AI is often framed as something that’ll change our future, but many people aren’t aware of quite the extent to which AI currently used in society and everyday life.
AI is often framed as something that’ll change our future, but many people aren’t aware of quite the extent to which AI already used in society and everyday life. While it’s important to recognize that AI is still very much in its infancy in regard to large-scale change, there have been incremental advancements in recent years, which have somewhat gone under the radar for those not regularly perusing the latest AI news or working in said fields. To explore just how prevalent AI is in our everyday lives, we have collected five spaces in which AI is shaping consumer behavior and practices.
The entire catalogue of movies and shows at Netflix is ranked, customized and ordered for each individual user in a personalized manner using Machine Learning (yes, you can blame your roommate for messing up the algorithms and data). Through the use of Machine Learning, Netflix is able to forecast, and facilitate your next series binge with five key stages to their recommendation model: ranking & layout, similarities, evidence and search, model improvement and exploit learning. In short, Netflix uses customer engagement to find similarities and patterns in your watching data, alongside the categories you most often visit and even the artwork to which you’re drawn to. These hypotheses are then A/B tested for accuracy and developed to enhance user experience.
Ok, so this may seem like an obvious inclusion on this list, but many often don’t realize what it is that really helps to power Siri, Cortana and their industry colleagues. The seemingly simple command to action transaction is not as simple as it seems, with mere short sentences going through a variety of processes including recognition of wake words, speech processing, natural language understanding, text-to-speech and more.
Many will remember that the initial virtual assistant offerings would often have many misspelled words, as well as an inability to understand the message and the obvious struggles with accents. However, projections for virtual assistants suggest that the already 20% voice search KPI on phones is set to increase, alongside adoption of VA into greater devices including cars and wearables. With the gradual implementation of VA into personal devices, it’s easy to overlook how far they have actually come!
AI being used in e-commerce and online shopping is suggested to reach revenues of over $30 billion by 2025, which poses the question: how is it being implemented? Well, prediction of customer behavior is a crucial aspect of its success, auto-populating websites through algorithms made up of history, third party data, content data and other information to offer the necessary reference to the user.
Alongside these examples, AI is also being used in the collection of data from consumers, listening to feedback, automation of review emails and product suggestions for retargeting. Alongside personalization and product recommendation, e-commerce retailers are now utilizing chatbots to provide 24/7 assistance, made possible through the development of self-learning capabilities and NLP.
There have been many early applications of AI in education, including, but not limited to grading, plagiarism detection, personalization and more. Recent developments in Machine Learning have seen companies like Gradescope using AI to develop assisted grouping techniques of handwritten answers from students. These groupings work on a points-based rubric allowing for accurate marking in seconds, which in-turn, can free up time for teachers and allows for greater lesson planning time, to name but a few benefits.
Current AI algorithms are also being used to assist in the fight against plagiarism. Tools including Turnitin are using AI to locate copied sentences, structuring and other stylometrics alongside monitoring the creation and moderation dates of work. The future of AI in education, according to many, lies in the development of heavily tailored courses and classes for students with specific needs, which may not be as far off as you’d think with education startups receiving over 20 billion dollars worth of funding by the end of 2018.
Ridesharing apps have been reported to use a variety of algorithms both for the benefit of the customers and organizations. It was only as recent as 2017 that Uber had admitted to adapting their pricing for rides based on customer data collected on socioeconomic factors of users. Seen as one of the first major adopters of AI, services like Uber have suggested that AI is central to almost every aspect of their business including fraud detection, risk assessment, safety processes, marketing spend and allocation, matching drivers and riders, route optimization and more.
Ridesharing apps also use personal data in an anonymized and aggregated form to closely monitor which features of the service are used most. Through this, Uber and their competitors are able to analyze usage patterns and to determine where they should further focus developments.These subtle implementations of AI across industries and in businesses may come as a surprise, but we have no doubt that over the next few years, AI adoption will not only become widespread, but increasingly publicized.
Sama provides ML Professionals and AI team Leads with an indispensable solution for Computer Vision data labeling.
