Will artificial intelligence impact every industry? This is a question I’m often asked. The short answer is yes. The more nuanced and considered answer is that AI’s use in any industry depends on what you are trying to achieve in the first place.

Put another way, every industry has at least one use case of AI, but not every use case needs AI. There are some instances where a skilled human will do extremely well and others where an AI algorithm is the better choice. Others need the best of both worlds, commonly termed “human-in-the-loop”. 

A business must first understand its needs and how it can leverage AI where it is most relevant.

Let’s look at the influence AI and machine learning has in leading commercial industries,  government and other fields.

In financial services, fraud is a risk for both the service provider and the client. Some providers, particularly fintech start-ups, use AI in how they monitor transactions to avoid fraud or money laundering.

Each of us has a unique way we type on a keyboard or move the mouse across the screen, almost like a “fingerprint”. A company called BioCatch uses these data points – in a growing field called behavioural biometrics – and other inputs to train AI algorithms to detect fraudulent activity. 

The company tested the solution with five retail banks in the UK . The results? An average of 83% real-time fraud detection.

PayPal has improved advanced fraud detection algorithms using similar principles. 

Similar AI algorithms from NuData helped retailers reduce the number of mistakenly flagged e-commerce transactions by 20%, thereby reducing customer annoyance. 

The medical field faces a complicated problem called protein folding. Simply put, this is the way protein gets its three-dimension structure, allowing protein to become biologically functional.

All lifeforms are protein-based organisms. Even viruses are simply organisms that have proteins folded in certain way. To create remedies such as vaccines, scientists usually need to understand the virus’s protein structure — in other words, how its proteins have folded. 

However, this presents even the experts with challenges. Researchers are turning to Google-powered AI for assistance. AlphaFold, from DeepMind, created an algorithm that took the Critical Assessment of Structure Prediction challenge to predict the structure of folded proteins. The result?  AlphaFold outperformed 100 other teams. 

Other laboratories that are focused on drug discovery using AI-powered algorithms have recorded similar successes. 

TV, music and movies form a large part of our lives. Studios and social media brands leverage AI for use cases in the entertainment industry.

For example, there are multiple stages involved in producing a movie. AI is used in breaking down scripts, storyboarding and managing budgets. It can also recommend a cast by understanding the summary of what the movie is about and matching it to the actors’ bodies of work.

Some Hollywood studios even use AI to decide what type of movies to make in order to maximise returns and profits. With the proliferation of on-demand streaming, Netflix has algorithms that learn your viewing patterns and suggest more content. 

AI is also used in social media. On average YouTube users upload 500 hours of video every minute and the platform uses AI to create automated subtitles.

Facebook uses segmentation recognition and image or object detection AI algorithms to identify if an uploaded video has sensitive content.

The private equity industry is becoming more competitive, with over 18,000 fund managers offering a private equity product. This is according to Preqin’s 2020 Global Private Equity & Venture Capital Report.

To generate leads and be the first to invest, some private equity firms use AI algorithms to uncover insights hidden in alternative data sources, such as social media mentions and satellite imagery that had previously not been used before in the market.

Natural language processing algorithms on job postings inform private equity firms the possible direction a company of interest is taking – for example, if a company is increasingly looking to hire machine learning and automation experts, private equity firms can reasonably estimate a strategic plan involving rapid digitisation for the near future.

Satellite images of parking lots at outlets such as McDonalds or Walmart help them better predict the revenue of these corporations and inform estimations of end-of-period financial results or value of a stock to decide whether it is under- or over-valued.

READ MORE: Bean Counters in Space: How Orbital Insights Sees the World

With the emphasis on environmental impact, other firms use AI algorithms to consume public disclosures from potential investees to check their commitment to environmental, social and governance (ESG) criteria in their operations and determine how much of a strategic focus it is.

AI also helps private equity firms understand when companies are launching new products, fundraising and expanding into new markets. It can also pick up on upcoming regulation that will potentially affect their investments.

On January 1, 2021, US Congress passed into law the National Artificial Intelligence Initiative Act of 2020, which will empower the federal government to  accelerate AI research and apply it towards economic growth and national security. 

National US agencies that will leverage AI include the National Science Foundation, NASA, the Department of Agriculture, the Oceanic and Atmospheric Administration and USAID, which provides social relief and aid programmes across the world.

An interesting use case is the US Department of Defense. As more sensors are used on the battlefield, the department uses AI powered algorithms to crunch vast and multidimensional data generated by satellite and drone images.

