Google is famous for its encouragement of technological progress, whether that means supporting companies interested in coming up with a driverless car or pouring its resources into investigating the feasibility of building a space elevator that would make it much easier to send shipments to space and receive shipments from the same sources. (1) However, Google has a particular interest in AI research, as shown by its past successes as well as its present endeavors.
To be fair, AI research is a natural extension of Google’s business interests. After all, the tech titan is most famous for its search engine, which is becoming more and more important as more and more people are making use of the Internet for a remarkable range of purposes. As a result, it has a strong interest in making sure that its search engine provides its users with the most relevant results for their search phrases, thus ensuring more user satisfaction than otherwise possible.
AI is important for search engines because it can make their search algorithms more capable than otherwise possible. For example, while AI is still a distant dream at the moment, its predecessors can learn to associate certain search phrases with certain concepts so that they can provide search results based on those concepts as well. This is important because it makes the search engines capable of recognizing that when a search engine user submits a search phrase such as, “Why did World War 2 start,” they are not actually interested in the search results for that particular sequence of words but searching for an answer to that question, thus saving said individuals a fair amount of time and effort by providing them with what they want in the first place instead of having them test their luck by entering search phrases until they chance upon the relevant search results. In fact, this is the reason that Google’s RankBrain has become one of the most important factors that Google’s search algorithm uses to rank Google’s search results. (2)
However, it is important to note that while RankBrain is capable of learning on its own, its learning capabilities are crude compared to the theoretical learning capabilities of true AI. As a result, if Google wants to tap into the potential of such processes, they have to progress their research into the same.
What Is Google Doing At Its Research Lab in Zurich?
Google has opened its first research lab outside of the United States at its European office in Zurich. (3) This is part of Google’s plan to establish research capabilities in Europe that are capable of matching its existing research capabilities in the United States, as shown by its relatively recent acquisition of a British startup with no commercial products as well as the broad net that it is casting for scientific talent throughout the content. Given the sheer amount of resources that Google is pouring into this project, it should come as no surprise to learn that Google is planning something big.
In brief, Google is focusing on machine learning, natural language understanding, and computer perception. In other words, it is focusing on making machines more capable of learning from their experiences, more capable of interpreting language in the same manner as humans, and more capable of interpreting sensory data in the same manner as humans, which can seem rather basic because we were born with the potential for such capabilities but can actually be pretty challenging to implement because they involve a whole host of interconnected processes that cannot be easily replicated at the moment. As a result, it is remarkably apt that Google’s lead researcher at the research lab called it teaching common sense to AI, which as so many cynical souls have put it, is not actually particularly common.
With that said, this project is interesting because the creation of processes capable of emulating even a crude simulacrum of what is called common sense in human would be an incredible leap beyond the current limitations of AI predecessors. After all, it would make it not just more capable of learning from its experiences, which is an important requirement for making sure that something is a true intelligence rather than something imitating true intelligence, but also more capable of communicating with humans in a sensible manner, which has an enormous range of possibilities both the commercial and other spheres.
What Does This Mean for the Future?
In main, the current research at Google’s office in Zurich is likeliest to lead to extensions of existing trends in the consumer market. For example, Google is already using machine learning for its search engine as well as a number of its other products, but as its capabilities continue to increase, it seems probable that the usefulness of its products will increase as well, thus leading to more effective and more efficient results for its users. Furthermore, as Google’s products become more and more capable of recognizing human speech and then deciphering it into something sensible, it seems probable that voice commands will become the next big thing in consumer products. For those who are having a hard time envisioning what this will mean, just imagine being able to find what they are looking for by asking questions of a computerized assistant in much the same manner as they would ask questions of its flesh-and-blood counterpart.
Of course, it is also important to note that Google’s efforts to teach common sense to its AI predecessors are just a stepping stone on the way to coming up with true intelligence through artificial means. As it makes progress, it seems probable that it will continue coming up with more and more advanced projects that will not just emulate more and more of the processes that make up human intelligence but also surpass more and more of them. Although this is something with troubling implications that have been occupying thinkers for a long time, it is also extremely exciting, not least because of what they will mean when it comes to the sort of products and services that consumers can expect to eventually hit the consumer market.