AI for furniture shopping!

As I wrote in my previous blog The Race of Cloud Solution for Machine Intelligence, we are about to enter an era that machines are learning to think intelligently and AI will be everywhere, for everyone.

Well, this is no dreaming. Besides serving as office assistants, marketing experts, Bots can now help you to buy furnitures and find the perfect car to buy. Soon we may all have our own J.A.R.V.I.S.

In life, we may come across things that we like particularly. Imagine if you can simply point your phones at something, and have it say, here’s what it is and here’s where to buy it. Well, Grokstyle is applying its machine vision API to find home decor and furnitures. Think is as the Shazam for image and shopping.

According to Techcrunch, Grokstyle started with a grant from the National Science Foundation for $225,000, and have parlayed that into a total of $2 million from a variety of firms and angels, unveiled in April 4th. CB Insight also select Grokstyle as one of its AI 100 from over 1800 AI startups. The cofounder, Kavita Bala, indicated that the current plan is to push hard into the furniture and home decor. However, the algorithm is generalizable.

On April 6th, The automobile research portol,, announced that it has incorporated Natural Language Processing (NLP) into its new search engine to make it truly intelligent.

Unlike the previous filter-search system, users can simply type or speak the car they’re looking for. Then, Carjojo automatically scans the internet for every new car available for sale with suggested brands and models, and returns every relevant car in the user’s area.

Moreover, last week, Talla introduced its office assistant AI — ServiceAssistant to support internal service team, such as HR or IT. Talla interact with users in systems such as Slack and Microsoft team.

It is quite true that internal service teams of large companies are bombarded by questions and help requests — often the same basic questions like “how do I reset my password?” or “do we have President’s Day off?”.

Talla can retrieve questions from Wikis and FAQs by interacting with users in natural language. By leveraging the cutting edge machine learning technology, it learns quickly overtime and ensures that administrators don’t have to answer the same question twice.

As entrepreneurs are enthusiaticly changing the world with machine intelligence technologies, tech giants such as Google and amazon are racing to become the leading enablers of the global Machine Intelligence movement. They all understand that developers will be the key to win this AI race.

After acquiring Kaggle in March, Google made another move to enhance its position in the developer community: on April 7th, Google DeepMind open sourced Sonnet, its object-oriented neural network library, on top of TensorFlow.

This is a smart move indeed. DeepMind explained that making Sonnet public allows other models created within DeepMind to be easily shared with the community. As the machine intelligence community becomes acquainted with Sonnet, the DeepMind’s internal libraries, they can also more feasibly contribute back by utilizing Sonnet for their own work.

Note that Sonnet is not intended to replace TensorFlow. Rather, it is designed to work with TensorFlow, so can be mixed with raw TF code.

In December, 2016, Google also open-sourced its flagship platform DeepMind Lab, and is currently working with Blizzard to develop an open source API that supports AI research in StarCraft II.