Machine learning, deep learning, and AI are all hot topics in the IT world, particularly now that computer science has advanced to the point that we are only just realising their full potential. Cloud-based service providers such as AWS, Azure, and Google now offer solutions leveraging these technologies.
In this article, you’ll find out what exactly AI, machine learning, and deep learning are, before understanding how the leading cloud computing providers are leveraging these technologies in some of their services.
Definitions
First, some necessary definitions of these concepts:
-
AI, or artificial intelligence, is an umbrella term referring to the ability of a computer system or computer-controlled robot to carry out tasks that normally require human intelligence. AI has been used in recent years to develop self-driving cars and facial recognition technology.
-
Machine learning is an application of AI that revolves around the concept of computer systems having the ability to autonomously learn how to do things using data alone, without requiring any additional programming. Examples of machine learning applications in everyday life include virtual personal assistants such as Alexa and Siri and search engines like Google and Bing.
-
Deep learning is a sub-field of machine learning that attempts to mirror the functioning of the human brain by forming lots of connections between data and building a neural network from all those connections. Examples applications include a computer system that can automatically restore colours in B&W photos, and the Google Translate app’s ability to translate images with text in real-time to a language of your choice.
Cloud-Based Implementations
Cloud computing continues to grow in popularity as more businesses adopt cloud services due to their cost-effectiveness, scalability, and convenience. According to Statista, public cloud services grew 18.5 percent in 2017, and the public cloud computing market is now worth $130 billion.
AWS, Google Cloud, and Azure are three top cloud providers who happen to be at the forefront in terms of offering solutions that leverage AI, machine learning, and deep learning.
AWS
AWS EC2 is one of the most popular Amazon Web Services (AWS) cloud solutions. The service essentially provides scalable computing capacity, allowing enterprises to run applications on cloud computers without hardware constraints.
Amazon offers Amazon Machine Images (AMIs) with its EC2 service. These AMIs are essentially cloud instances pre-configured with operating systems and components of various application stacks. You can also get Amazon EC2 instances pre-installed with popular deep learning frameworks. Organizations often use these AMIs on AWS EC2 in the context of disaster recovery, providing a way to resume mission-critical app functionality in the event of a disaster.
Aside from AWS EC2, Amazon Sagemaker provides a dedicated platform for data scientists and developers to build, train, and deploy machine learning models. You also can add intelligence to your applications through an API call to pre-trained machine learning services, helping you to leverage machine learning without developing your own models.
There are also services within AWS that use machine learning to enhance the functionality of AWS, such as Amazon Macie, which is a security service that uses machine learning to discover, classify, and protect sensitive data automatically.
Azure
Microsoft Azure is a cloud computing service for building, testing, deploying, and managing applications and services in the cloud. A cornerstone feature of Azure, marketed by Microsoft, is the ability to develop intelligent applications and services using the power of AI.
Azure Machine Learning Studio allows teams to build, test and manage different versions of custom AI models in production. The Azure Bot Service provides a platform to accelerate the development of conversational AI, integrating your chatbots with Facebook Messenger, Slack, and more.
Azure Cognitive Services lets your organisation use AI to solve business problems, infusing apps and web services with the power to see, hear, speak, understand and interpret user needs.
There’s a Face API that allows you to authenticate app users with a selfie image for an extra layer of security, which can come in useful for apps affected by the threat of cyber attacks. Other APIs you can leverage in your app include a Computer Vision API and an Emotion API.
Google Cloud
The Google Cloud Platform is a suite of cloud-based services that use the same underlying infrastructure such as Google’s own end products; Google Search and Youtube. Services offered within the Google Cloud Platform range from data storage to developer tools to networking. Also included is Google’s Cloud AI.
The Google Cloud AI provides machine learning with pre-trained models, and you also get the option to create your own tailored machine learning models. Cloud AutoML enables developers with limited machine learning expertise to train high-quality machine learning models for use with their apps. Google provides its own hardware accelerators to speed up and scale machine learning workloads.
There’s a Google Cloud Video Intelligence API that makes it easier for users to discover videos by extracting metadata and identifying key nouns about the video. A Cloud Vision API enables you to classify images into thousands of categories quickly. You can also build your own product recommendation engine that uses machine learning to recommend relevant products to users in online stores intelligently.
Summary
Each of the three popular cloud providers, AWS, Azure, and Google Cloud offer a plethora of available options for organisations that want to leverage the power of machine learning. Within each platform, you can either build your own custom machine learning models, providing your developers have the knowledge, or you can make API calls to existing models created by the service providers to easily add AI capabilities to your applications.
With cloud providers competing to make machine learning more accessible to developers, the availability of tools and frameworks for AI is rapidly rising. As part of their 2018 machine learning trends post, Forbes predicted that machine learning would soon extend to IT operations, with models using application and hardware log data to find correlations and insights, transforming IT ops from reactive to predictive. No industry has been left untouched by the surge in growth of AI and machine learning, and these technologies look set to change the way we interact with the world drastically.
