Amazon says ‘aye’ to AI
Artificial intelligence is changing business models all over the world. Balqis Lim finds out how Amazon.com is embracing AI
ARTIFICIAL intelligence (AI) is where a computer or computercontrolled machine is able to perform tasks generally associated with the intellectual processes characteristic of humans. These machines are seeping their way into our lives, affecting how we live, work and entertain ourselves.
From simple tasks such as moving stuff to carrying out complex ones such as solving mathematical problems, the revolution is vast and fast emerging around the world.
Amazon.com, the largest Internetbased retailer in the world, also provides on-demand cloud computing platforms via its subsidiary, Amazon Web Services (AWS).
In maximising the value of its facilities, Amazon has spread its wings into AI for the past 20 years to optimise business operations.
AWS head of Emerging Technologies (Asia-Pacific) Olivier Klein says Amazon has a rich heritage when it comes to AI implementation across the company.
“At the beginning of our retail business, we optimised robotic picking routes. Our Alexa, for example, is an intelligent personal assistant made popular by Amazon Echo devices. You can use it to search the web, order food, or even shop online.”
Within AWS, Klein says the company is focused on bringing that knowledge and capability through an array of services. It’s dubbed the Amazon AI.
INTELLIGENT SERVICE
Amazon AI is the intelligent service powered by deeply customisable machine learning and AI platform.
It is also designed to be scalable to match individual business needs and requirements.
But this AI is not capable of thinking on its own. It requires human assistance to be pointed in the right direction to complete tasks.
Klein says AI should not simply be looked at as a replacement for people in the workplace. Instead, it helps perform some of the mundane activities.
“For example, in a fulfilment centre, AI can optimise picking, inventory accuracy and reorganise data to accelerate the order fulfilment process. It functions collaboratively with human workers. Companies will still need human force to operate the AI systems, ensure accuracy and pack orders.
“It does not mean companies will stop hiring. Eventually, AI helps improve accuracy and boosts productivity.”
Facial recognition in machine learning models.
Amazon Go plans to introduce cashier-less shopping experience.
THE AMAZON AI STRUCTURE
Klein shares some of the AI services and platforms available: They include Amazon Rekognition, an image analysis and facial recognition service; Amazon Polly, a textto-speech service which lets users create applications that speak in over two dozen languages; and Amazon Lex, which uses automatic speech recognition and natural language understanding to create conversational interfaces, commonly called chatbots.
For developers and customers who want to focus on building custom models, the AI platform services include Amazon Machine Learning, which provides visualisation tools to analyse data without the need to learn complex algorithms.
The Amazon Elastic MapReduce (EMR) is a managed Hadoop framework designed to perform log analysis, web indexing, scientific simulation and other big data cases.
Meanwhile, Spark runs inside of Amazon EMR as an open-source, distributed
processing system for big data workloads.
Amazon has made available applications and tools to get started on AI.
AWS Deep Learning Amazon Machine Image (AMI) is pre-installed with software frameworks such as Apache MXNet, TensorFlow, Caffe2 (and Caffe), Theano, Torch, and Microsoft Cognitive Toolkit.
According to Klein, the AI solutions are currently in use and service a host of clients worldwide.
AI IMPLEMENTATION
There are many examples and applications of artificial intelligence in use today.
Klein says AI is optimised in document processing. Computer vision models are also used in self-driving cars. AI is also used in drones for maintenance activities such as electricity and cable repairing, and used by organisations for fraud detection.
The Amazon Go convenience store also does not require checking out or waiting in line.
“With our Amazon Go app, customers enter the store, pick the products they want and just go.”
Amazon Go uses the same types of technologies used in self-driving cars: Computer vision, sensor fusion and deep learning.
The system automatically detects when products are taken from or returned to the shelves and keeps track of them in a virtual cart.
When a customer is done shopping, that person can just leave the store as their purchase will be deducted from the Amazon account and a receipt will be sent to them.
As of now, however, Amazon Go is still in its beta stage and has yet to open to the public.
Moving forward, Klein says AI is definitely becoming part of our lives and a component to make the customer experience better.
“Our vision is to democratise AI, making it available to everyone,” he says.