by a happy customer than by a disgruntled one?

Is that a person in front of my car?

You are subscribing to some form of AI and its future growth, like it or not… So let’s get the lowdown…

Many consumers who own a voice device use it to make select purchases. The popularity of Alexa and Google home (Siri is way behind) have added to the growth of voice-based search and purchase. Say you tell your phone “I am looking for Charlotte Tillsbury’s Airbrush Powder”. Natural Language Processing algorithms will transform your speech into words, will exclude “I am looking for” as extraneous noise, and will return the product which you need even if there is a slight misspelling in the product name.

Another famous example quoted of late is that of “Amazon Go”. A customer can visit the Amazon Go store, get a few items and ask Alexa to look for a recipe and the product recommendation engine can suggest ancillary products the customer might need. This integration allows Amazon to create an extremely detailed profile of their customer. As more customers shop at a Go store, Amazon collects every bit of data possible to build an even more robust object identification and people tracking algorithm to be leveraged in Amazon.com.

From the world of beauty (think Sephora & Coty) – Machine Learning can also be utilized to analyze an individual’s skin tone & type. A questionnaire first and then an algorithm analyzes the data to provide a skincare recommendation that aligns with the individual’s wants while considering their physical features. The beauty of AI is that it allows one to try on products virtually including lipsticks and eyeshadows – without having to do so manually. AI’s predictive analysis can also help customers pick a color they will likely be happy buying if past purchase habits are indicative of their future preferences!

Similarly, AI-powered chatbots can enhance the customer experience – They engage customers, and efficiently handle order queries, to minimize response time.

In recent decades, data has become abundant and processing power has grown by leaps and bounds, allowing for increasingly powerful tools such as Neural Networks to provide increasingly accurate answers (“predictions”) to increasingly complex questions. Still, algorithms are imperfect, and some may also have systemic limitations and biases which we have a long way to go before we understand them fully.

Share this post