24 Apr 8 Things AI is Already Doing That You Probably Didn’t Know About
Some companies are embracing AI and there are many things AI is already doing, while others practically run and scream when they hear any mention of the term. At least part of the divide comes from those who see how AI can help a business to be more efficient, and those who only see AI as “robots taking our jobs.”
While both of these attitudes are present in today’s culture, a recent survey found 60% of companies had implemented AI and/or machine learning strategies. That shouldn’t be too surprising when we start to look around at many of the things we use on a regular basis. Simple things such as the movie recommendations on our Netflix, or directions we get on our phone from Siri, both use AI.
And AI market-use is only expected to grow. By 2020, research firm IDC predicts, AI influence will push worldwide revenues to over $47 billion, up from $8 billion in 2016.
Before we get to that point, let’s look at where we are now with AI.
I guarantee you there will be at least one thing on the list you haven’t already heard of. Try and prove me wrong.
1. Getting rid of spam
Let’s look at a company you probably already expected to be using AI.
Since late 2016, Instagram has used a deep-learning AI algorithm called DeepText. DeepText can sort through thousands posts per second, and in over 20 languages. It looks at the textual content and sorts out what is spam, which posts are only trying to get loads of followers, and help users to see less of that. Now, Instagram’s algorithm is also being used to detect offensive language and cyberbullying.
2. Screening patients for eye disease
AI isn’t just about smartphones and virtual assistants; it’s also reinventing other areas of life. Take health care for example.
Google has recently been working with Aravind Eye Care System in India to easily screen patients for diabetic retinopathy, a condition that causes blindness. When detected early, the blindness is completely avoidable, but access to regular eye examinations is a big problem in much of India.
AI offers a new hope, however, and Google taught an image-recognition algorithm how to detect signs of diabetic retinopathy using a dataset of 128,000 retinal photographs.
According to Google, clinical testing has proven the algorithm to be as accurate as human ophthalmologists.
3. Helping veterans with PTSD
A recent project by The Institute for Creative Technologies at USC created a virtual version of a therapist, meant to help recognize PTSD symptoms in patients. Though the AI technology has technical and ethical limitations, and would not be used to replace a human therapist, research suggested that people are more willing to disclose their symptoms when they know the data is anonymous. This can have a monumental impact on getting people diagnosed and on the way to the help they need.
Read more about this story here.
4. Making videos on Snapchat
Instagram is only one of many cases AI is being used across social media channels. Another such case is with Snapchat and its CrowdSurf video feature.
Using AI machine-learning technology, Snapchat is able to recognize when multiple users are all recording the same event. Twenty different Snapchat users, for example, could all be at the same concert taking a video of the live performance. The CrowdSurf capabilities take portions of these different Snap videos, and make one longer 10 second video.
5. Picking Coca Cola’s new drink flavor
It’s not surprising that the biggest companies are rushing to use AI to improve their brand.
One of those big companies we all know is Coca Cola, who just used AI-based learning to launch their newest canned drink flavor. How did they do it? By using monitoring data collected from the latest generation of self-service soft drinks fountains, where customers can mix their own drinks. Coca Cola was able to see which flavors customers were combining, and make its own can on the aisles ready to sell.
That’s not all though. Coca Cola is also in the process of producing AI-powered vending machines. Partnering with Facebook Messenger and Pandorabots, Coca Cola has set up a way for customers to chat with a Coca Cola machine via smartphone while on the go. Users can then pick up their drink at any selected vending machine of their choice. Using data from the consumer’s Facebook activity, location and conversation tone, the AI bot adopts a local dialect and persona tailored to each user.
All this certainly takes drinking a Coke to a new level.
6. Influencing influencer marketing
There’s no doubt that AI is being used for marketing purposes, but there is one aspect of marketing you may not have thought of in terms of AI. It has nothing to do with why you click on a picture of a handbag online and then continue to see that handbag and similar handbags for weeks.
It’s influencer marketing. Those celebrities, bloggers, or Instagram junkies who get you do follow a brand or buy a new product.
CEO of Unmetric, Lux Narayan, has said, “It’s imperative that brands have more intelligence into how they associate with certain influencers, making sure that they’re a fit for the brand’s DNA.” Platforms like Insighttool filter through influencers across different platforms like Twitter, Facebook and Instagram, and help brands find the right match for them.
It’s not random of course; it’s AI learning.
7. Taking bias out of the hiring process
A company called Blendoor recognizes that humans have the tendency to hire employees based on unconscious bias, and is using AI to do something about it. The company holds the mission statement “To drive better decisions about people based on merit, not molds.”
Blendoor’s technology can identify where bias may be taking place during the hiring process and communicates a true figure of a candidate’s hiring potential for employers.
One point for AI for helping the job market, not harming it.
8. Stopping cyber-attacks
We’ve already looked at healthcare, beverages and the workplace, but where else can AI go? How about the obvious- online.
Recent privacy attacks online have led to compromised personal information for millions. As a result, many companies are looking to AI-based solutions. Whether it’s insurance companies or retail stores, AI cybersecurity companies reach across numerous industries.
One such company is Cadence Design Systems, Inc., who continually monitors threats to its intellectual property. The company’s CISO, Sreeni Kancharla, points out that security-related data flows at a rate much too high for their 15 security-analysts to ever properly keep up with. AI steps in to back up the human efforts and allow the employees to solve problems while AI finds them.
Did any of these cases surprise you? What other uses of AI have you come across? And where do you find yourself on that spectrum of attitudes towards AI?
Some things are for sure, whether we like AI or not, it is going to continue to be more prevalent. If you run a business, are an active social media user or marketer, or even just love to drink a Coke, you need to think about the implications AI has for you.