Key trends in machine learning & artificial intelligence

Archived Development

How Google & Apple are shaping the future

This piece is part of our intern blogger series.

The two major powerhouses of computing are Google and Apple–they are consistently putting out impressive technological advancements and creating a better world for both the customers and developers. There were huge announcements at the two most recent conferences: WWDC 17, Apple’s developer conference, and I/O ‘17, Google’s developer conference. At both of the conferences there were a lot of mentions of machine learning and artificial intelligence. The big question is, what are they and how are they going to help you?

What is machine learning?

To start, what is machine learning? Machine learning is a computer applying statistical analysis to a set of test data to identify patterns that can be applied to other things. The computer is typically given an initial set of data, known as training data. Once trained, it should be able to differentiate between the items or respond to certain stimuli.

The goal of machine learning is to allow technology to be more functional as well as capable of being more flexible in how it can be used. Machine learning is also an important step toward developing artificial intelligence; the adaptability is a necessity for artificial intelligence to be a success. The other functionality of machine learning is to allow computers to act faster, allowing them to recognize patterns in a set of data and make predictions that can help to speed up processes.

The two developer conferences were huge successes, and if you missed them, here is a short recap of what each event had to say about machine learning and artificial intelligence.

Biggest announcements at Apple’s WWDC

At Apple’s WWDC 17, there was big talk on their new operating system for the iPhone, iOS 11, which would include a revamped siri as well as new camera functionality. Apple has worked to integrate machine learning into the camera and photos to increase the quality of low light photos, as well as provide better facial recognition and tracking. The machine learning enhancement also looks to speed up certain tasks on Apple devices.

Biggest announcements at Google’s I/O ‘17

At Google’s I/O ‘17, there were announcements pushing artificial intelligence to the mainstream, even providing their TensorFlow Research Cloud to the world’s top researchers free of charge. Google is really looking to make artificial intelligence not only available to everyone, but also functional for everyone. Google is already working to enhance their email services through machine learning by including Smart Reply, which offers a recommendation of how to reply to an email.

Trends in machine learning and artificial intelligence

Both of these companies have made huge advances in the fields of machine learning and artificial intelligence. They share similar goals in wanting to provide the most comprehensive systems of artificial intelligence and machine learning. Apple introduced the HomePod speaker at WWDC 17 that looks to provide a fresh in home experience for electronic assistants. The HomePod is a speaker with all of the capabilities of siri, plus the HomePod is capable of spacial awareness, meaning it can judge the size of the room and tune itself to better fill the room with sound. Apple also introduced their new machine learning API, Core ML. Core ML is the framework that will allow for enhanced speeds and better camera functionality. Through Core ML, image recognition can occur as much as six times faster than the Google Pixel. Core ML also looks to improve security for Apple devices by keeping everything on the device instead of constantly transferring back and forth with the cloud.

Google has done great things for both artificial intelligence and machine learning, especially with the Google Assistant. With Google Assistant now being available for both iPhone and Android and having more available languages, it is a great tool for a lot of people. As more advances are made to it, machine learning and budding artificial life will be available to a wide variety of people.

At I/O ‘17 Google announced several programs to connect machine learning to new areas, such as finding jobs for people, as well as introducing machine learning to the medical field. With the job searches, Google is hoping to connect unemployed individuals with well fitting jobs in order to increase the number of good job applicants, such as what they did for Johnson and Johnson. The work with Johnson and Johnson showed an 18% increase, after one month, in job seekers applying after using the Google cloud analytics.

Google also revealed to the world AutoML, a machine learning system capable of using neural nets to create new neural nets, which can greatly diminish the amount of time and effort needed to set up an artificial intelligence system. The AutoML could also be used to help no experts still be able to create functional instances of artificial intelligence.

These last two developer conventions have brought huge news with them for machine learning and artificial intelligence. The companies pioneering the way are setting up successful paths for those that follow and use their devices and applications. While machine learning and artificial intelligence are still in their infancy, the future looks brighter every day.

Pierce Cusick is a Digital Scientist Intern and graduate of Alpharetta High School.

The Latest

How is mobile technology disrupting healthcare?

Healthcare providers are seeking partnerships with tech companies to help supercharge their innovation efforts. The result is unprecedented innovation in mobile health technology. Learn how mobile health apps are transforming healthcare.

What is HIPAA and why does it matter?

HIPAA sets forth important guidelines relating to the privacy and security of personal health information. If you’re a software developer building healthcare apps, it’s critical that you understand HIPAA and its Privacy and Security rules.

Defining and building a new data service – a standard approach to data architecture

Converting data into meaningful information requires a system of tools and technology that can overwhelm even the most experienced data scientists.