Zank here from Bennett Data Science.
A.I. is transforming business today and will continue to do so over the coming years. But many companies struggle to realize benefits of the data they capture. Are you interested in keeping up with trends and understanding how A.I. can impact your business?
That’s exactly what this newsletter is designed to help you do.
A bit of history: years ago, I created Tech Tuesday as a way to keep fellow colleagues apprised of developments in A.I. in a conversational way that informs business decisions. It was an immediate hit and garnered praise years after I stopped writing. Today, Tech Tuesday is back and better than ever!
Every Tuesday you’ll receive exclusive content and a curated list of articles designed to let executives and technical managers know what’s important in A.I. right now.
Of Interest
Deep Fakes:
If you’ve heard of deep fakes before, you may already be aware of just how spooky they can be. Here’s a video from Ctrl Shift Face that doubles down on the creepiness.The implications are huge, from generating fake news to adding celebrities to adult videos or creating social influencers who don’t even exist. There are now digital forensic companies springing up to work on detecting deep fakes. How should the public react? At very least, we should use care before believing anything we read, see or hear. And that type of vigilance comes with its own down side. Expect there to be a lot more A.I. movement in this space, especially as the 2020 elections grow nearer.
A.I Informs Effective Weight Loss:
Several years ago, Ariel Faigon kept a personal journal of all the food he ate, the sleep he got and each time he exercised or fasted. He used this information to show which behaviors increased or decreased weight loss. You can find all the details here: Discovering ketosis: how to effectively lose weight. In his very accessible writeup, he describes how he went from 143 lbs to over 190 lbs and what he did to lose the weight again. He uses an amusingly named and very capable tool, Vowpal Wabbit to effectively uncover the factors that accelerate weight loss, including intermittent fasting, increased fat intake and increased sleep duration. It’s a good example of how to use A.I. to uncover the factors that influence an output, aptly called variable importances. Be sure to take a look at his variable-importance chart.
Your Face Knows how you Should Dress:
In the late 1980’s a man named David Kibbe authored a book called Metamorphosis, a personal image and style book for women where he introduced a style type theory known today as Kibbe Types. They work like this: Given measurements of a person’s face, the person can be categorized as some combination of the five main Kibbe Types: Dramatic, Romantic, Classic, Natural, or Gamine. This article gets into greater detail of each Kibbe type. The central idea to Kibbe types is that each type is associated with styles that best compliment that type. The challenge is, it’s quite difficult for people to know what type they are. At Bennett Data Science, we’re working on an application where a user can take a selfie and immediately learn her Kibbe type with associated style recommendations. More to come in the following weeks.