
AI Has Entered Our Daily Lives More Than You Know!
With all the hype and excitement about artificial intelligence (AI) of what is “just around the corner” like self-driving cars, immediate machine translation, etc. For the average person, it can be tricky to understand how AI is impacting our lives from day to day. What might be some examples of artificial intelligence that you are already using right now?
You may be surprised to find out you have used AI on your way to work, communicating online with friends, searching on the web, and making online purchases.
Artificial Intelligence In Work & School
Commuting
Texas Transportation Institute at Texas A&M University reports in 2015 that commute times in the US have been regularly climbing year-over-year, leading to 42 hours of rush-hour traffic delay per driver in 2014 over an entire work week annually, with an expected $160 billion in missed productivity. Clearly, there is a massive chance here for artificial intelligence (AI) to make a tangible, noticeable influence in each person’s life.
1. Google’s AI-Powered Predictions
Using refined area info from Smartphones, Google Maps (Maps) can assess the speed of flow of traffic at any particular time. And, with its addition of crowdsourced traffic application Waze in 2013, Maps may more easily integrate user-reported traffic events such as accidents and construction. Access to vast amounts of information being fed into its proprietary algorithms calculations means Maps may reduce commutes by indicating the fastest routes to and from work.
2. Ridesharing Apps Like Uber and Lyft

Image: Uber
How do they determine the purchase cost of your ride? How do they reduce the wait time once you hail a car? How can these services optimally match you with other travelers to decrease detours? The answer to these issues is ML.
Engineering Lead for Uber ATC Jeff Schneider explained in an NPR interview how the company uses ML to forecast rider demand to make sure that “surge pricing” (short duration of sharp price rises to reduce rider demand and boost driver supply) will shortly no longer be required.
Uber’s Head of Machine Learning Danny Lange reinforced Uber’s use of machine learning for calculating meal delivery times on UberEATS, ETAs for rides, calculating optimum pickup locations, as well as for fraud detection.
3. AI Autopilot Are Being Used In Commercial Flights
AI autopilots in industrial airlines are a surprisingly quick use of AI technology that begins as far back as 1914, depending on how broadly you establish autopilot. The New York Times states that the average flight of a Boeing plane involves just seven minutes of human-steered flight, which is generally reserved only for takeoff and landing.
1. Spam Filters
Your email inbox looks like an extraordinary location for AI, but the technology is mainly powering one of its main features: the spam filter. Easy rules-based filters (i.e., “filter out mails with the words ‘Nigerian prince’ and ‘online pharmacy’ which come from unknown businesses”) are not useful against spam as spammers can quickly upgrade their messages to work around them.
Instead, spam blockers must always learn from a variety of signals, like the words from the note, message metadata (where it is sent from, who sent it, etc.).
2. Smart Email Categorization
Gmail uses a similar method to classify your emails into primary, social, and promotion inboxes, as well as labeling emails as necessary. In a research paper named “The Learning Behind Gmail Priority Inbox,” Google describes its machine learning approach and notes “an enormous variation between user preferences for the quantity of important email. Thus we need a manual intervention from users to tune their threshold.
Grading and Assessment
1. Plagiarism Checkers
Several high school and college students are familiar with services such as Turnitin, a popular tool used by teachers to examine students’ writing for plagiarism. While Turnitin does not reveal precisely how it finds plagiarism, research shows how ML can be used to create a plagiarism detector.
2. Robo-Readers
Essay grading is extremely labor-intense, which has supported companies and researchers to develop essay-grading AIs. While their selection varies amongst educational institutions and classes, you (or a student you know) have likely interacted with these “Robo-readers’ in some manner. The Graduate Record Exam (GRE), the principal test used for graduate school, grades essays with a single human reader, and a single Robo-reader named e-Rater.
Banking/Personal Finance
While the guide explains Machine learning in an enterprise context, your regular, daily financial transactions are also profoundly reliant on machine learning.
