Google and the Machine Learning Product

Google has always been a company that was well positioned to capitalize on the machine learning/artificial intelligence age. A central service that’s sole focus is to organize the world’s data and make it easily accessible is a perfect combination for deep learning. Two companies have benefitted the most from the last few years of breakthroughs in deep learning algorithms, and they are Google and NVIDIA.

It makes sense in this machine learning era to think about Google’s product basically being artificial intelligence broadly. That product manifests itself as search, email, video like YouTube, Google Photos, Google Assistant, and many other current and future products and services. Google is essentially THE AI/ML company. I say that because I don’t see any other company that is solely oriented around AI/ML. Focus, and more importantly, how a company focuses their R&D and CapEx is very telling about their priorities. All other companies who have aspirations in AI/ML are generally treating it as a complimentary feature for other parts of their business or products. But for Google, AI/ML is the product, and because of that, the vast majority of their spending is focused on having the best machine learning tools at their disposal. Given Google’s absolute focus on AI/ML, it is very tough to see any other company be better at general purpose artificial intelligence than Google.

Establishing Trust Remains the Center
If you caught some of the news from Google i/o, you might have seen one of the more talked about demos being one where Google Assistant can simulate a human making a call to make a reservation or book an appointment on your behalf. This demo was pretty mind-blowing and as I watched it I couldn’t believe it was real. This demo was also met with mixed reactions, and for good reason. While the promise of AI has a lot to offer humans in helping us be more productive, it will also take us a bit of time to wrap our minds around the whole experience. Which will lead to the question of trust.

AI is inevitable, but whose AI we trust to be our computing companion is the central question of the next decade. Now it may be somewhat easy to think that because of Google’s business model that most people will not trust Google. I’m not sure that is entirely true. There is a dramatic difference between how the mainstream public perceives Google vs. Facebook around the topic of privacy. In our own primary study, and I’ve seen similar results from other companies research, Google ranks much higher on the trust scale than Facebook. In fact, in our research study, Google and Amazon were nearly tied for third place of all the companies we tested, and Facebook was toward the bottom.

One of the things Google has going for them is their services generally add much more value, or are perceived as adding more value and convenience than Facebook. This is an important point in this conversation. It has long been argued that most consumers don’t mind giving up some privacy for convenience. Essentially receiving what they perceive as great value for free in exchange for some information for example. The big difference between Google and Facebook, in my opinion, is consumer see more value in Google’s services holistically than they do Facebook’s. Which means they will be more tolerant, and perhaps more trusting that Google will not abuse their information. Google also differs from Facebook in that what you use Google’s services for are relatively private. What I search for is not public domain for example, for everyone to see, where most of what I do on Facebook is public domain. This is a distinct difference in the services and a consumers mindset around using and trusting these services.

Therefore, I believe, a consumers tolerance for Google services and their openness to trade some privacy for convenience will be much higher with Google than with other companies with a similar business model. This is a primary reason why I think Google’s AI assistant is a force to be reckoned with when it comes to assistant platforms.

Competitive Conclusions
While I maintain, Google will be the best at general purpose AI, that does not mean Google’s AI will be the only ones we use in our daily lives. While predicting the future is impossible, it does seem highly likely we will use some assistants in our lives for what they do best. We are still a long way off from this reality, and the initial implementations and coordination of assistants will be awkward and clunky at best. But how these digital assistants like Alexa, Siri, Cortana, and Google Assistant collaboration and work on behalf of the needs and requests of the consumer will be critical. These assistants are platforms, and humans have been able to manage using multiple platforms in their daily computing lives already so it makes sense they can handle using multiple assistants to get stuff done.

But a key focus of all of these assistants that I think is important to understand is the positioning Google used that the ultimate goal of Google Assistant is to help you get stuff done. That mission statement is not unique to Google assistant, but I did like how Google positioned it as something can help you get time back.

This, I think, is still the ultimate direction the technology industry is slowly heading and where I think AI will transform how we use computers. We do not, and should not, need to sit in front of these screens all day to be productive. There has to be better ways for technology to increase our productivity, save us time, and give us hours of life back.

When it comes to future platforms and buzzwords like augmented reality, and voice and visual computing, this is the direction that I think will be paradigm shifting in how we live our lives. Technology can enhance our lives and be less of a burden. Work more on our behalf and allow us to spend less time wrapped up in our existing workflows and build new ones that allow us to have more of our life back. That’s a future of technology I’m all for.