By Rishi Sharma
On 8 May 2018, Google’s chief executive officer, Sundar Pichai, unveiled the Google Duplex as a part of the 2018 Google I/O, an annual convention for developers. The Duplex is Google’s latest artificial intelligence (AI) powered innovation that has the ability to make phone calls on behalf of the owner.
Recently, Google has been putting most of its investments into machine learning, natural language understanding, deep learning, and text to speech technology.
Duplex will run as a part of Google Assistant, a feature available on Google phones. It works similar to iPhone’s Siri. Google Assistant has to be given a command, for example, “Book me a table for 2 at Indigo at any time from 7 pm to 9 pm tomorrow.” From there, Duplex will get the job done – it will call up Indigo and get a reservation for 2 people at 7 pm. If a table is not available at 7, Duplex will agree with the receptionist on any time that is within the time frame given by the owner, which is, in this case, 7 to 9 pm.
What sets this software apart is the immense similarity it bears to the human voice. Google has integrated human factors such as “uhhs and umms”, short pauses, changes in intonation, and many more to ensure that Duplex sounds like a real person. A restaurant is just one of many places Duplex can call up.
Pichai said in the Google I/O convention, “Done correctly, it will save time for people and generate a lot of value for businesses.” Indeed, owners can leave simple calling tasks to Duplex, saving time and increasing overall productivity in business firms. Pichai gives an example of how it would do so in his own firm, Google. Google has timings for small businesses available on its search engine. These timings are, however, not regularly updated, especially during holiday seasons. Through Duplex, Google can call multiple businesses at once and get accurate details pertaining to the timings and addresses, reducing the number of calls received by these businesses.
A further development is to implement such a technology into call centers of various companies and even government organisations. This will allow firms to save an incredible amount of money and would also encourage humans that previously took similar jobs to pursue higher level jobs.
The plan has been criticized for its prospects with increasing unemployment. However, a report acknowledges that 1.8 million jobs will be lost, a total of 2.3 million jobs will be created, making a net gain 0.5 million jobs. The report suggests a general increase in productivity in an economy as AI is implemented into simple tasks such as call centers and technologies such as Duplex are examples of such AI.
It is foreseeable that eventually, firms’ AI will be communicating with households’ AI. In reference to the aforementioned example, an AI could call a restaurant, and instead of an actual receptionist answering the call, it could be answered by another AI. This saves time on both sides of the phone – the consumers’ and the producers’.
Bill Gates had said, “The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.”
Whether or not Google Duplex will bring the efficiencies promised, will only become clear when it is released to the public.. The system is still under development and a release date or price has not been unveiled as of yet, although trials of the system will be rolling out within a few weeks.
Duplex is a one-of-a-kind, which stands as a premier model of future innovations in AI. Google’s team has successfully created an AI system that can be easily accessible to the public and can save numerous hours for its users. This is a prime example of how AI software is developing and how other similar innovations might function.