AI Products You (Most Probably) Already Use

When thinking of artificial intelligence, people tend to readily embrace the wildest fantasies about what it is and how it can impact our lives. But we unnecessarily focus on the bleak scenarios – evil machines getting out of hand and wiping humanity off the planet. There are two problems with that thinking: it ignores the good side of technology, and it looks way too far into the future while ignoring AI is already a part of our lives.

There are many AI products you may actually be using while not even being aware of it. Many of these technologies are serving humans, saving lives, and making them easier and longer. In this post, we discuss the role of AI in everyday life and present and the most common misconceptions people hold about it.

What is AI?

Let’s start with a bunch of definitions.

Intelligence is commonly understood as the ability to perceive or infer information, and to retain it as knowledge to be applied towards adaptive behaviors within an environment or context.

Artificial intelligence is understood as intelligence demonstrated by machines.

This sounds super simple, but also quite vague. Artificial Intelligence is often presented as something that will shape our lives sometime in the future. But many people do not fully realize the extent to which artificial intelligence is already being used today. And AI is literally everywhere – it permeates our lives in ways we don’t even comprehend.

Those who find themselves looking for the next AI revolution may not live to witness one. And that’s because, in many ways, the revolution has already begun, and there is a good chance your everyday life is rife with AI-enabled tech.

FUN FACTAccording to a 2020 HubSpot survey, 63% of people actually do not know that they are using artificial technology.

AI is not a thing reserved to sci-fi movies

AI is not always spectacular by Hollywood standards. Quite the contrary – it’s often plain boring – and that’s not necessarily a bad thing. AI is usually in all the most tedious, boring things people don’t like doing or are just pretty bad at. AI powers the personal assistants that thrive in our mobile devices, automates marketing strategies, recognizes patterns, analyzes images or meticulously scours big data at mind-boggling speeds.

Sometimes AI can be playful and quirky – like those apps that show us how we’ll look when we’re old.

On other occasions, AI is actually very useful and beneficial, making huge improvements to human lives – helping to predict the risk of cancer and the impact of cyclones with 99% accuracy.

Common misconceptions of AI

The term AI has been problematic and people’s relationship with it is not always clear. While we generally appreciate and use technology, in certain areas like customer experience we’d rather interact with a human being than an AI system. When given the option, many people would still prefer to interact with a human being rather than an automated customer service system.

Statistics show that most consumers don’t understand AI and have deeply held concerns about it:

• A recent study made on 6,000 adults (North America, EMEA, and APAC) to find out what they think of AI. It turns out many consumers are excited to embrace the benefits of AI and see a promising future ahead, but many still harbor deep-rooted fears about it.
• 43% of respondents in the survey reported that they were not sure what AI was or how it is currently being used.
• 7% of respondents say they do not know and do not care what AI is.
• Only 26% of respondents reported feeling “great” with Artificial Intelligence.
• 60% had a warm acceptance of AI, allowing for its ultimate future potential but also realizing that we need to be mindful of how it is used. It could be a sign that the air is open to public opinion, only a very small percentage of respondents (6%) say they feel “bad” about AI or that technology poses a near-human threat.
• A Blue Fountain Media survey found that 38% people felt comfortable with how quickly AI was introduced into their lives, while 26% said they felt neutral because they did not realize one way or another.

Also, artificial intelligence has lost much of its original allure and impact, watered down by overzealous marketing or unimpressive technology that claims to be AI when it’s anything but.

The term AI has been seriously abused over the past couple of years, becoming a shorthand for automation. This, in turn, has led to a situation where anything that uses a bunch of conditional statements bundled together is immediately labelled as AI –  and many such applications don’t quite live up to the expectations.

A famous tweet by Nick Craver, Architecture Lead at Stack Overflow

Based on their experience, people have been trying to make sense of AI, coming up with rather inconsistent understanding of its purpose.

What’s the purpose of AI

Most people do not understand that one of the main purposes of AI solutions is to enable machines to learn new things. Few realize that AI is essential in solving certain problems or understanding human speech.

When asked to describe how they understand what AI can do and what the purpose of artificial intelligence is, people come up with really interesting answers, ranging from quite accurate to absolutely ludicrous:

• 35% ability to replicate human interaction
• 51% thinking logically
• 19% play games
• 18% surveillance on people
• 31% replacing human jobs
• 14% feeling emotions
• 8% controlling your mind
• 10% taking over the world
• 57% ability to learn
• 50% solving problems
• 37% interpreting speech

AI in drug discovery

Among its many applications, AI helps to make painstakingly slow processes a little faster. A good example of such a process is drug discovery and drug development – it’s a meticulous, and expensive process which involves months of hard work and lots of trial and error. To learn more about drug discovery and development and the role of AI in the pharmaceutical industry, head over to our blog.

