Despite being a staple for science fiction fans, Artificial Intelligence remains one of the most divisive topics in the world of technology. Everyone from Elon Musk to Mark Zuckerberg to Jeff Bezos has weighed in on the pros and cons of AI 'taking over the world.' For better or worse, we can find applications of artificial intelligence pretty much everywhere. From chat bots to recruitment, AI is almost all pervasive and growing in sophistication. We don't want to scare you; at present most are pretty benign. However, future advances might not be (especially if you agree with Musk).
In the following article, we'll take a look at 17 applications of artificial intelligence in the world today. Like other lists, this one is far from exhaustive nor is it in any particular order.
1. Smart Cars
Let's start our list of applications of artificial intelligence with one most are familiar with. Self-driving cars are becoming increasingly a reality with each passing moment. From Google's self-driving car project to Tesla's "autopilot" feature, it won't be long before AI is standard issue. In fact, some high-end vehicles come with AI parking systems already.
The Washington Post recently reported that Google is developing an algorithm that could let cars learn to drive, much like your younger self. Granted without the panic and fear of the instructor.
In fact, Google's AI managed to learn to play simple computer games. They plan to unleash it on driving computer games before letting out into the real world. The idea is that it will "look" at the road ahead and make appropriate decisions. It will also learn from mistakes.
Tesla's is a little less advanced but is already being road tested. It won't be long before we see this technology in its full form.
Traditional security camera monitoring is usually conducted by a human operative. Humans, as we all know, are prone to boredom and have trouble tracking multiple monitors simultaneously. And you can't blame them; it's far from exciting.
Wouldn't it be great if you can have a computer do the heavy lifting? Like other applications of artificial intelligence here, AI can be trained using supervised exercises, security algorithms etc to take input from security cameras. Eventually, they can identify potential threats and warn human security officers to investigate further.
Currently, they are pretty limited in what they can perceive as a threat. Wired detailed that they can see flashes of color, for instance, that may indicate an intruder or someone loitering where they shouldn't be. More sophisticated misbehaviors like identifying potential shoplifters are far beyond its capability at present. This is likely to change very quickly as the technology improves.
Of course, there is also the potential for physical patrolling robots, but they can be hit and miss...
3. Detecting fraud
Financial service companies have found some interesting applications of artificial intelligence. If you hold any kind of credit card or bank card, you've probably had a potential fraud message of some kind. It could have been an email or text or even phone call. Many high street banks have offered this kind of service for several years now.
They use AI to detect unusual activity on your account to block potential fraud. You'll often be asked to authorize the transaction or simply say it was you.
In most cases, the AI is "fed" a large sample of fraudulent and genuine purchases and made to learn to look for the warning signs of fraud. After enough exposure and learning, the system is then ready for deployment in the real world.
4. Fake News (Yup, it's true)
AI is increasingly being used to write simple news stories. Wired recently noted that the Associated Press started using AI to write simple stories like financial summaries, sports recaps and fantasy sports reports. AI certainly isn't being used for traditional journalism. However, the systems could help aggregate data to make a reporter's life a little easier. The AI can be very well equipped to churn out simple articles lacking a lot of analysis.
Automated Insights, the guys behind the Wordsmith software, believe that e-commerce, financial services, real estate and other data-driven industries are already benefitting from its services. Wordsmith still needs human direction but the concept is proven and is likely to become more popular with time. Moving beyond data-driven writing will, of course, need a lot more advancement in the technology.
AI writers are not here yet, but it's just a matter of time.
5. Customer Service
There are many websites now that offer customers the chance to chat with customer support. It's one of the more ubiquitous applications of artificial intelligence. You'd be surprised how many of these are not actually living breathing human beings.
No, they are not bored office interns but are in fact a basic form of AI.
A lot of these chat support bots are little more than automated responders. Some, though, are able to extract information from the site and present it to you on request.
The most interesting thing about them is this. Chat bots, given their intended use, need to be adept at understanding natural language. This isn't a simple task. Computers and people hardly think or communicate in the same way. Plus there is a massive variety of speech amongst individuals anyway.
Teaching a machine to do this is no easy feat. Rapid advances in natural language processing (NLP) means they are getting better all the time.
6. Video games
Here's an example of applications of artificial intelligence, you might not have noticed. AI has been used in video games for some time now. You are probably already familiar with a lot of its applications in your favorite games when you think about it.
You could argue that it has been there from the very beginning, more or less. Its complexity and level of "intrusion" into the virtual worlds offered by games have been increasing exponentially for decades.
