Siri, the first virtual assistant with a voice that responds via technologies provided by AI (Artificial Intelligence), was created with a desire to make our interactions with computers more “human-like.” It is presently part of the Apple Inc.’s iOS, iPadOS, watchOS, macOS and tvOS operating system. The software adapts to users preference, searches and individual language usages by using natural-language user interface, voice queries and focus tracking to answer questions.
Siri first appeared in the IPhone 4s in 2011 and has now been deployed on Apple’s iOS, macOS, and other iOS-based devices. It is also available in various verticals (niches where vendors serve a specific audience and their set of needs) such as audio calls and video calls, messaging app, payment apps, contacts, photo search and booking apps.
But Siri did not start out on a handheld device. It began on a more grander and visionary platform.
History of Siri
In 2003, SRI international was awarded $22 million by the DARPA (U.S. Defense Advanced Research Projects Agency) program to develop a personalized cognitive assistant. SRI named the project CALO, for Cognitive Agent that Learns and Observes. Their goal was to make a program that helps military personal with office work and making decisions. It was used for organizing and scheduling meetings as well as providing the necessary documents for participants. In short, it would be a virtual assistant that learned from watching people work.
CALO even made decisions in circumstances when someone important was unable to attend a meeting, they can cancel the meeting and reschedule whenever.
SRI International spun off a company called Siri, Inc. dedicated to developing Siri. The initial Siri assistant had a dry sense of humor and a penchant for bad language. Apple acquired Siri in 2010 and an integrated version of the digital assistant appeared in iPhones in 2011. Apple cleaned it up, beefed it up and integrated it with the iPhones software. They also stripped access to competing websites to prioritize their own services and gave it the voice we know today with multilingual capabilities. Still, Apple keeps improving the features of Siri, despite only one of the co-founders of Siri still works with the company.
According to urban legends, Siri is an acronym for Speech Interpretation and Recognition Interface. This is untrue. According to the Siri team, Siri is an obscure woman’s name in Scandinavia. They used the name because they wanted something short and unusual.
How does Siri work?
Siri is among the top voice assistants in the market today, due to its outstanding abilities using high end natural language processing. It responds to voice commands and greetings and enables the user to use the device without touching it. It answers queries, calls contacts and read messages, sets alarms, check the weather, play your favorite music and much more.
Siri works on two principles. Namely: Speech Recognition and Natural Language Processing.
In this part of the technology, human speech(this includes tones, words, slangs etc.) is typed out into text and entered into anywhere on any program, just by speaking. It is an exciting but challenging task because the voice of one individual is far different from that of another individual; the voice of a male person differs from a female. It’s really handy, and it’s much faster than typing on glass- but it’s not 100 percent accurate.
When you start-up Siri by saying “Hey Siri”, a powerful AI speech recognition system by Apple kicks in and converts the audio form of your voice into its textual format. This may sound easy but it’s definitely not. It is an extremely challenging task because we humans have developed a myriad set of tones and accents that cuts across countries; but also across cities and states in that same country. Some people speak fast or slow or even with a lisp; and all these must be taken into consideration by the speech recognition software.
Apple trains its machine learning models by introducing it to large datasets of recorded voices of large group of people. This has drastically reduced the word error rate of speech recognition engines by more than 10% due to the availability of such large a dataset. The voice datasets are diverse to ensure Siri is able to identify and cater to various accents.
Once Siri has understood the words you’ve said and converted it to text, the converted text is sent to Apple servers where the servers run Natural Language Processing (NLP) algorithms on the text to identify the intent of the speaker. The right and specific task is then sent to the device for execution. It goes without saying, this requires an internet connection.
Natural Language Processing
At the heart of every voice interaction between a human and a computer is the NLP. NLP is a branch of Artificial Intelligence that uses both computer science and linguistics to aid computers in understanding ‘human language.” Human beings have millions of thousands of languages and their other variants while computers have just one language called machine language. All of the processes in our computers and smart devices communicate via millions of ones and zeros to perform a specific function. Machine code is unintelligent to humans, which makes NLP a critical part of the human-computer interactions.
NLPs offer new ways to press buttons, enter text and swipe by leveraging on the users voice. When people talk to a virtual assistant, their request is converted from audio to text and analyzed by the NLP software to determine the intent of the request. Just as computer is rendered useless without set of programming instructions, a virtual assistant can’t do anything without an NLP to provide intent.
At the heart of every voice interaction with an NLP, three basic concepts must be defined:
What is the user attempting to accomplish?
“What is the weather?” There are several ways are user may be expressing their intent. Maybe they want to know the weather for the rest of the day or they might like to know the weather at that specific time. An NLP will take what the user says and, based on hints provided by the voice application, attempt to translate that into intent.
What can Siri do?
Below is a list of some of the commands and questions Siri is able to respond to, and some of the actions it is capable of taking.
- Make calls/initiate Facetime
- Send/read texts
- Set reminders/check calendar
- Set alarms/timers
- Send messages on third-party messaging apps
- Play music/podcasts not limited to specific songs, artists, genres, playlists
- Identify songs, provide song and artist info
- Split a check or calculate tip
- Check stocks
- Play TV shows, movies and answer questions about them
- Translate into other languages
- Check sports scores and weather
- Solve math questions
- Apple maps navigation and directions
- Make reservations
- Play voicemails
- Tell jokes, roll dice and flip a coin
- Control settings like brightness and loudness
- Create and search Notes
- Send money via Apple Pay
- Find files (Mac)
- Search twitter and other apps
- Check movie times and ratings
- Open up the camera and take a picture
Is Siri useful?
The default answer to this question would be a resounding YES. Despite over a billion searches every week not many people have jumped on the Siri bandwagon yet. Reasons usually have been the constant need for an internet connection to take advantage of the full features of Siri and secondly, the security issue of Siri recording private conversations when it has not been activated by the user.
The advantage of Siri far outweighs the disadvantages. Siri can perform tasks such as adjusting the user’s calendar on the fly. Siri works with both iMessage and email. Simply dictate your message using speech-to-text. You can call or text anyone with another Alexa device as well as call anyone on your contacts.
Siri is a voice command app; it is used in making life easier for people who are too lazy or busy to type. You can ask Siri to post statuses or tweets to Facebook and Twitter just by saying, “Post to Facebook or Twitter” and then dictating your post.
Siri is one of the most intelligent voice assistants. It can open any app on your iPhone by request. It can manage your smart home, it can integrate with HomeKit-enabled accessories to turn on lights, change the temperature and work your security system.
Despite Siri not being specifically an educational app, it could soon come to play a very big part in the post-pandemic e-classrooms. Teachers have started using Siri to maintain documentation on their kids. Siri has streamlined the process of note-taking and documenting video conferences with students and parents. Siri can be used in problem-solving by school kids. It can keep track of homework, tests and projects; it can even be used in transcribing lectures for much older students in high schools and colleges.
When used correctly, Siri can be a mathematical wiz. It can solve simple and complex equations.
On the subject of Siri listening to private conversations; Siri only does this to improve its intelligence by getting datasets from the user's conversation. This can be undone by deleting your conversation history on Settings.