The impact of AI in the pharmaceutical industry is undeniable. Even the most experienced and knowledgeable chemists can’t do what machine learning does in a fraction of the time. With the help of AI, a scientist can now explore 1060 potentially drug-like molecules (that’s more than there are atoms in the solar system) at a scale and speed previously unachievable by humans or any other technology.

But there are areas where the application of AI is accepted with limited enthusiasm. People are generally sceptical of the use of AI in customer service.

AI in customer service

Blue Fountain Media surveyed over a thousand US people aged 18-65 about their true feelings about AI. 87% of the respondents said that they would trust the diagnosis of a human doctor over an AI one. When it comes to customer service:

• 41% say they would rather have their issues resolved by a human agent.
• About 40% of the respondents agree that the future of customer service looks bright with AI in it.
• 38% of the respondents do not believe that AI is capable of delivering the same (or better) level of service as people.
• Only 27% think AI is capable of delivering a comparable level of service.

These statistics show the advancement of AI in the field of customer service is invariably held back by the general preference of human agents. Even consumers who prefer interacting with customer service via online chat still appreciate speaking to a real person. Bots just don’t seem to cut the mustard and the experience typically falls short of expectations.

Actually, it seems AI works best when people don’t realize it works.

Do people know where AI is used?

Most people don’t realize they’re actually using AI. In a recent study, only 34% of respondents thought they had previously used some form of AI. In actuality, however, 84% of the respondents had used AI-powered smart assistants, or predictive product suggestions. The very fact that you’re using Gmail makes you a blissfully ignorant beneficiary of AI-powered tech.

Based on Pega Research

The number of people who unwittingly use AI is striking. In the study mentioned above only 41% of the respondents realized Google Home or Amazon Alexa were powered by AI. At the same time, consumers see the presence of AI in Apple’s Siri (57%).

Most people assume that there is AI behind every voice support system. A recent study by Edelman, shows how the general population understands artificial intelligence compared to tech executives.

What are the examples of AI our everyday life

Although many people don’t realize it, AI is used in many programs and services that help us perform everyday tasks. Connecting with friends, using an email program, or using a mobile sharing service – let’s look at the examples.

Gmail

Google’s legendary spam identification engine heavily relies on AI. Over time, the filter learns from a variety of signals, from email metadata to specific words or phrases in an email. Thanks to its machine learning capabilities, Gmail itself is able to successfully filter 99.9% spam.

Gmail also offers the Smart Reply for canned responses, giving users a way to respond to emails with simple sentences at the click of a button. These responses are tailored to the content of each email. Users can respond by typing the answer manually or choose a smart one-click response instead. For example, if you send an email to someone about an upcoming movie release and they respond to let you know they are interested in seeing a movie, Gmail offers “smart replies” like: Let’s do it!, I’m in!, or I’m no longer available.

The Gmail “navigation” feature reminds you to track emails that you have ignored or forgotten. Using AI, Gmail tries to determine which emails need a response, and highlights them after a few days of neglect.

Social media

Artificial intelligence makes it easier for users to find and connect with friends and business partners.

Twitter algorithms suggest people to follow tweets and stories based on individual user preferences and monitor and classify video feeds depending on the subject. The company’s photo capture tool uses AI to figure out how to crop images to focus on the most interesting part. Twitter AI knows how to identify hate speech in tweets. In the first half of 2017 alone Twitter closed 300,000 accounts, 95% of which thanks to the use of non-human, artificially intelligent machines.

Facebook uses AI Messenger conversations, in its algorithmic news, photo tag suggestions and in ad targeting. Among other things, Facebook also uses a combination of AI and personalization to combat spam and harassment. With the number of users and content on Facebook, it would not be possible to use manual labor to identify potential threats and assist the police in fighting criminal activity.

In November 2017, Facebook introduced an active detection feature to scan posts to find patterns that could indicate whether a user is considering self-harm. The AI-powered system sends mental health services to a person and, sometimes, to his or her friends. Facebook supports this AI program with staff resources such as trained presidents, partnerships with local mental health organizations, and local first responders where appropriate.

LinkedIn uses AI to help match the best candidates with jobs. LinkedIn uses a prediction system not only to determine who will decide to apply for the job post, but who will actually get hired.