1. Mobile Check Deposits
Most banks offer the capability to deposit through checks through a smartphone app, reducing a requirement for users to deliver a check to the bank physically. According to some 2014 SEC filing, the huge majority of major banks depend on technology produced by Mitek, which uses ML and AI to decode and change handwriting on checks into text through OCR.
2. Fraud Prevention
How can a financial company decide if a transaction is false?
In most cases, the trade volume is much too high for people to examine each trade. Instead, AI is used to produce systems that learn what types of transactions are fraudulent. FICO, the company that produces the well-known credit ratings used to ascertain creditworthiness, uses neural networks to divine deceitful transactions. Things that may influence the final output of the neural networks include the recent frequency of transactions, transaction size, and the kind of retailer involved.
3. Credit Decisions
Whenever you apply for a credit card or loan, the financial institution must immediately decide whether to accept your application and if so, what particular terms (interest rate, credit line amount, etc.) to offer. FICO utilizes ML in determining the specific risk assessment for individual customers. MIT researchers observed that machine learning could be used to lower a bank’s losses on unpaid customers by around 25%.
Artificial Intelligence In Home
Social Networking
1. Facebook
Facebook also uses AI, make sure to customize your newsfeed, and ensure you see posts that interest you.
2. Pinterest
Pinterest uses computer vision, an AI application where computers are instructed to “see,” to be able to automatically identify objects in pictures (or “pins”) and then urge visually similar pins. Other uses of machine learning at Pinterest include search, spam prevention, ad performance and discovery, and monetization, and email marketing.
3. Instagram
Instagram, which Facebook obtained in 2012, utilizes machine learning how to recognize the contextual significance of emoji, which have been substituting slang (as an example, a laughing emoji could substitute “lol”). By algorithmically recognizing the emotions behind emojis, Instagram can design and auto-suggest emoji hashtags and emojis.
This might seem like a trivial application of AI. Still, Instagram has witnessed significant growth in emoji usage among all demographics, and having the ability to interpret and examine it at large scale through this emoji-to-text translation sets the base for additional analysis on how people use Instagram.
4. Snapchat
Snapchat included facial filters, called Lenses, in 2015. These filters facial movements, enabling users to add animated impressions or digital masks that adjust when their faces moved.
Online Shopping
1. Search
Your Amazon searches ( “pizza stone,” “ironing board,” “Android charger,” etc.) promptly return a list of the most appropriate products correlated to your search. Amazon doesn’t reveal just how it’s doing this. Still, in a description of its product search technologies, Amazon notes that its algorithms “automatically learn how to connect various relevance features.
2. Recommendations
You see suggestions for products you’re interested in as “customers who viewed this item also viewed” and customers who purchased this product also bought,” as well as via personalized recommendations on the home page, bottom of item pages, and via email. Amazon employs artificial neural networks to make these product suggestions.
3. Fraud Protection
Machine learning is used for fraud restriction in online credit card transactions. Fraud is the main reason for online payment processing being more expensive for merchants than in-person transactions.
Square, a credit card processor favorite among small corporations, charges 2.75 percent for card-present transactions, in comparison to 3.5% + 15 cents for card-less transactions. AI is set up not just to avoid fraudulent transactions, but also reduce the number of authorized transactions failed due to being falsely recognized as fraudulent.
Mobile Use
1. Voice-to-Text
A regular feature on Smartphones today is voice-to-text. By pressing a button or stating a specific phrase (“Ok Google,” for example), you can start speaking, and your phone transforms the sound into text.
Nowadays, this is a relatively routine task. Still, for many years, the right automated transcription was past the capabilities of even the most superior computers. Google employs artificial neural networks to control voice search. Microsoft claims to have acquired a speech-recognition system that may transcribe conversation somewhat more precisely than humans.
2. Smart Personal Assistants
Now that voice-to-text technology is reliable enough to rely on for essential communication, it soon becomes the control interface for a new generation of smart personal assistants. The initial iteration was simpler phone assistants like Google Now and Siri (far replaced by the more complex Google Assistant), which might set reminders, perform internet searches and integrate with your calendar.