Many modern games have player performance based difficulty modifiers to keep the game challenging, whilst not making the player frustrated. For instance. Others will learn your behavior, the way you respond to stimuli and try to react in unpredictable ways.
Middle Earth: Shadow of Mordor is a good example. Non-player characters within it that have memories of past interactions and variable objectives. Another interesting example was in the form of the Alien/xenomorphs in Alien: Isolation. The game was marketing as having an AI monster that learned from the way you played the game.
Monster Hunter was another series with some interesting AI. The monsters become increasingly harder to fight and more unpredictable as the game progresses. If you fancy a challenging game series, you could do a lot worse. Yes, I am a bit of a fan boy, no apologies.
Compared to other AI applications, AI in computer games is relatively simple. Owing to the massive revenue this industry generates, there has been heavy investment in perfecting AI in recent years.
7. Predictive purchasing
Honestly, this one is a little spooky. The likes of Amazon and Target stand to make a killing if they can predict your needs. Amazon has an anticipatory shipping project that, they hope, will send you items before you need them.
In theory, it could use your past spending habits for regular stuff and send it to you before it runs out. This could save you a lot of embarrassment on anniversaries or birthdays.
This principle is nothing new, traditional stores have been providing a similar service or years. Tesco, in the UK, has had a coupon service that provides you with offers based on regular purchases for a long time. In the U.S., stores like Kroger, Walgreens, and CVS sometimes tailor coupons based off of frequent and consistent buys.
This AI could come in the many forms. It might be sending you coupons, offering discounts, targeting advertising or stocking warehouses close to you with products you tend to like.
As you can imagine, these kinds of applications of artificial intelligence is far from without controversy. Does keeping tabs on what you put in your shopping cart serve as a breach of privacy, for instance?
8. Smart recommendations
Here's another "breed" of applications of artificial intelligence. Though much simpler when compared to other forms of AI things like Spotify, Pandora and Netflix provide a useful service to its users. They can recommend music, movies or anything really, based on the customer's interests.
Though somewhat similar to predictive purchasing, your historical "tastes" in stuff could leave you exposed to future spending. 'How could you resist that new album or series? We know you like this stuff, go on!'
These AI systems monitor your choices and put them in a learning algorithm. From this, they can recommend similar things you might like.
Most of their usefulness is predicated on your input. This is as true for this one as it is for other targeted advertising in any industry. As intrusive as this might sound, it can help you discover something that might become your new favorite thing.
9. Smart Homes
Our homes are becoming smarter every day. Not just you or the people you live with, but the actual devices within your living space. Many smart devices can now learn your behavior patterns and help you save money.
Thermostats and building management systems can help automate building heating and cooling, for instance. In effect, they learn and can predict when to turn your boiler on or off for optimal comfort, whilst factoring in outside weather conditions. Amazing really.
Future ovens, needless to say, actual chefs, could have your food ready for you when you get home from work. Lighting is another example of a home appliance getting the AI makeover.
They could be used to set defaults and preferences for lights around your home. If tandomed with occupancy sensors, they could turn lights on or off as you move around the house or what you are doing.
They might be able to dim lights when you're watching TV for instance. The future is indeed bright with AI, or dim, or just plain dystopian. We don't know anymore.
10. Virtual Assistants
Virtual assistants work pretty much in a similar way to other AI applications we've already covered. They do, however, conduct a some very particular tasks unique to them. The likes of Siri, Google Now and Cortana are all examples of this technology.
If you aren't aware, though we'd be surprised, you can actually speak to them to help you find something. It could be location, general information, help schedule your day or set reminders, etc. These clever assistants then provide you with the requested info or send commands to other apps, depending on the request.
The core of virtual assistants is their reliance on AI. Cortana, Microsoft claim, can actually continually learn information about you. Eventually, the company said Cortana could predict your needs and get accustomed to the way you speak.
11. Preventing heart attacks
Artificial intelligence can now be used to potentially save your life. They are being used to scan medical data and help predict if a patient is susceptible to heart attacks and strokes.
A recent study even found that AI was more accurate than actual doctors in making these kinds of predictions.
The AI can actually look through the patient's medical records and cross reference them with historical data from other patients who have suffered from heart attacks and strokes.
Futurism found that the AI made 355 more correct calls than those made by doctors alone. This should not really be surprising given the processing capabilities of computers compared to the human brain.
Other potential applications can be expanded to other factors that could prove fatal for a patient. In effect, it could one day predict when you might die.
12. Identifying criminals
Yes just like Minority Report, scientists from Shangai Jiao Tong University have managed to develop a form of AI that identifies criminals with an accuracy of around 89.5 percent.
It achieves this by using machine-vision algorithms. The AI uses still photos of suspects and real criminals all without facial hair.