Pinterest Lens uses AI to identify objects in images. Take a picture of that beautiful maple food set in your friend’s house using the Pinterest LENS tool, and its AI-driven feature will help find similar tables. In some cases, you may even be able to find a product vendor so you can purchase the item. Pinterest uses a computer concept, the use of AI where computers are taught to “see,” to automatically identify objects in images (or “pins”) and recommend the same visual pins.

Slack

Slack AI uses a data structure called the “work graph” to collect information on how each company and its employees use the tool and communicate with others. The information can be used to train artificial intelligence models and make Slack easier to use.

Slack uses machine learning and natural language processing in a feature called “Highlights” to surface most relevant messages. Slack’s search also uses artificial intelligence to help users identify knowledge experts and channels where they can be accessed based on an analysis of who’s talking about what and where.

Smart assistants

Digital smart assistants such as Siri, Alexa and Google Assistant are among the most popular examples of AI in everyday life. They understand voice commands and translate them into actions.

From getting directions to your destination for lunch to inquire about the weather you will be escaping over the weekend, digital voice assistants quickly become our pilots who can stay healthy. These tools from Siri and Alexa to Google Home and Cortana, use native language processing and AI-powered generators to get answers to you.

There is no question voice assistants will continue to grow and become less of a party trick and more an indispensable tool in our daily lives.

Google search

AI starts working the moment you type a word in the Google search box. Search suggestions are based on data previously collected by Google about you, such as your location, age, and other personal information. Using AI, the search engine tries to guess what you might be trying to find.

The scope of AI in Google services is staggering. It makes ads “follow you” based on your search history and are personalized to help you find the things that the algorithms believe might be looking for at the given moment.

Online shopping

Ecommerce has undergone a major overhaul over the past few years with the introduction of Artificial Intelligence and Machine Learning.

Automated AI-enabled storage management and storage systems help marketing organizations improve their performance, while emotional analysis enables them to stay in line with their customers’ preferences and behavior.

Amazon leads the charge and uses artificial intelligence in almost every step. It shows recommendations for products that should be purchased, as well as improves their logistics chain by predicting spikes in demand and distributing stock accordingly to the nearest locations.

Banking

AI in banking ensures the security of transactions and detects fraudulent activity. When you check in with your phone, get a low balance warning, or log in to your online bank account, AI works in secret. AI crunches lots of data to decide if a specific transaction is “suspicious”, and blocks it if needed, or helps banks to determine which applicants are safe borrowers.

AI dynamically customizes the interface of the mobile application by showing customized alerts such as: reminders to pay bills, suggested cyclica transactions or categorization of payments.

Ridesharing apps

Uber uses ML to predict passenger demands to ensure that surge pricing (short periods of sharp price increases to regulate passenger demand and increase driver supply) will soon be gone. Uber uses machine learning to better predict supply and demand and minimie price manipulations.

Recommendation systems

Netflix and YouTube have great recommendation systems to decide what you might like to watch next. Both These services analyze your previous responses and media preferences, and even dynamically change the miniatures on the main screen to make certain shows more appealing to you.

Music streaming apps also use AI to make sense of your listening habits. They then use the information to suggest other songs. For example, Spotify acts as your music buddy, offering suggestions for new albums, bands and playlists based on your listening habits.

Google Maps

Google uses algorithms and traffic information to determine the best route you can take, whether it’s on foot, by car, bike, bus or train.

Many popular navigation apps use Google Maps technology and AI to monitor traffic to provide real-time traffic and weather conditions and suggest ways to avoid traffic jams.

In Mountain View, California, a self-driving car from Google’s sister company Waymo can take you back to and from work.

Smart home

Components of the smart home like Nest thermostats learn about our preferences and daily routines to best adjust the temperature with minimal human interaction. Smart refrigerators make shopping lists based on what products are inside them.

One of the most common examples of smart technology is the Roomba robot vacuum cleaner which uses artificial intelligence to scan the room size, identify obstacles and remember the most efficient cleaning routes.

Private cars and planes

Unauthorized Aerial Vehicles (UAVs) or drones are a major component of our aerospace, performing surveillance and serving as delivery services for a growing number of programs, including the delivery of medicines and essential goods to citizens detained in their homes due to COVID-19.

While the market for private or driving cars is still very small, there are enough plans to test whether such cars and trucks will expand as the technology of installation and IoT (Internet of Things) technology grows and grows.

Speaking of AI, there is no better and more outstanding display of this technology than what car manufacturers and drone drivers do with it. Just a few years back, using fully automatic cars was a dream come true, however, now companies like Tesla have made so much progress that we already have automotive ships on the road.