America is not alone in its military applications of AI. The RAND Corporation’s published research highlights the increased investment spend by China and Russia. 

In 2020, a company called OpenAI released its GPT-3 algorithm. The AI-powered solution has many different use cases, one of which is giving the user the ability to describe the high-level details of a mobile application they want to create in plain English and the algorithm writes the code in the background. 

While practical use is a long way off, the fact that one can describe an app like Instagram or Tik Tok in normal English language and an algorithm can not only interpret what has been said but translate it to functioning code is mind-blowing.

The approach is not new. Four years ago, Microsoft already had its own version called DeepCoder.

Does this mean that AI will replace software developers? The truth is not any time soon.

However, programmers do leverage AI algorithms through dynamic coding. Diffblue, for example, uses AI to automate unit testing and check for errors in its codes. Traditionally one would have to create and manually code these tests. The new approach makes it 100 times faster to test. 

In 1985, Dr KG Kempf, who later joined Intel to lead a team of decision scientists, published a paper discussing the use of AI in manufacturing. He argued that manufacturing problems that couldn’t be overcome by traditional data analysis and human ingenuity could be solved by AI.

Back then AI algorithms were largely informed by principle only. In the 21st century, we now have richer data, better algorithms, deeper investments and more computing power to put these AI ideas into practical use.

Although not yet ubiquitous, companies that combine machine learning algorithms and sensors in the manufacturing or engineering process often see the positive benefits of AI.

According to Moor Insights and Strategy’s paper Internet of things (IoT) Analytics at the Edge, using AI to process performance-level data throughout a manufacturing pipeline, generated by sensors on of machines, results in a 28% reduction in annual maintenance spend and 30% reduction in parts maintenance. The faster time to insight helps reduce the typical maintenance window by up to 75%, the report indicates.

AI therefore has a material benefit for a company’s bottom line.

One might be forgiven to assume that AI is only leveraged in coding-related industries. According to Ernst & Young, AI algorithms help the multinational services company identify outliers and analyse whole population datasets.

It cites benefits such as time savings, faster data analysis and increased levels of accuracy. It also produces more in-depth insight into business processes and better client service.

Due to the nature of some auditing tasks, EY has to analyse huge data sets of unstructured data, such as emails. This is where AI algorithms from a field called Deep Learning are leveraged. 

Similarly, Deloitte has an AI-powered system called Argus.  The company uses it to analyse customer documents such as leases, sales and derivatives contracts.

PricewaterouseCoopers (PwC) has named its AI system Halo. The system is used to trawl large amounts data and identify higher risk transactions that need further interrogation. 

In all respects, 2020 was a significant year as we faced a global crisis not seen in generations. Scientists experienced a deluge of scientific papers regarding the Covid-19 pandemic.

At one point, there were 4,000 journal papers being published each week. By mid-2020, over 23,000 journal papers were published, and the number doubled every 20 days.

The White House and tech firms launched the CORD-19 dataset, considered the largest storage for research papers regarding Covid-19. It houses more than 50,000 journals.

Sifting through these papers and coming up with connections and useful conclusions is humanly impossible. Data scientists, therefore, leverage AI algorithms to get to the insights desperately needed in fighting Coronavirus.

Similar to PwC’s Halo system, IBM’s research AI employed algorithms to assist in extracting information — especially information from tables, a tricky task given that the format appears differently in so many journal papers.

After extracting the information, the company leveraged Watson Discovery’s Smart Document Understanding. which went on to “understand” what the paper was about and categorise it alongside similar journals. This helped researchers link publications together and connect results from a vast trove of research.

Today data scientists who have the skills to leverage AI apply their skillset to different thinking about how we can preserve our environment and save humanity.

In February 2021, US finance giant MSCI partnered with Microsoft to use Microsoft’s cloud and AI capabilities to support environmentally conscious investment decisions for clients.

Microsoft has teams dedicated to leveraging AI for ESG in a department called “AI for Earth”, with various use cases —including using AI to track sources of fresh water, algorithms that model population growth and suggest ways to combat the spread of diseases such as malaria. AI-powered recognition algorithms also help conversationists identify and track animal species.

AI algorithms are used in different industries, from the medical field to military and everything in between.

What is important to remember is that AI is not a technology that is reserved for programmers or IT professionals. It is one that each one of us needs to understand 

We need to know how to leverage it, not necessarily how to code it.