Not only did the AI identify the criminals with a high success rate but also managed to provide typical facial features that might indicate the person is less than law abiding.
Such determinations definitely brought up moral concerns, particularly whether or not it was just to associate these features with criminal activity. (And again, we're not suggesting judging people by these features at all.) Some of the traits found by the AI included:
- inner corner distances of their eyes
- specific lip curvatures and;
- nose-mouth angles
13. Preserving Wildlife
Wildlife preservation is notoriously difficult, especially when attempting to analyze population sizes or track animals. Scientists, simply, cannot possibly track every animal or tag them all with GPS devices.
A team in Chicago have successfully implemented a form of AI, developed by Wildbrook.org, to do this for them. How might you ask? Well AI inspected photos that are uploaded online and with a clever use of algorithms, analyses each photograph and looks for distinguishing markings.
It can also track habitat ranges from GPS coordinates provided by each photo, estimate the animals age and even determine its sex.
The ran a massive campaign in 2015 that managed to determine that lions were killing too many baby Grevy's Zebra in Kenya. This prompted local officials to alter their lion management program. Impressive.
14. Search and Rescue
Applications of artificial intelligence can be used to disaster events for finding stranded survivors. Traditionally, you'd either have to go out in person and search on foot or look at aerial footage of the disaster areas.
Sifting through footage and photo is very time consuming indeed and any time lost could potential mean someone dies before help can arrive.
Drones are already in use to provide real time footage of disaster areas which still rely on humans to analyze the footage. AI permits the examination of large amounts of data, photos, and footage to help find missing people, sometimes in less than 2 hours. It can even find piles of debris in flooded areas that may have trapped victims in them.
AI can also analyze social media like Twitter to learn about who is missing during disasters.
Michael Walker, a program manager with the Defense Advanced Research Projects Agency's (DARPA) Information Innovation Office believes that applications of artificial intelligence has a place here.
"Attackers can spend months or years developing [hacks]" he says. "Defenders must comprehend that attack and counter it in just minutes."
AI would be a great way of countering this. DARPA held a challenge in August of 2016 that seemed to prove the concept. One of the contestants discovered a new attack in binary code, worked out what it was doing. It then returned the favor and breached the attacker's defenses.
The attacking system noticed this and ended its attack -- in only 15 minutes.
16. Work automation and maintenance prediction
Recent developments in fields such as automated cars will potentially lead to a net loss of millions of jobs. Do not be fooled into thinking this will not expand into other working roles. This is likely to be one of the more popular applications of artificial intelligence in the near future.
This might sound like some impended employment crisis, but that's likely to be an overreaction. This is not unprecedented. The industrial revolution led to a loss of a lot of traditional jobs yet there was not a mass unemployment of the populace.
It will simply mean companies will become more efficient. They will also be able to predict potential problems in their production stream or infrastructure early on. Engie, a French Electronics company uses drones and AI-powered image processing to monitor its infrastructure for instance.
GE reduces machine downtime by collecting and analyzing data from smart sensors installed on its equipment.
The IoT and AI allow and will continue to improve operating costs, productivity and perhaps produce many new industries for all those defunct employees.
17. Hiring (and perhaps firing)
To date, there are billions of devices and gadgets that are connected worldwide. The internet of things is growing at an exponential rate with large amounts of data being generated and collected. Smart sensors, for instance, will likely reach 400 zettabytes by 2018. Applications of artificial intelligence will, likely, play an important role in the future.
Making sense of all this "noise" would be a Herculean task for the best of us, but for AI it's just a walk in the park. Analyzing this information and making some useful insights into it has unlimited benefits.
Sean Captain wrote a book on how this could apply to recruitment. “Can Using Artificial Intelligence Make Hiring Less Biased?” is based on Fama's program for screening job applicants social media activity. This program actually scans peoples social media activity to flag up some warnings about unacceptable content.
Incredibly, around 40 percent of companies access candidates social media profiles in a similar way. Using AI instead of manually "digitally stalking" saves HR departments hours of time.
Of course, this is only as good as the instructions to the AI. 'Garbage in, garbage out' as they say. We are not sure about the ethics of such a process -- especially in the realm of politics which is notoriously subjective and fickle.
The Final Word
Applications of artificial intelligence are with us today and are set to become ever more invasive in the future. Are you ready for it? Some will probably save you life, others might just bring it crashing down. Horses for courses. This doesn't mean you should be overly worried unless of course we inadvertently build Skynet.
Have we missed any applications of artificial intelligence? Feel free to make some suggestions below. Have any scared the wits out of you? Comment below.