Self-driving cars and parking, also known as private cars, use machine learning to identify the area around the car so that you can drive the passenger without having to control the car. Tech company Nvidia uses A.I. “[enabling] autonomous vehicles to process large amounts of sensor data and make real-time driving decisions.” and partners of car manufacturing companies such as Tesla, Toyota, Mercedes-Benz, Audi, and Volvo.

AI autopilots in commercial aircraft are an amazing use of AI technology dating back to 1914, depending on how freely you define autopilot. The New York Times reports that a typical Boeing flight involves only seven minutes of human-guided flight, usually scheduled only to take off and land.

The common AI-related fears, myths and dilemmas

Among the many fears connected with technology is taking over our lives and making us lose independence to make free, educated decisions about things. We can all become normal creatures and can predict more than we would like to admit.

• Over 70% of consumers have some fear of Artificial Intelligence.
• A worrying 34% of people said they were worried about an AI system that could injure or kill someone unintentionally.
• 32% are concerned that AI will eventually take our jobs and significantly affect the global economy. Science fiction hardly tries to alleviate these concerns.
• 25% are concerned even about machines that take over the land. This is the basic mistrust that businesses must face and overcome. This takes time and well-thought-out strategies to introduce AI benefits and gradually increase consumer comfort levels.

Will we really lose our jobs?

Short answer: no, we won’t.

According to a report from the World Economic Forum, 85m jobs will be replaced by AI machines by 2025. However, the same report also states that 97m new jobs will be created by 2025 because of AI. This goes to say that AI will make some jobs obsolete, but create others.

AI-powered machines are less likely to make mistakes

As humans, it is only natural for us to make mistakes. Computers can eliminate human error from certain jobs.

AI does not get tired or distracted

Machines don’t suffer from a lack of rest, personal stress, boredom and never get a hangover. No matter what, their performance is always the same, no matter how tedious or  repetitive a job gets.

AI-enabled machines can do dangerous things

Dangerous chemicals, falling objects or extreme heat – AI-enabled machines can work regardless of the conditions and perform consistently all the time.

Machines are cheaper than human labor (in the long run)

While this point is a bit controversial, AI is cheaper than human labor. It is true that the initial cost is high, operating costs are much lower. It also eliminates all the costs connected with employment – hiring, training, salaries and benefits.

While the above considerations remain true, the hand wringing is mostly excessive.

Why we won’t lose our jobs

It is true that computers may take over jobs which require certain skills, but it will only allow people to better utilize what they’re really good at.

Let’s cover some of them:

Creative thinking and problem solving

Computers work best with repetitive tasks. On the other hand, people are great when it comes to creative ideas. Any work that involves ingenuity or problem solving – think musicians, painters, writers, marketers, founders – will not be replaced by artificial intelligence anytime soon. This also includes all the jobs that require a certain dose of intuition and empathy (eg. psychotherapists).

While it is true that there are paintings, drawings and music made by AI, it is only possible because a sufficient amount of data has been fed into the database to teach the machine what people typically understand as art.

In other words, AI art is based on data rather than intuition and the very love of art and the inner need for expression.

In 2016, a computer was taught to paint like Rembrandt. A group of scientists thoroughly examined Rembrandt’s work pixel by pixel and used this extensive database as the foundation for creating The Next Rembrandt. Over 6,000 facial landmarks were used to classify the relevance and recurrence of specific features of the style. The machine thus learned how to paint a Rembrandt “autoportrait”. The effect was rather impressive.

But who is the actual painter? The 3D printer, the engineers, the algorithm? Or perhaps Rembrandt himself, because the algorithm used the diverse features of his style as input?

Similarly, In 2019, Chinese technology company Huawei announced that the AI ​​algorithm was able to complete the last two movements of Symphony No. 8, an unfinished composition by Franz Schubert in 1822.

But suppose there were no human writers and composers, what data would we feed machines and algorithms? How would the concept of copyright change if AI could freely paint and compose music based on existing art? Should the algorithm also be recognized as a writer, and enjoy the same rights as an artist? Should people pay money for its work?

Computers can’t simulate truly human communication

Many jobs need to establish trust and personal connections with the aim of making people relax and openly share their feelings. Because AI can’t easily assume the role of the teacher, therapist or nurse, there just can’t be a substitute for the unique, intimate interaction created by humans in such roles. In other words, computers, though capable of communicating, lack that essential human touch.

