Today, many companies make business decisions after analyzing huge amounts of data. E-commerce services need to process data about lots of orders to offer products customers are likely to buy. Marketing companies and marketing departments need to analyze the behavior of website visitors to send personalized emails. Video hosting platforms like YouTube have to learn their users’ preferences to deliver relevant content.
To create value for users (the output), digital businesses need to analyze a huge amount of user data (the input). In other words, they need to apply algorithm technologies.
The market for algorithms is quickly growing. We might soon say “there's an algorithm for that” just as we currently say “there’s an app for that.” And today, you don’t even need to be an expert in semantics or facial recognition to use algorithms in your business – you only need to find the right algorithm that does all the work for you.
Let’s consider why algorithms are being integrated into businesses across the board and, in fact, creating new markets.
What are algorithms and what is the algorithm economy?
Algorithms are nothing new. But as information technology has developed, algorithms have become more affordable and thus almost ubiquitous. Google, Yahoo, and Bing use algorithms to provide relevant search results. Apple uses algorithms to make Siri recognize speech. Apple has even filed a patent for an algorithm that would enable Siri to provide smart replies when a person can’t answer the phone. Facebook's algorithms is constantly evolving to ensure a better experience for users. Supply chain companies like Dell, Colgate-Palmolive, and Li & Fung continue to optimize their operations with algorithms.
And there are tons of other spheres where algorithms are already used. Algorithmic tools are quietly but persistently being implemented around the world. They’re everywhere from healthcare to law enforcement, making decisions that influence our lives.
[The Siri virtual assistant]
Even the recipe for your favorite cake, in fact, is an algorithm. An algorithm, in a nutshell, is a step-by-step set of operations. So how is a computer algorithm different from a list of ingredients and instructions for putting them together?
For a more academic definition of algorithms, we can look to Gartner: “Algorithms are a mechanism to capture knowledge and insight in a packaged form that can be simply reused in a consistent fashion.”
Broadly speaking, a business becomes an “algorithmic business” when it derives value from data and the algorithms that process that data.
Gartner offers the following definition of an algorithmic business: “Algorithmic business is the industrialized use of complex mathematical algorithms pivotal to driving improved business decisions.”
The algorithm economy isn’t about selling smarter apps. It’s about selling tools that enable developers to make their apps smarter. It’s about a new stage of innovation where developers can monetize their best practices and know-how.
Why the rise of the algorithm economy is inevitable
Algorithms have already found their way into a variety of business spheres. Any business can benefit from algorithms, from healthcare to education. The main reason why algorithms have become so widely used is the development of machine learning and artificial intelligence. These technologies make algorithms more accessible to the majority of businesses.
An equally important factor in the rise of algorithms is the emergence of algorithm marketplaces. One of the brightest examples of an algorithm marketplace is Algorithmia. The Algorithmia marketplace provides access to many different algorithms through an API, letting algorithm developers monetize their inventions while offering app developers an easy way to improve their apps.
Algorithmia’s customers include over 60,000 developers, who use over 4,500 algorithms. In 2017, Algorithmia launched a service that allows data scientists to create their own machine learning models, functions, and algorithms and share them with others inside their organizations. There are two ways for developers to access this service: using a Serverless AI Layer for setting up models in the cloud or using an Enterprise AI Layer for hosting the service in any public or private cloud.
[Algorithmia’s AI Layer]
With the emergence of services like Algorithmia, you no longer need to be a brilliant mathematician or a superstar developer to apply an algorithmic approach to your business. You can develop an app using an algorithm in just a few simple steps. Your code can be written in any programming language and then integrated with the algorithm you need through a single API, reducing the time for development, testing, and scaling.
How does the algorithm economy benefit technology companies?
Algorithms enable innovation in various spheres. Let’s take a look at some of them.
Internet of Things
Perhaps the Internet of Things (IoT) is the sphere where the influence of algorithms is most obvious. IoT is becoming more and more important. Gartner forecasts that the number of connected things will reach 25 billion by 2021. This means that growing numbers of previously human-run processes will be automated by means of devices and algorithms.
