top of page

How Do Artificial Intelligence Assistants Interact with People?

Updated: Sep 21, 2023

Medha Singh Yadav,

Editorial Intern (The Indian Learning),

Indian Society of Artificial Intelligence & Law.

As the people are evolving so is the technology. From a reusable rocket to an autopilot car. An example of this evolution is also an artificial intelligence assistant at home. From the beginning, text was the best way to interact with an assistant app (typing in a phrase triggered a response). Presently, voice has dominated. Assistant apps or smart speakers are continually tuning in for their wake words. As a matter of course, the words "Hey Siri," "OK Google," "Hey Google," and "Alexa" are the guidelines on their particular gadgets, however clients can also customize their wake words. "Alexa" can turn into "Echo," "Amazon," or simply "computer." The capacity to create these changes can be particularly useful in the event that someone named Alex or Alexis lives in the home.

Wake words depend on an exceptional algorithm that is always listening for a specific word or expression so a phone, smart speaker, or something different can start speaking with a server to tackle its job. Wake words should be sufficiently long to not to mistaken, simple enough that a human can speak it and clear enough that a machine can recognise it. This is the reason you can't change your wake word according to your preference.

Voice assistants don’t really understand what you're saying — they simply tune in for their wake word and afterward start communicating with a server to finish a job. Natural language processing (NLP) is a type of computerized reasoning that assists innovation with interpreting human language.

Many companies own their own voice-enabled virtual assistant like Amazon’s Alexa, Apple’s Siri, Google assistant and Microsoft Cortana. Like its competitors, Cortana can be described as the next stage in human-computer interaction. There is a neck-to-neck competition among all these companies. These voice-enabled assistant makes our life’s easier for us whether it is to set the alarm or looking for records or exercises. According to Microsoft, Cortana is currently used by 148 million people.

Cortana’s Key Characteristics

Cortana is the AI-controlled advanced aide first showing up on Windows telephones in 2014 before growing to PCs running Windows 10. Cortana is intended to get familiar with a client's propensities and expect their necessities.

For instance, if a client regularly requests that Cortana check the morning traffic, Cortana will start to offer the data without provoking. Cortana's consideration in Dynamics 365 CRM implies that deals and promoting have more devices available to them than any time in recent memory. Cortana can keep up with the client's timetable, set up updates, show client records, make new records or quest for contacts.

The more that Cortana finds out about a client, the more point by point the reactions become. For instance, Cortana can give refreshed entryway tasks to a booked flight, offer knowledge gathered about the purchaser that a salesman is meeting or give showcasing experiences dependent via web-based media destinations.

Cortona can make up a list, pull legit information from LinkedIn like the professional background and the company details and plan accordingly the meetings and everything. It can also track upcoming travel reservations.

Microsoft’s new Play My Emails feature for outlook has given Cortana an ability to read and relay messages.

Microsoft is utilizing Artificial Intelligence and Cortana to upgrade Dynamics 365 CRM

Microsoft Dynamics 365 is a product line of enterprise resource planning (ERP) and customer relationship management (CRM) intelligent business applications. Different devices in Dynamics 365 use Cortana's AI without straightforwardly alluding to the partner. Microsoft has expressed that its will probably dispose of storehouses among advertising and deals while solidifying applications through brought together route and client experience. Clients are shown all applications for which they approach, disposing of information storehouses while keeping up with the client's knowledge of dashboards and route devices. A portion of the highlights that are important to sales and marketing experts are:

  • Client insights provide marketing and sales a complete image of the company’s customers. By uniting data from different sources, creating singular client profiles and examining KPIs, organizations can acquire experiences into their missions and exercises, measure achievement and even get ideas on approaches to further develop commitment.

  • Relationship experiences are particularly useful for the sales group. Utilizing AI, sentiment analysis, natural language processing and data recovered from Dynamics, deals experts can get point by point writes about the situation with their associations with their clients.

  • Significant arrangements controlled by AI can help in an assortment of ways. For instance, experiences may uncover that it is a happy chance to seek after a client for another deal or that move should be made promptly to hold a record.

Microsoft's utilizing Machine Learning (ML) as an approach to speed up the training of these AI models. In 2016 the organization started looking at utilizing field-programmable gate array (FPGAs) within servers as a method of expanding performance level. While a universally useful CPU like an Intel Xeon can be customized to run a calculation, a devoted fixed-work ASIC (application-explicit coordinated circuit) is by and large the quickest execution. However, as ASICs don't take into consideration much improvement in the hidden plan, a FPGA compromises between execution and adaptability. Microsoft has stated that it’s making available the hardware-accelerated Azure Machine Learning models that run on FPGAs.

According to a new Patent, Cortana will use AI and ML to investigate message information from various sources — Microsoft Teams, Skype, WhatsApp, Twitter, messages, calls and instant messages are completely delineated in the drawings — to become familiar with the significance of each message. To investigate message information from various sources — Microsoft Teams, Skype, WhatsApp, Twitter, messages, calls and instant messages are completely delineated in the drawings — to become familiar with the significance of each message. Cortana would score the messages and generate a text summary that would be transformed into speech and shipped off to a listening device. That could be a telephone, vehicle, earphones or keen speaker.


Microsoft has the potential and money to make further innovations and development in the area of voice enabled virtual assistant. It is well positioned to expand upon the scenarios Cortana can accomplish. The size of Cortana will likewise be controlled by how effective Microsoft is at persuading more significant outsider sellers to foster abilities and backing the innovation. Speech recognition is turning into a standard component in big business applications, and advances in regular language comprehension and voice synthesis will give organizations much greater adaptability to pick the best for human-computer connections. There are more modifications and corrections to come in the advancement of Cortana. We can expect to see more enterprise applications and conversational features in the near future.


The Indian Society of Artificial Intelligence and Law is a technology law think tank founded by Abhivardhan in 2018. Our mission as a non-profit industry body for the analytics & AI industry in India is to promote responsible development of artificial intelligence and its standardisation in India.


Since 2022, the research operations of the Society have been subsumed under VLiGTA® by Indic Pacific Legal Research.

ISAIL has supported two independent journals, namely - the Indic Journal of International Law and the Indian Journal of Artificial Intelligence and Law. It also supports an independent media and podcast initiative - The Bharat Pacific.

bottom of page