AI cannot develop critical thinking

While this may not always be true, AI is not capable of developing complex strategies or critical thinking. There is this elusive common sense that is especially important in senior executives and business leaders. Many people in these professions may turn to AI to help them think about problems, but in the end, it takes a human to make the decision.

AI still needs managing

In order to have fully automated staff, someone needs to manage the AI. Not only that, but as technology advances, equipment will need to be regularly upgraded, maintained and replaced. There will always be jobs for those who understand the internal functioning of AI and machines. And those who don’t, will need to gain a level of competence in this area.

Today, experienced human data scientists are still needed to frame the problem at hand, structure the data, prepare training data sets, remove potential bias from training data and, most importantly, continue to update AI software to be able to integrate new information and data into the next learning cycle. This is much work, even in seemingly simple tasks. We’ve covered this subject more broadly in our post on gesture recognition.

Machines are not yet self-aware

The concept of self-conscious robots has long been popular in science fiction and Hollywood. From a psychological point of view, “self-awareness” symbolizes the ability to acknowledge one’s own existence.

The concept is important because building artificial intelligence that works and interacts effectively with humans should be able to demonstrate (or mimic) self-awareness. This involves the use of first-person pronouns, ability to describe themselves, and respond to environmental-based stimuli.

Since we still know relatively little about human consciousness we don’t know how to test and replicate it in machines. Alan Turing conducted a test that would determine if a computer could think of trying to imitate a person. The ability of AI programs to pass the Turing Test is debatable and largely refuted – although many scientists do their best to prove otherwise.

The future of AI: self-learning machines

The use of artificial intelligence mostly boils down to computers learning to recognize patterns with human supervision. The future of AI is in systems that can learn without human supervision.

Supervised machine learning depends on the specific data: pictures, sound or text provided by a large number of staff teaching the algorithms what to look for. After inserting millions of labeled images, the algorithms became experts in knowing what they were taught to see.

Supervised machine learning still has many practical uses. It brought us self-driving cars and machine translation. But the scope of real-life applications of supervised machine learning is very limited and in most cases people still do better.

Ethics

The advent of artificial intelligence in machines entails some serious ethical dilemmas. For example, should a self-driving car’s AI avoid a pedestrian but could be at risk of injuring or killing passengers, or should it decide to plow into pedestrians to save its passengers?

The problem demonstrates the importance of ethics in the development of AI-based technology. Ironically, however, AI is used in global justice systems investigating cases and making decisions in a better, faster, and more effective way than a judge.

AI approaches can have a profound impact on a wide range of areas, from legal and judicial functions to assisting decision-making of civil society and legal entities. For example, AI can increase the efficiency and accuracy of lawyers’ work in counseling and litigation. AI tools can support them in writing new decisions.

While this is in the realm of science-fiction for now, it is argued that AI can help create a better criminal justice system. Machines could be taught to detect and measure what is more appropriate, processing exponentially more data (especially precedents) and make better judicial decisions than human beings.

Does my business need AI?

According to Gartner research, 69% of current manager activities will be automated within the next four years. A growing number of businesses will be trying to incorporate AI into their strategies. The arms race may have already begun. Every organization should look at the potential impact of AI and investigate if it can be leveraged to solve the business problems of its clients. In today’s competitive business reality foregoing the use of AI can put businesses at a competitive disadvantage.

Calculating a predictable ROI from AI projects might be very challenging (you can read about it in another post on our blog). According to a study by Capgemini only 27% of data-related projects are considered successful, and up to 85% of AI projects fall short of expectations. Read more about data here.

But a certain degree of experimentation and error is ingrained in making AI an integral part of your offering. The decision whether to implement AI or not shouldn’t be taken lightly. The right use case for AI might not immediately present itself, but it does not they don’t exist.

• According to a MarketMarkets report, the AI ​​market will grow to a $190.61 billion industry by 2025 at CAGR of 36.62%.
• According to Gartner, 80% of emerging technologies will be based on AI by 2021.
• Gartner’s research shows that 37% of organizations have used AI in other forms, an increase of 270% over the past four years.
• The IDC predicts that 75% of commercial enterprise applications will use AI by 2021.
• Believe it or not. Artificial intelligence affects our choices and our daily lives.

While most of this is guesswork – we don’t have the capacity to precisely estimate the disruption caused by artificial intelligence and machine learning – it’s good to gear up for the future ahead and be the first to reap the benefits.

One thing is certain – passing on innovation may make your business irrelevant in a matter of years. A smarter future comes with a degree of uncertainty, but we know it’s laced with artificial intelligence.