Applications of AI inn our day to day lives

Netflix Movie Recommendations To Google Maps: Top Applications Of AI In Our Daily Lives

Muskan
Muskan

More often than not, we underestimate the impact brought about by artificial intelligence in our very own day to day lives. In fact, many people live under the delusion that emerging technologies such as artificial intelligence have nothing to do with them. Contrary to this belief, it turns out that with AI and ML becoming more pervasive by the day, we might be living under the rock if we fail to recognize the way we encounter the technology on a daily basis. 

Following are the applications of AI in our daily practices. 

Reshaping email intelligence

Filtering spam emails

As per a survey conducted by Statista, about 306.4 billion emails were exchanged every single day in 2020. Of this, spam emails accounted for a whopping 47% of the traffic. In such a situation, it becomes imperative for email providers such as Gmail, Yahoo, Outlook, etc., to make sure that the users receive only the messages that they want to, without drowning in the sea of unnecessary spam emails. Many of the email providers use machine learning algorithms to weed out these spam messages. Although it might seem that AI does not manage to filter out every last spam email, a quick glance into the spam or junk folder will show just how many annoying emails we receive on a daily basis which fortunately manage to get filtered out. In fact, Gmail claims that it filters out 99.9% of spam messages using machine learning algorithms. 

Smart email categorization

Many email providers such as Gmail classify all the emails into several categories such as primary inbox, promotions, updates, social and junk folders. Gmail uses AI to segregate emails in their relevant folders. The primary goal is to make sure that just the right ones appear in the primary inbox. This is extremely beneficial in avoiding any types of clutter in the main inbox. Moreover, it’s relatively easy to find emails when they are categorized properly. For example, if we are searching for an email sent by Facebook, say a birthday email, we can simply head over to the social folder and find it there. 

At the end of the day, the main goal is to save the users’ time by having them read only the mails which might be relevant to them. In fact, a research paper on Gmail’s Priority Inbox – a feature that uses machine learning to categorize emails – showed that people using the feature spent 6% less time in reading all the emails combined and 13% less time in reading unimportant emails. 

Optimizing the subject lines

In the modern fast-paced business environment, everyone’s mailboxes are often flooded with thousands of emails. In fact, in order to save time and energy, most people just check the emails based on the subject lines. If a subject line does not catch the user’s immediate attention, there are high chances that might end up leaving that email unopened, even if it could potentially have some valuable leads for their business. Whether it is a cold email, sales email, marketing email or a job application email, sometimes subject lines can be the first and last shot at making connections with a customer or a business. Therefore, it becomes imperative to write a good subject line to stand out from the crowd. Thanks to AI, many email providers incorporate features to automatically understand the subject and recommend suitable subject lines. AI also helps in email marketing by generating appealing subject lines in order to attract a higher click-through rate. Certain algorithms can analyse the performance of different marketing campaigns and use that as a basis to improve the subject lines in the future. 

Apart from the subject lines, the algorithms also enable dynamic autocompletion of email content. This way, we can compose better emails in a much shorter time. 

AI-powered voice assistants

From asking Google assistant about the weather to asking Alexa to play our favourite songs, voice assistants have come a long way from the early inception in the 1960s. As per a report by Juniper Research, there will be a whopping 8 billion virtual voice assistants in use in the next two years. This just goes to show how popular these voice assistants have become in recent times. Under the hood, these voice assistants make use of a subset of AI called Natural Language Processing to generate answers to our queries. 

Most of the voice assistants can do everything from answering questions, performing internet searches, setting reminders, playing music, telling stories, cracking jokes, playing games and most of all controlling smart home appliances. 

Our homes can listen to us

Voice assistants and smart home appliances are transforming the way we live. It’s fascinating to see people talk to their devices in order to control them with simple voice commands. A smart hub integrated with one of the voice assistants can be used to control all the compatible devices. Artificial intelligence and IoT have given rise to a slew of smart devices including smart thermostats, door locks, cameras, refrigerators, energy management systems and smart lighting. In fact, most of the aforementioned devices can be controlled remotely, often through an application. So one does not have to worry any more if they forgot to adjust the thermostat in the rush of leaving for the office. They can simply control the device through the smart app. 

Improving social media experience

AI is used heavily to enhance our social media experience. This can be backed by the fact that we are only surrounded by the content we would like to consume, which keeps us hooked to the applications whether it is Instagram, Snapchat or anything else. Now, whether this is a boon or a bane is a debate for some other day. For now, let’s look at how different social networking applications make use of various branches of AI to enhance user experience and make social media addictive.

Instagram

Instagram uses machine learning algorithms to show users only the content that they seek. These algorithms take insights from the collected data and learn over time to show only the content which is relevant to the user. Apart from creating a personalized feed for the users, Instagram also enhances the platform by identifying content that may violate its Community Guidelines. Instagram warns its users if they’re about to post captions that could be potentially offensive. The AI identifies “offensive” captions and warns users by stating “this caption looks similar to others that have been reported.” This way users can be more conscious of their actions and put responsible content on the platform. 

