“After being a casual user of Tinder for a while, I realised just how much of my time was being consumed by swiping and responding to introductory messages,” says Justin Long, creator of Tinderbox. “Though they say the "chase" is half the fun, I try to hold more value for my time.”

With this in mind, Long created an automated piece of software that could attach to Tinder, not only making intelligent vision choices, but also holding down a conversation.

Tinderbox attempts to make predictions based on profiles you've previously liked by using elementary Machine Learning (facial analysis) methods. Tinderbox will examine previous likes/dislikes and develop a model based on colours and shapes found within faces of profiles. The more likes/dislikes made, the more accurate the model will become.

Tinderbox also has built-in automated messaging. The built in "Tinderbot" looks for conversations that fit patterns within pre-set message trees. The tree directions are based on positive or negative sentiment, and when the conversation no longer fits the tree the bot will notify the user to take over the conversation.

“There are many things that could be benefitted from AI, whether its personal scheduling or other communications,” says Long. “I think AI can be tremendously helpful in menial or marginal tasks, and it would allow us to spend even more time focusing on more quality experiences.”

This is certainly true; Google is currently working on a new artificial intelligence called Im2Calories that will use “sophisticated deep learning algorithms” to analyse a still photograph of food and detect how many calories are on the plate.