A Day in the Future
An advisor sits down in front of a monitor and drinks her morning coffee. He rdaily personalized news starts to come in. An alert pops up in the financial planning software: the Jones household is welcoming a baby daughter, born one week ago. The happy parents have been taking photos of her and sharing them on social media with trusted friends and advisors.
Artificial intelligence software connected to the advisor’s wealth management platform has recognized the new addition to the family. Based on this information, the advisor selects the family’s financial plan, and with a few commands sets in motion the starting investment for a 529. Although the advisor could have automated this step and allowed the software to create and fund the account, she prefers to be more hands-on with her clients. Everyone is on track.
Software is merely a tool – no different than a hammer or an industrial machine. It allows us to become more productive and connected to others that need our help. Image recognition, often referred to as Computer Vision, can help advisors provide better service by teaching them more about their clients. Whether highlighting life events or mapping out personal relationships, the power to understand and parse recorded visual images will bring unparalleled business opportunities and regulatory challenges.
Would you be surprised to know that, in certain categories, computers can categorize images more accurately than people do? They can find and track human faces, recognizing the age of the subject and their emotions. Machines can read handwriting, and mimic it, creating playful cursive scripts. They can tell apart hundreds of breeds of different dogs and un-blur blurred family photos. And they can do it at machine scale, enriching all the visual data that phones and wearable technology capture in our daily lives.
To get a glimpse into this future, try downloading the Google Photos app and connecting your digital photo library. It will store them on the cloud without cost or meaningful space limitations, and then unleash its army of helpful robots. These robots are called recurrent neural networks (RNNs), and each one is trained to find a specific type of image. As a result, your photos become searchable for objects Google finds in the images themselves. Instead of remembering the when and where to locate a photo, the user can search for a “red car” or “wedding night.”
Imagine the possibilities! The software can: follow a human face from childhood to old age, understand whether you live in the suburbs or the city, know whether you climbed Mt. Everest or renovated your home. The photos people take in their everyday life are a track record of their choices, personalities and creative potential. This is powerful, free stuff that exists today!
But photographs are just the beginning. Live, streaming video can finally be supported by the telecommunications infrastructure. Services like Periscope allow users to become instantaneous broadcasters, filming with their smartphones, attached to Twitter accounts, and distributed to everyone they know. Snapchat has created video-sharing functionality that combines videos from friends into an ongoing stream of media. Image recognition makes this overwhelming set of data searchable.