Google Remains Committed to Artificial Intelligence and Machine Learning


Google CEO Sundar Pichai recently released his first earnings call, revealing a solid third quarter.¬† As the call drew to a close, some investors asked Pichai and Google CFO Ruth Porat about the company’s progress with machine learning.

Pichai responded that the development of artificial intelligence and machine learning are among Google’s top priorities.

Google has been devoting time and resources to the development of machine learning since the beginning of the “Google Brain” project in 2011. Google Brain is one among multiple simultaneous and independent Google-funded projects aimed at cracking the problem of artificial intelligence.

“The progress, particularly in the last two years, has been pretty dramatic” reported Pichai. “We’re thoughtfully applying it across all of our products, be it search, be it ads, be it YouTube and Play.”


One can’t help but be reminded of Google X’s Self-Driving Car (SDC), which has been raising neck hairs since lead engineer Sebastian Thrun led the award-winning project to fruition in 2005. In 2014, Google created a new SDC prototype that functions without a steering wheel or pedals. This newest model has undergone on-road testing in the San Francisco Bay Area since 2015. The cars are to be made available to the public starting in 2020.

In light of the unprecedented technological advances made possible by developments in AI, many are wondering exactly what machine learning is and why it is useful.

Charles King, principal analyst for Pund-IT, has the answer: “In essence, what machine learning provides are ways for computerized processes to learn as they go, thus improving performance over time and through experience. As a result, ML could be applied in virtually any industrial situation, from enhancing the time and results of product design processes to monitoring and maintaining complex machines.”

Although ML has great potential, it also presents major challenges in terms of the level of processing power it requires. For example, in 2012 16,000 computer processors were required for Google Brain to teach itself how to recognize cats on the internet.

google brain cat

That said, the prevalence of utilities such as GPS and speech-recognition for smart phones speaks to the potential payoff for the winners of the AI game. ML tends to lead to products that, once obtained, consumers hate to go without.

Take facial recognition software for example. Social networking sites like Facebook make uploading and organizing photos of friends easier by using programs that can actually recognize who is in each photo and sort the photos accordingly.

There are the occasional hilarious mix-ups of course, and then there was the not-so-hilarious instance in which a Google Photos app labeled an African-American woman as a gorilla. Google was quick to apologize, and a spokesman reported, “There is still clearly a lot of work to do with automatic image labeling, and we’re looking at how we can prevent these types of mistakes from happening in the future.”

Here’s to hoping that Google’s Self-Driving Cars don’t suffer from these kinds of hiccups, seeing as they’re already on the road.