Instagram also uses ML to identify sentiments behind different emojis. This information is used to auto-suggest emojis to the users. 

Facebook

Facebook uses AI to solve some of the most prominent problems faced by the company and its users. This includes combatting hate speech, terrorism, spam, violence, drugs and self-injury related content. In fact, in the fourth quarter of 2020, Facebook took action on 6.3 million pieces of content involved in bullying and harassment partly due to the use of artificial intelligence. 

Facebook also uses a deep learning-based tool called DeepText to understand the context of posts with “near-human accuracy”. DeepText has the ability to enrich the user experience by extracting information pertaining to the intent, emotions and entities from the posts. This also helps in weeding out unimportant content such as spam. Yet another application of DeepText is to identify the most relevant comments on a particular post. This is especially essential in content posted by celebrities which often garners hundreds and thousands of comments including spam comments. Apart from this, a very famous use case of AI is for suggesting tags in posts based on facial recognition. 

Snapchat

Snapchat is famous for its fancy filters and special effects. Whether it is a time machine lens used to age and de-age selfies or a special lens that can superimpose animations on videos, Snapchat has never failed to surprise us with its filters. Snapchat uses machine learning and deep learning to detect faces, locate facial features and generate 3D meshes to move along with facial movements. 

LinkedIn

LinkedIn uses AI to recommend connections and jobs to users. It also helps in filtering out irrelevant content from the user’s feed. In the background, LinkedIn uses AI to remove malicious content and ensure that the users do not receive annoying or spam notifications. 

Making commute a whole lot easier

Google Maps

Applications of AI in Google Maps

Google maps can evaluate the agility of traffic and weather conditions in real-time and can use this information to recommend the best routes. This helps in reducing the commute time considerably. Users can also report accidents, constructions, traffic and other obstructions that are ultimately fed to the algorithms in order to make appropriate recommendations for the fastest routes. 

Fare estimation for Uber, Ola and other riding apps

Apart from optimizing maps, ride-sharing apps such as Uber and Ola use ML to generate ETAs and up-front fare estimates. In fact, Uber also leverages ML to alert drivers of regions that can have a high demand in the future. Predicting demand enables the companies to surge the prices during the peak hours which drives in more profit. 

Secure banking

Artificial Intelligence is used in our banking systems for a variety of applications. The most important application is to enforce the security of transactions by detecting fraudulent activities. The systems can identify and monitor unusual transaction frequency and size to issue alerts to the concerned customer. Banks also use ML algorithms to assess the risk involved in approving loans.  

Online shopping

To understand search queries 

Several e-commerce companies such as Amazon use AI for a multitude of purposes which in turn helps in improving the business efficiencies. One such application is to better understand what their customers are actually searching for. Not only this, but it is also imperative to understand the context of the search query which can be used to recommend complementary products. This not only enhances the quality of the search results but also optimizes the user’s future search queries. Amazon, in a blog post, stated:

“Once we determine which items are good matches to the customer’s query, our ranking algorithms score them to present the most relevant results to the user.  Our ranking algorithms automatically learn to combine multiple relevance features. Our catalogue’s structured data provides us with many such relevant features and we learn from past search patterns and adapt to what is important to our customers.”

Providing personalized product recommendations

Most e-commerce websites use shoppers’ past behavior to make product recommendations. This is a very powerful technique both for the business and the consumers. The success of AI in making personalized product recommendations is due to the fact that there is a plethora of dependencies between users and items. The data is so vast that it is almost impossible for humans to go through it and make any meaningful recommendations. Moreover, the traditional way of showcasing popular items only on the basis of the number of sales is wearing out as compared to AI-based personalized recommended systems. 

Movie and music recommendations

AI seems to have a command over entertainment as well. What’s better than browsing through your favourite streaming service and having just the right stuff recommended for you? With so many tv shows and movies, it is extremely difficult for a viewer to choose what they want to watch on their own. Netflix’s recommender system comes as a saviour in such a situation. Netflix’s personalized recommendation service generates revenue of $1 billion per year from customer retention. In fact, almost 80% of Netflix views come from the platform’s recommendations. 

Applications of AI - Netflix recommender system

Popular music stream service, Spotify, also uses AI to curate personalized playlists for the users. The engine uses natural language processing, collaborative filtering and audio models in order to make music recommendations. 

As great as the innovations sound, this is just the beginning of an AI revolution as Fei-Fei Li, former Chief Scientist of AI/ML of Google Cloud states: “These are some of the tools and technologies we’ve developed are really the first few drops of water in the vast ocean of what AI can do.”

Muskan
Technology

2 Pingbacks

Leave a Comment

Your email address will not be published. Required fields are marked *