Think about home repair, for example. With algorithms, we might get to the point where our house tells us what needs to be fixed, when it needs to be fixed, and who we can hire to do the work.
The use of algorithms in medicine helps us diagnose diseases faster and increase the efficiency of medications.
Remember House? In each episode of the hit TV show, Dr. Gregory House and his team had to identify which disease their patient was suffering from. Before finding the right treatment, they made mistaken diagnoses in nearly every episode. These mistakes make the show interesting; but in real life, they could cost lives. What if medical professionals could apply algorithms to reduce the time spent on diagnosis and instead spend this time on treatment?
Prognos is an artificial intelligence platform that uses clinical algorithms to diagnose over 30 conditions. The company boasts more than 1,000 algorithms to ensure early prediction of diseases, reveal gaps in care, and improve medical outcomes.
Another notable company specializing in clinical algorithms is Paige. They’re currently building disease-specific modules to improve the clinical diagnosis and treatment of cancer. All the data Paige collects can be used by doctors to more accurately assess patients and plan treatments.
Furthermore, machine learning algorithms can even predict premature death. Researchers from the University of Nottingham collected data on more than half a million people. Then a machine learning model helped the researchers analyze a wide range of factors to predict the mortality of the subjects.
Human resource tools already use algorithms to sort employees and applicants by roles and identify the most suitable candidates for positions. But algorithms can also be used to allocate workloads inside organizations. In the future, when replacing employees with smart agents becomes a norm, algorithms may even choose the most suitable agent.
PredictiveHire is a service that suggests what specialists you should hire based on the data you provide. Backed by a team of data scientists, mathematicians, and recruitment experts, PredictiveHire claims they can change your hiring experience through their algorithmic approach. The company has recently partnered with the management platform PageUp. This partnership aims to help Australian businesses reduce bias in the recruitment process.
[PredictiveHire predicts how well a candidate fits a company]
Retailers have long been using algorithms – for example, to show customers relevant products through filters and “recently viewed” features. But with the rise of algorithmic business, retailers will also be able to automate pricing, product management, and merchandising.
Target helps retailers improve their product offerings and predict customers’ choices based on machine learning algorithms that analyze customer behavior.
[How retail analytics is shaping the future of retail]
How algorithms are predicted to change digital businesses in the near future
1. By 2022, robots will take over 75 million jobs globally but will create 133 million new ones as advances in computing free up workers for new tasks.
2. By 2023, there will be an 80 percent reduction in missing individuals in mature markets (compared to 2018) thanks to Artificial Intelligence face recognition.
3. IoT will be the greatest catalyst of business transformation in the next three years. For many companies, IoT will drive a significant shift from selling goods to providing services.
These predictions suggest that artificial intelligence, IoT, and robotics are likely to make our lives easier and more comfortable.
The challenges arising from algorithmic business
On the one hand, the rise of algorithmic business promises a bright future. But there are still some challenges we need to overcome.
Lack of contextual understanding
Currently, not all algorithms are able to understand context. Therefore, developers should take care to teach their algorithms such understanding. This is the biggest challenge in the development of algorithmic businesses.
[A messenger chatbot]
Ethical problems with using smart machines and algorithms
The algorithm economy will likely mean the replacement of human workers with robots. Of course, this is good news for businesses because they can focus their resources in other directions. But this creates a moral dilemma as we elevate machines over humans.
Not enough algorithm maturity
Caught in the hype of using algorithms, companies tend not to answer one important question: will implementing algorithms reduce or increase inequality? In Indiana, for example, an algorithm categorized incomplete welfare paperwork as “failure to cooperate.” As a result, one million individuals were denied access to food stamps, healthcare, and cash benefits over a period of three years. Omega Young, who suffered from cancer, was one of them. She died as she couldn’t afford her treatment.
Using algorithms with malicious intent
Algorithms are available not only to benevolent businesses and organizations. They can also be used by criminals and terrorists. Making algorithms do good, not evil, is a massive task for both developers and authorities. We need to work together to solve this problem.
At Yalantis, we’re looking forward to using more advanced algorithms in our projects. And we’ll definitely share all the latest improvements in the algorithm economy.